GT ĐÁNH GIÁ CẢM QUAN THỰC PHẨM PDF

01234567
8946
4
9
!"#$$%&'("( (%)$#(*

+,%&#)$&%("&($$&+"((#$-$%(%("
'$&./&+00(&#&(%)$#(
1&0(*&)$%*&&(-2+"(&#-
&""+!$%$&(#0-./+,%&#!%(
0%!(($$(%&.'$(!-$-&""&
0$+!(%)$#(&03- &+!(%
)$#(&(00(&%-&-#&*"( 40$*&%&#%.
56789:;6<=>?=@A97A8=BCA>D9EA7D?8@:;FG@H=88I=H76DA7@C6>=@6JKL8M;@C6:=IADN:9H@C9E@C6
B978?O6:I:9D?B@B9OI=7A68PO=A7>;E99DB9OI=7A68QQQFR?@?:6D6<6>9IO67@A786789:;
6<=>?=@A97HA>>D6I67D?I9786<6:=>E=B@9:8P9769E@C6O98@AOI9:@=7@M6A7N@C6I69I>6=7D@C6A:
I:6I=:=@A97=7D@:=A7A7NF
S T$U31VWXW2
0
94594Y 778946
4
929565
Z[Z \]^_à^]]_bcdcade .......... Z
V.V.V f'(............... Z
V.V.g h#0............. i
Z[j kaâlm]_no]_pdqpddrld
^sqdtdp^_ ............... u v'$&("(%)$#(-! &$%
V.g.V f""v-........... w (&$"("!#%*$(-!4
V.g.g f)x$%.......... y (("(&"((&&(#0 (&#
V.g.z x"")v- ........... { &#.U(%)$#((0("
V.g.| v}$f(-0Sx$%~#
&(v............. #"( #0#0("#0 (
Z[i laa^]pr^lldd]^... Z ("((&&00($$%+-""
V.z.V f"""(0h3- ("+&&%&( "(0(,#(
h(& ............... Zi (#0 (.x#*0(($
V.z.g f"""(0v&($(&#
&-U%0 ........... Zw (%((""((&0$)&(
Z[u `nn]_r^]l̀a^].......... Zy )&0(&#"#$"(0(((&#
dsdd]d .................... Z{ &)$(*"((&*&0-+(#
(%("(&#.v'$&!
(0)$%)!&+%x0*-+(*&
($VW*&0(4)+#+
$&+%h(3(!$.1g2*U(&U&$
1g|2*&h$-&$.1g2.v$
(#$!(30&(%$
.v.¡!$*.%0*56789:;¢<=>?=@A979ER99D*£((&Uv4U*
f¤V.V¥¦W¥XV||VW|XX§V*Ü-U©ª#h&*¡¡}gV
0 1 234567894
63
. "
! :
" #"""$% $ ; ##"
"$ % ".#"
! # #
&" % % <6#
# "!'#'(' # #"'
" "#'
'#''' )& #$#
& '*++,-. % =##'
"##$ # ' #
"(/ #
0 . 1#& / $
. 2. " $
"""" ! ' #
# %3 $ 4 "'
"!5&"" # ("
"# > $
##( 0 6#
6#' '""
" 7 #" $$ $
$ ."
#&# $ ## 1"6
#7 # 6#/6#'
'61 #$
6#$ # " 6' ! !$
# ' #
8 !3# $
& # 7 # ''
##'"#' #' 3"" # '
$ " #" #6#' !
6#$! ."
96' $ '4#5& ###
"8"$# 6#' #
$ %
$ # # . $
&# $"# ' '
8 "" ."
8#06#'$ (
# #
## $
6 #$## ( .
"16#" #
# $!
$ ##.8 #
" 6# 1$
$## #
8 # $ # %.#
# '##'
010 234567894
6337 5

!" !,!'
#! $%!&'()**+,- E!*& &,)
.)))&!/0 /&$'(
)$* //*$&/)$)"
!' >?/@AABC'(/!!
)! !*$)
121234567895:5;< ,&'
(*//H/"
)/!!)!'
+&)!!/)/'=, F /!!)
!!)&) !)'N )I!
!! !!!)!& )/.)
&*&)>?/@AABC' ! )!&$
D!/!!$& /)/!'M/!"
!!$'E!)) !&
,$$*/) !/*)!'(
'F ))/&!) )//
$&))* !'J?*!K
&)$#!/'F /&!)'F,"
!) ! )/" & */)//H&! &
!!* $$!)#/!$!"
'G&/!) * 'F
$//H&!/)! /)&&)0!)/$/)/
$ &" )*$!/
! /)/'F !& &)$)! /'
$&/'($&" E!*)*!*)!
!//H* &*FEE//HO",'
#)I)!!*/ P!)&/
'(! !!&!!)Q R
!!!' (#'
(!!!!!)!'E F /!!))
!!!&$ 0*&) !)
)/)!&)!J)K *)"
&)0$/)// )!)'M!*&
)/)/&$ )))/!!*"
!$'($&" !!$)!'E/
!!!!& )/&&)
'F !!$!) FE!*)*'S!!
/) */)' *!//!!"
L$)/$'M)) 0!0*&*,
!!&& ) 'F !/!!&/)$
&)!* )!!))" ,!!)!*)I!
/)/'E* ! )),$)
#/))I!! !!&'F !//!)"
!)//*$))!' !&)! &/
E)) )$0*!
*H*)'F ) #'+!!))!,
/*!)//)! /!/$&))!*RT
*&$*!/ !)!'
0 1 234567894
63
23$4(56768%1 !
!! "! ! !!
!#$ ! ! ! )
! !! ! $
!!%&!$ ' !
'$ ! %* !!
! %&$(! ! ( '29'(
$'( !' !( 566:8%+! !
!! ) %* ' ;! !
! '$! %+
! ! $ 3$2566<8$ !
$( ! ! #! !
! %+ )!!$ 5: %
!$
%,! !
!!! (
!! "! =>?@ABCDEAFGHIGJKLGEMBGJKCNOPNEOO
!%
-( QHGBBOBDRPOBCSOCNDKB
$ !!
!! ! %.!$ , !
! $ !' ) %.T
! !'%.! $$ '2UV!
' ( $! (
! ! ! $ ( ($ %W#&1(
! ! %* '$$ 56578%, T
) ! ( )" !
$ ! %* (! )
($ ! ( !
$ ) %. ( $$!! !
!'!!! !! '
$! '%* ! $!! !%,$
! ( ! '! ! $ ($ (
!!' % T( ! $!
* ! ) %X !
( $ !
! %/!'! !"! '
!!)! %+ %,!!!
! ! ) ! %- ( $!Y T
) ! %0 $
! ! T '
!! ' ! % ! ! !( ( (
+ ! $ ! ( $ '!%
! 0! !$
! !%+ ! %+!(
! !%1!! ;! $ ! !
! ! $ ! %,
!$ ) ! ) ( $
! ! ! 2 !8
012 34567849

85
68
555756675
!"#$$%#&'("%"" AQ7STU!#* #&4:#V+AQW68'X&"..+:
#&)&)*#+,%##-&.$&..%"#. $%.:)*".#*/#":"#"$2
/&"$0%&.1$...&.*&%..+"! %:#.))&'Y%.:%/#.;$;"
"#-/&#%#!$#/!*)*#+,%#&2 .#*$)*#*&""'(/!.*2
#-+#&&/.-#&.34&#&4+ &#))&.:%/&))* &..&
56678' ".#/-" - =$/!"#&'
9"%&.&.!-#%#&*".*$.# X&/:.+"...-$*#!#.#*#&.
.)*#.%*&"&,%.:".#/."#; .&Z%.)/-#%#&+&.%"#"!
.)%.&&%.!#&##*.#"' $.....%)@&.*&#&#/.'(&"
<%"):"#:&..#&#$%.*. *%$2"))&.:#.-$##/%
)*$)#.#&$/*.&%&&"$#2 ".#**&..)$%$..),%#!
# =$&).&.!.$#..-" #.>6 & 3U!#* #&4:#V+AQW68'X&"%[
!#.))#&&.%*$%.#"+#&". $%+#)&.#*$:#.-&#&"&:
=$&.&.#/'9"$*#!&&)#&! ..#*$.'\&)"..#*$.*#"")2
.&.!-#%#&.$#..&.%"#". &:""":#.)*#))&$%+
*".#$$$##&.:",%.&./& /#"$..'9"$#$#&:%!*#"
#.;#/%"$%&".'?".#.&+ ".#*$")&+:"#"#&
..#%.%#!#..@#&"$*#! $/#/!)&2"#)'(.&"#&.+#$$2
$%$.#&*.-#%.'9"!$.).&.! &)".#/-"#=$/!"#&
.&#**&!%.+#":"#))&# .&.-&)#$-#/))&
#&#"%.&$#$#&..%.&))&2 /:&$%.'("$$%#))&.:#.
#'(.%**#!)""*#&!$.).&. "$#*$#.&+&:""$#$#&.:%/
-&&9#/A'A' #.;".:""):$%.:#..&
*&&.&#-&#/%'U#!%"
)#"#"$#&.].#&&.#.$@
BCDCBEFGGHIHJKHLHMNFJO #/%+"...-!.&.-))&.'9".
"**&))&..#.":&&?'A'A'
4*$))&.."#-$-&-!%.)% &
9".*$..&.!..*!#*$#&.: #$$#&#&#&:.$#%.#!'9!$#!
:""#&!$$/))&=../:&: #.*&#&.: /&%:"5W[76
!$.)$%.'9".#".*&#&.. $#$#&.:""#-/&.&)".&.!
.*$))&.&$%.'(&#!... #%!**&$%))&.#&:"#
%.%#!/#.&".#..)),%&.#&$2 )#*#:"".$%.'9".&#!$2
$&.3%&&"#&:&#&.:.8'?*" -.#&#,%#.#*$.V)%*&&#
..%.+:&)))&./#.&"$$2 .&.!))&.'\)&#$#..$)*
&.)$.&.:"##/".#.$% :"".$&&.#$.&&".&.!.
!)*#*&#.).*#& $2 )#!'X&$#+"$$%#!)"....%
%.'(#..=#*$)"..:#."#& ".*$!)###&#!..'4#.##/.-
$%+%.&"P#.//:.#&&" )*"/&*#./%&-"*&*%*&%*2
4##*...&"AQ76.3R*#&9+ /).$&..&&%.#.#
^_àbcdc P#..@#&).*".&.&.!-#%#&
P#.. e%.&)&. 9!$). U#&."##..
f.*&#& ($%.$$/!))&&#&!:#! 0(&#!1 4&).&.!#%!+&.
*"+.**.#&
f.$- R:$%.))&.$@.&.! 0(&#!1 4&).&.!#%!#&*-#&+
"##.. #&""!#&
())- R:: #$%.;:""$%. 0R&1 4&)$%.+%&#&
#$)
0 1 234567894
63

!"!
##
#$
%&#' #*-)&&
#$()) ?+$
*&+#' S!#+Q'
!&),-) +&O+#+#+
#$ )##
# +#$
S++%#+
./0/012345678692:;<=>323 ),*$D#
T@()GHUJ
+#%,)
(?) P! O+#%&
,)#+' ,%O+##KV$!GHUW!
)#$(# SQQ-$!GHMLN$(,!(P
-*#+)$(% A%! %P'
*CB#?* #'&Q'
@+A% +#DE$ ##*
F#GHIJK!GHLMN$ +*$('
(#**+##' +##)+
+OP&)O++ &-*O*$@P#!#+
#&++ #O+ #),#
#,&)# #$A+-O
+#$ ##)+)$(P#%
() ++P+' #*Q#%'
#&Q )#!#
O+ #'&$
#) X##*+##+
##$(+**) )#&$DSBR
+$R## )GHMJ!###
P#?K&*!!N #+)*)C
*#+! Q # #
@
+
A
%
B
012 34567849

85
68
555756675
!"##!$%!&!!''" +$(%&!!)&"#$%!(-
##&$'(()&*+'&&'&&,- !"&$%!''"&'&#%+0A ,#!)&7C'&
&+./&&!0)12345067$' &78'&=< $&"C'%+&&$@8$&7&7$%$''067
9+&$&&$:'%$#&$;!"'$' 9:; ( #!('+%7!"'$'8+!%'$'&(#7#'1?T
=<
'7&./&/$!)>??4509:;=< #%+' 1>8!!-&$$$$+!')878$&&&7
87$!"( &7&$&$'(7$! $@'(&7&'@$#%&$%8$&7,-
'%7+',#$&!'$@''&&$'&$- #!'0J&'$&" (%'&,#!$("'%!#$&'
%!!"'''+%7'!"'$'($%067$'$'+ !'@$$'&%7$H+'0V&&7+&
$#&*+@&'(#!$'&''&&$'&$%!&'&- (&$!$(&$&7&%#$$&7$'
$@)$%&'&&&7@+#$'=<%+''$%''+' ,#! $$&7&&7$'$'!"!G-
#%+'(&7A!BC! &70D&7$- $@&&7#+%&\'&,&+]^ $@7&(
&$''%$#&$#%+'8&$%7$=< H+!!"&$!''"!"'$')#7#'8$&7'#-
'##+!$&")'+%7'&7/#%&+ E&7 &&$#!067 !&$!"'!!+ (
.E$!@&!0)>??F5&7&$%!+7$@7@( #!$'&'.K '5$'*+'&$C+&&7$!!
%!$&$(#!$'&'( $&'$&"'%!#$&')+%7 (%!$&$0/$%&7"7&$&+'
!$G&76,&+BC!0/&$!!&7 '%7'7 &&$+&'%!'$'$$! ) $%$'
#!"7"$&%7$H+'&7&$%!+'(&+' !8'&&$'&$%!#8 &'&''$&$$&"
(&7$+''%$#&$##%7'.I$'&$)122150 $&$$'#$&((8 '&$'.(8 &
6""#+%&!#&@+#'+'7"$ #$&'# #+%&50/$$! ,#!'(&,&+)^-
##%7''&7&@'(%7"##!"&&7 )(@%)&%&$!!"''%(+$
#+%&' '+%'(#&$%+!%#"0 E$!@&!0.>??F50
:'%$#&$!"'$'7'#&&7'&%-
#7'$$(&$''"!+&$&!0J&
$'##!$%!&&7%7%&$K&$(8$$-
&"(#+%&%7@' '%7H+'&$'$( _àb̀cddefghiejekghlm
#+%&!#&067$(&$% !&
&%'+%%#&%$(&$&$'&+- 67&7$* %!''(''"&'&'$'&7'&7&
&!'+'"'('&&$'&$%!&%7$H+''+%7' &&#&&H+&$("&7@(!$G$@ $'!$G$@
@''$% !&$0
;,#!('%$#&$!!&( &,&+ 67 (#+%&)%!!7$% ((%&$&'&&7'0
''''&(%G$#+%&$''78$6!10>0 & '&'&$@7&(8##%7&&7$'#!$'
67#+%&$'''''&$((&&$$&!'$ '$(&7$ (
( # #!%7$%@!&&$#+%&'
'%! #(%(&7*$&"(
'#&'067#!8$&7%7$%&'&'$'&7&&7"
&"$(&$+&&7@$&+(!$G$@
LMNOPQRS :'%$#&$!+&$(%G$'T&,&+&&$+&' $'!$G$@( '#&'0;7$'&$%!!G
B7' ;&&$+&' U%7' $&7$'%!''(&'&'8'&77$%'%!!#&
/+(% <+@7'' /&7T+@7 &7Y0/0;"AW&$J'&$&+&$&7!&
B&$%!' VT" 124?'.n'&!0)12oo5067$'&7#$!-
:"'' D$!"T" %2-#$&'%!( !$G$@8$&7%&+&!
A$'&$& A%&+$!$&"
X''
W+!"T$&&!
/(&T7 %&@"&&#&&#+%'%!#$&!!'
B&$%!'$K /!!T!@ 8$&7'&7& #'&#'"%7!@$%!!"H+!
A$'&%78 :''' ;$"T' '&#'%7@'$7$%&0J&78')$&8'
Y$($&"(%78 IT+ '%!8$&7 +!-!$G##&$'87'H+!$&!'
W788 E$'&+'#&$ VT+%7 8+!!&'&&$'&$%!!"'$'0
W7'$''('' Z'T%7'$ ;,#!(&72-#$&'%!$''78$A$@010>0
6&7#%G$@ VT+%7 6"#$%!!"7$%&'&&"8+!$!'#!
[$&&$''
<'$+! D$!$''
VT+%7
:"T$!"
(3oT1o?%'+'878 @+! +''(&7
B&$%!' VT" #+%&067&'&8+!$!'!!&&$-
W7!G" V&%7!G"T"%7!G" ' $'(&7#+%&%+%&$'%&!
!%&$ ''"&'&(%$!$&"067!@ #!'$K
0 1 234567894
63
3%/+ 1#C2'%$
%&%''%&%3
%2+ $4%
+%+%%&%%:
%D%%#
EFGFHIJKLKMNOPQRSTUPVWMPQXNYZ
[KO\]\^K_SMYZIK\N\
%%3%%&'
+4+4`'
3#,"%;#=223
!" %%623%+844`'
!"#$%"&'(#)#*%+&, 3'"#3A"$%&
-%./0 %%+%-%1$' '%++ 42%&
2% +%"33# +"%%%33%+#
*"3#.+%8
2%%%&33%+"
333'%"'%4&3 %a.3$%+#
'%3%+ .832%+%!
2%+4%3% 2%4%%''%a*
2%'&#%' +%#@2!2
%&!3%"+32+& 2%4'"%&%%
!33%&3%$3%5+$33% 32%a%'&
%%6'%#.+&%'%& %%3%#)++%%+
%43%"3%+%" 4`'8333%%8%
%3%!"%!"%# /b2-$;=1#%%4
7%!%3%&2%& +32%%%8%% %
2&"2 '%+%%%"+'%'
'83%"%& +#c3++%`43%&
3/9!2:$;<=1#'+3 3%&%%
%'"%3 %"A3&33%+
24%:2%%42 4&%#"+&%8%
5%+&"3#*& +4%3'%$%'2233%
"+%+8% $%'2%%83&2&
"/-%>12 ''%4 3%+42%
3%4#% 24 +!"A4
'4%34!? 84!#%&
%%3+%& 4 % +3
2%5%+&"#. %%224%%
%% %2% %'%%%3%+#
%&::2&%" %%2+%%%%#
2%+%%'#@33% %&"''&
""$"$5%+3%44& 3%%+4
%4+3A#B&"2% 3%%%&#
3%+%&'%2 %3%&4'2
%&'2%3%+%& &$&%&"%+
012 34567849

85
68
555756675
!"#$%&'#!()*
+,&#-'-,&,%+()$&()./$',-
)&#,+$0% 10,"&(2,'$)-
%,',$%"#34,'&()'5-(66,%,)&
',)'%7&,'&+,&#-'$%,
',8,"&,-.9(+(8$%-,"('()
:%",'','$%,+$-,():$),8('&
',8,"&()5',&&()*4:%,':)',
'"$8,'5()"# '()*
,;:,%(+,)&$8-,'(*)'5
'&$&('&("$8$)$87'('5$)- &#,%
&$'<'()-,'(*)()*$',)'%7
&,'&=%,:%()&,-!(&#:,%+(''()
6%+>$!8,''5?@@AB.
$)--,'"%(:&(2,+,&#-'$)-$66,"&(2, %#,-)("&,'&' &,'&'5:$%&("(:$)&'+4'&1,"#',)"$%,64887& ()'4%,

'4"#$'&#',()282,-()$'',''()*")'4+,%8(<()* &#$&&#,%,'48&'!(88*,),%$8(G,& &#,::48$&() 6
%:%,6,%,)",'=>$!8,''$)-C8$$'',)5?@@AB.D%&#, ()&,%,'&.K$%&("(:$)&''#48-1,6%,34,)&4',%' 6&#,
$)$87&("$8&,'&'5:$),8('&'$%,',8,"&,-1$',- )#$2()* :%-4"&5'()",&#,7$%,+ '&8(<,87& 6%+&#,&$%*,&
$2,%$*,& * -',)'%7$"4(&76%&#,"%(&("$8"#$%$"E +$%<,&$)-!(881,6$+(8($%!(&#'(+(8$%:%-4"&'.F#,7
&,%('&("'=&$'&,'5'+,88'5&,;&4%,'5,&".B 6:%-4"&'& :'',''%,$')$18,,;:,"&$&()'$)-$6%$+, 6%,6,%E
1,,2$84$&,-.F#,7$%,6$+(8($%(G,-!(&#&#,&,'&:%E ,)",!(&#()!#("#&#,7"$)6%+$) :()()%,8$&(2,&
",-4%,'$)-+$74)-,%* *%,$&,% %8,'',%$+ 4)&' &#,%'(+(8$%:%-4"&'&#,7#$2,&%(,-.
6&%$()()*5-,:,)-()*4:)&#,+,&#-.H)&#,"$', F#,$)$87&("L#,-)("-('&()"&()*(2,'%(',& '+,
6-,'"%(:&(2,$)$87'('5&#,7$-:&$)$)$87&("$86%$+, #(*#87(+:%&$)&%48,' 6&#4+1$)-'+,!$%)()*'
6+()-56"4'()* )':,"(I"$':,"&' 6&#,:%-E $14&+$&"#()*&,'&+,&#-'$)-%,':)-,)&'.H&('
4"&$'-(%,"&,-17&#,'"$8,' )&#,(%34,'&())$(%,'. 4)!(',& $'<&%$(),-:$),8('&'$14&&#,(%:%,6,%E
F#,7$%,$'<,-& :4&:,%')$8:%,6,%,)",'$)-#,-E ,)",' %!#,&#,%&#,78(<, %-('8(<,$:%-4"&.F#,7
)("%,$"&()'$'(-,5$'&#,(%01(' )87&':,"(67!#$& #$2,1,,)$'<,-&$''4+,$-(66,%,)&5+ %,$)$87&("$8
$&&%(14&,'$%,:%,',)&()&#,:%-4"&$)-$&!#$&8,2,8' 6%$+, 6+()-$)-& :8$",:,%')$8:%,6,%,)",$'(-,.
6',)'%7()&,)'(&75,;&,)&5$+ 4)&5%-4%$&(). D4%&#,%+ %,5&#,7#$2,)&),",''$%(871,,)',8,"&,-
H)")&%$'&& &#('$)$87&("$86%$+, 6+()-5")E & 1,6%,34,)&4',%' 6&#,:%-4"&5' &#,7$%,)&
'4+,%'()$)$66,"&(2,&,'&$"&()$+4"#+ %,()&,E :$%& 6&#,&$%*,&::48$&()& !#("# ),! 48-8(<,
*%$&(2,6$'#().F#,7:,%",(2,$:%-4"&$'$!#8, & *,),%$8(G,#,-)("&,'&%,'48&'."++ )$)$8*7
:$&&,%).8&#4*#&#,(%$&&,)&()(''+,&(+,'"$:E #,%,('&$)$)$87&("$8()'&%4+,)&.M 4! 48-)&$'<$
&4%,-17$':,"(I"$':,"& 6$:%-4"&=,':,"($887(6 *$'"#%+$&*%$:# %$:N+,&,%!#,&#,%(&8(<,-&#,
(&('$1$-54),;:,"&,-5%4):8,$'$)& ),B5&#,(%%,$"E :%-4"&5' !#7$'<74%$)$87&("$8-,'"%(:&(2,:$),8
&()'& &#,:%-4"&$%, 6&,)(++,-($&,$)-1$',- =OJP$#)75?@Q@B.
)&#,()&,*%$&,-:$&&,%) 6',)'%7'&(+48$&()6%+ C )2,%',875:%18,+'$%(',!#,)")'4+,%'$%,
&#,:%-4"&$)-,;:%,'',-$'8(<()* %-('8(<()*.F#(' $'<,-&64%)('#2,%7':,"(I"()6%+$&()$14&:%-E
""4%'!(&#4&$*%,$&-,$8 6&#4*#& %-('',"&() 4"&$&&%(14&,'.C )'4+,%')&)87$"&()$))E$)$87&("
6&#,:%-4"&J'':,"(I":%I8,.H) &#,%! %-'5")E 6%$+, 6+()-14&$8' 6&,)#$2,2,%764GG7")",:&'
'4+,%'$%,,66,"&(2,$&%,)-,%()*(+:%,''()'1$',- ) $14&':,"(I"$&&%(14&,'5")64'()*'4%$)-1(&&,%
&#,()&,*%$&,-:$&&,%) 6:,%",:&()'.H)'4"#")'4+,% &$'&,'56%,;$+:8,.H)-(2(-4$8' 6&,)-(66,%+$%<,-87
01 0 234567894
63
! $"
# ! ! " ##
!!"#" "!" "%."
"!$" " "% # !!
"% """ "" ,""$ " "#
&%%$#! ' > % %'
" #% #(! "" #!# !"
"% "" " ,"%%% "%'
) "*" " "%"" "- !
% " " % % "%
%%+ # !' ((" # $""'
+ %% , " %" "!
" "# >!" "
%%+ % - " !C"""%'
"" .!% " ""
% "% " " $
/#$!#" % !+' % !!"%(
' ! "" " " %
"! " 0#$ #&""#$ "
!" "% % !" % ""@% %
# '# "%!" "% "% %
1!2"3 456789# #"" !% % #
% !! "# "% >
:";" ! <,=2456779 " "% #: <
% " # > "" "%" ,"$
&>&&&>&# $' !" D%"%
# (" $ " % " # " !
#%"( " % % %% %'
" %"!" &> "
&&# $ " " ' $ '
# "%! $ % ! # 4 %!
# % "" > "92")34566E9"
%" "% !"%" !
?# %!" % !" " #
#' %"%>
>" % % %'
" ' % % %$
% # % %
"% ""
@ "% "
A"% ##
""" !%"" FGHIJJKLMNOLPQRSTUVWPKKXMOYXQRPZV
> "#% % [NON\
! ""'
0#$% " $
" %% " "' /%" "%"
% " $" "%
%$ # $B' $#' "- !" " $"'
">"" " "% % %""
012 344567896

5579
898 00
!""#$% &'($$)'" ( '$& 0%$(##(#$"'&"
""( *#&+!' #(""# *#!&( &*'!(#$"*#$!&#()*)#"'
#$"'($("'",-$($!""#$"'($" !$"'(",G!.%"&'$!#$"!#" &
!."$"##)# &'($"$"""("/"( *'!#$&(#$$*$*#(%5" "$2
#"% *0(,-$"'($"(#(#! "!!#"#!"$*0!#$"!$!$%"
(!$#!($#'#$ && *"1$"*$ &'$!#$ &1 #"#$&(#$!$!#*
* (#(#!" &#2!+#$% $**#"*!.%#$%(#$*#"#$
!!'#$#%#!3'#*"#."#) ('!'"!$ ),G!.%#$%$*#$#$.(!'"!$%"
!'"!(#!##$#$% &3)(#",4 " #$3) (*' 3)$"& ' &
#($5#$"'($""""($"%#))'" *'!$*"(#("$"& & &&23)"#$
(#""$#($!'!""6#$2 *'!,7!.%" ")""$#($2
#$ &"$"0#$!'($#$# # 0#*#)!$%"5$#*#'"
"$*#$%,7#$#"#$"*1$"$" &&!" &#%2!E*!#$"5$* (#!#
#$'$*%$#$ &"$""&( ' #$&"#$"$* $'#"$!",
*,-#"(""#)2*#"#'*!"" 7"$""#"!$*'!* "'*1"
$*!('#$$%#$5! &#*&" & *'!($#'#$"1#!!#)*!$%"
$!%$##$,-!(""$")'#$ '($ ")",-$#""$"5"$")'#$
".!(1#"$#"$*0#$# #"#$"*##$" &"!"#!"5 *2
&( &&$!,8$"0#$!#"#$*5 "$! &"!#$#+!"!%5("
%#)$($#$%1##$&( &&$!5)'* "!#&#$"#"1$*#&&$$%)2
#) 0!#$"$*!$#$))#$%#$ "#(#$%#$%'$"$" %$">"#!
&('#"#('$'" "9'$##$'",:#$ &"!"#!"B$*'($"$">
;'*%($"$**" '*,7'"#" "!%#!B,C&$5$!$$*#!0!
<!#$ &!#$=$"#("#(''"$* 1"$"!$%1#"&'$!#$ &#$%2
"$">4#%*,5?@@AB, *#$"5!"""5!.%#$%5##")*#&+!'
C$'($"$"*)#*"( *2 * " "#$!& *"$*!$"'(*'!"'"'
"&1 !$"'("#. !#)$* 9'#!(0""(",:)"$*("*$*
! & **'!"#$#&,D!!5$2 '$!(0(#0'" &($)#!(#!",
E5$*#$"$"* &( *#!#$" -$&("#$%#$($#$%#
'1 *'!")!$%**'#$%*2 "'&+!#$$"1 "$"9'"#$",7$!
'!*)($%(,-$& *$*!$"'( #$*)($#" !;'*%
*'!"#$*'"#"5"!$%"#"&( $#"$"#(!1#*#"!#$"5!(#$%
#($&!"6#$%*#$"5!"""5$*!.%2 *"5$$"$**#%#$%5$*" $,:#$5
#$%,F&'!$"#*#$#" &$1*2 $"5"5$*$%' &*'!*)(
'!%"5#$ 1 *"#""&#&5'1( $"$#) &( " 1 1#
!$"#*"&"## $"!#!" &2 '*'!,8 #""('$!#$'
!""#$%5#'"')""#) $>'" 1!$"'("1#)#1*'!"!#'$*
!$"#**'!"0"* ('3'!'#$5 ( $'!$*##$",
#%2!E* 0#*#$5(#!#!$(#$#$5 HIJKLMNOIMP#".,-& '!( &"$2
$* <'""=B,-$%*#$!$%"#"& ""#"&!.$1$$**#!5#"$
$'(&"$",7(#$*'!*#() $* !$*'!&()'#$,Q$&'$5
*'!9'#5 *'!!"" &*'!#$5"#(2 '""""" &$9'"* &"$""2
!'"!#$"' &1 (#"" #$%%'#$#$*'"#"#$%,7'*$ &
!('$)#,G!""#$%!$%"#.1#"#" '"""'#$""#""&( )$$!*
&(( #()9'##$(" &"$2 *'!*)($"9'$!"5!*##$"5
"5$'##$5(#!#%#!"##&!"5 (*"# ! $"&&((#$
*'!!"" #()($'&!'#$%*'!#)2 !" &'$0!*&#'",R 1)5"$"
#,G!.%#$%!$%"#"&( !$"#*#$" & "#" $"'"&'"( '$ &*'!#$#$
*'!"## 9'#&!"5,%,5!#$ '$!#$!!'",-&#"$ '$!#$5
01 0 234567894
63
:!79 ) :&# / ;#
! #&
" " +4)'/
"" "# 7898.% &#<=
# $% # &#! "(",# -
& # ## % ## "
# # "&>?% #
#& # '! 0" !
"# "( 0 #"0
& # & &
) &# # #! # = +7898.
% *###& 0" @
# *# +,-.
*## /0 +7.) ' !
*# 0 #
*# ! #+ . #
,- ## & 1# / +788A.
B)4
2 C #
*# *#*#! # & "# "
! ,## &# /
&& 0 # (#
## *##
"# ! & &/
*# 3 *# !
" & '&#
# #% " #! #
# ## #(
'# !
"# # )## ' #
# # &#
##&# ( #! D# #<# ) )#
& &&( ,# ;
&# + . & ( " #
# (# ##
" +E.; & # =
% #!#! !
&# #*#& " #
4 # # #!!#!#
# #&"! #-# +4)'/
# ##5"( :78?9EFF9GH#7887. &!
" #22" ## ! 4 !
# # & "# 0 #
##6 # !I&J K# !
- # &# #! &
# *## " 0
"## ##
'# # +A.: 0 ' !
% 78984)'/ - # & *#
012 344567896

5579
898 02
!""#"$%"%& U+(-"%"-#"+("!#(*(()$
"%'("%(""!))(*(!(") %(9((%#!"()(%%&'+(!"()
+(!,(#!(%(%-(%$%(%") )(%$%"))($#+')(-$!+-%
!"(%. "*"",(%/$()(%&!(%%(%1("("
/012*(""),+("#"%.3"( )-""$(#%(%-+"$(%.U+(("#
($4()+"!(#(((%"-+"$("%"# $")()!+-%"+"$(%%(%-(%!%(%+(
#+(%"/#)#("%(%,(( (%%((#!%-+!+-%"+F$.
)%")'+(-"(+$-1.5(%(" /R1G(%-("%.=)%")%(!)
/677819(%() '+(+(%#:(%(%- %"()(%".U+((%(%-%
(%("+%,($!(#()("%("$ (%%)(%")("%"$+""(
%"%+""-,((!()*(# %%%+"%(+"((")"%F$(##(%&
%(("%&")((-%(%-%(%% +(,((+(-"(!(+((%%#%(
+"*(%"()(;!(+%"("#(%("+/# %-(%%")!*)(#()V)$(%")
(;"!(&<,("")=(-&>??7@A((".& ("%()%%"))(!$V%
>??7@5(,-")=B%&>??71. ,(%")%"%"%$:"(.
/C1D(%*$)%(!"(%,('((","-") /71G(%-()".2)%(%#%(%-)""")
:()%)(%.E+(%("+#(",(&)("()& !(!('+(9(%%(%-(%%#((;!((
")!(%(""-"(+)%+(%(%-", "%'("9(%%.3"(()+(
"-&%(""-!()$:()(% %!()+-,('((""-"")+()
(%%"-,(%.5""$((%,("'"( (%$/(.$.&)%"*(%%!(#(((1
#+(!("##"%(!%*(($"*((%% ")+'(;!"$+%)##((('"%"%"
#"#(%$%(9((%"()&.(.&# "%F.W(+("F#')(%!(")"$
%+%"(")(+((%$%(9((! %(%-%((&+("%F#%(%-()"%
"F($"('!).G(%-(*"" %'+%)"-&")"%(%-!#(%%"%
%!("%%)%-)"'"-%+"*(( ,(",((;!"+(""(,(+)(%(+
%)-+((*(#(",('(("," )%")"(+(!"(")$#
-"):()(%%&,"!)(%(%-!$" %(%-(+$-%(%-%(%%")
'))(!()+(F%+%%%(. ""$(%.
/H1E)*)")##(((%.G(3"(I%("&"$'
$("(+"%"()+(#"+"+"
!"(%%"()("&(+"$(",((" XYZYX[\]]^_^à^b]_cdef_g^h\ì
%$%(%.2"+(%'+)##(( j^b^f_ake^hkclb
!+-%$"(9!(&)##((#"(%#(#
(((&)##((",(%#%")""
"(&"))##((*""(%(%.J% J+(+"($(+((##(*("#
"(;"!(#)##(((%!+-%$-&'(+"*( %(%-(%%(%+((%(,"(#%(%-
+($'$("(%!(:"%"%K%( )""+%($(("()#+((%("+(+)%.
L,)(%%(%M%!(:+("!)% O,(%""%()(+("!!"(%"-#
"$!(%%'++('%("%(%(%# %(%(%-%((%%+%()(),-
%(/N-((".&>??H@O(".&>??H@ "F($(%("+%(*(%.A'(*(&%(!"
P-%F")Q",'%&67R01.E%+),( )##(((%(;%"%%+'U",(6.8.G(%-(%%
%!%$+"%(#"-+""(%%"( "("%"'"-%)()",)",((),"%%.
)#:#(*("()!"(%%(*""(") U+"%&!))(-%%"-,%()+(
("$(((/N(")S+%&677H1. +"+("#"+""'%+(!)
/T1D("$%(%-)##(((%!)*"",(%. ,((*""()+(!!("($-/(.$.&)
U+%%("-+(("#%(%-%(( ,("F#"%(("1.E"%&"F($(%("+(%%
)%"!"(.A'(*(&"-!))(*( #()(*((;!(!%","!)K",(
!(%)%#:(-**(+(%(%- "%&")*(%$"$(-&"#"&
%!("%%+()(-$(%("+9(%%. "-+(#"+""-((+(;
01 0 234567894
63
&'()*+,-')-.(/0.-12*()34'+)
!" 5!6!78$67!!9
#$"%!$ :7;87"6<7"!7#$""!!6"
="66766"9
>?6!87"67!7$!9
@776A6667!!!6
BC+D'-+')'C+21-')-.(/E2*(2'F-C(GH*+F+*G32--')-I
5!667!;96$!766!!6
!6!"J"J7"!?6!
>?6!77"!!6A!"7"67!9
K8"66"!!67?6!
797"%$67!!!666; 8#$"%9!$7
9LM9MJ899!!! !$#$!"6MS$68!8
66NM b8ML\]]hN7$8$668
O$689! 7L\]]hN!?"6$
667#7$687 M
66MO$7!P! =" 6778$#
QJ#7 $8 ; 678$ "6M^;
L97J6!9J6!%99!$9N7 !7!7 "67!768
66!$; #$$$$"$76"!
8J7!!!6R!M 9"6!667!!$!;
S$ 7""T$ R!9 !MS$"!887#$$
!97?6!97" !"; "$7 68"$R!
6?!!6"MS$"6 !!669MS$67!76$
687%7"7$7 9$%# $67!"9
U9V 7# 967!M k$"%7%# $67!"
5!6%6$?6"7 !"669$!!6J
9!#$#$$$%$6 69J779"?MO$!67;
#MS$9$!!; !J9"6"!
!$878!"6M57! #$8$6"M
" !!"6?##$!" ^""$#9#966;
77?6!!LWMJXYYZ[: !R!9LmJXYdaNMn#J
7J\]]Z[77^#JXYY\[ "$$ U9$V7
MJ\]]Y[=T7_!`JXYYa[b"cTMJ $$U#9MVS$"676
\]]d[e"MJ\]]d[_ 87`fJ\]]Y[ 7!7!77
5%7gJ\]]h[^$6$7MJXYYXAXYY\NM 6!6M`?"6J9 "%
>?6!!!"!; !$"8!7!7$6;
#7#$?6!77"!7 #$6?66!$
77$!7#$#!!J 667!!6MS$"8J$
$!97!#$?6!" $%6!$J88"7$$67;
L^97J\]]Y[gMJ\]]a[i" !!7!7Co-'+-1',1Cp'Cq+'CG,
MJ\]]d[jMJ\]]ZNM5!%99787 'rF+'))'G)*4'F*).-.p'C--.-3G')766%
"#!k79"L= 8!Mn#J87!;
MJ\]]Z[=TMJXYYY[>%9MJ\]]lNM "!7!7"6967!
^$66"8!877 J#$6!9!!6
!!6;77"%!$Q 7M^$!$"";
M6"9"78"77 666!$779
$ "67!MS$!8 $76M
012 344567896

5579
898 0
!"#$%&'()$!'* ?=63<E32C;39>89M3<2E3?A32H1?25P563;39256^2;957B
C4<73:34<>8@L3865<E39^2>89M3<=6GL5686_8P82=6
+$(,")-'(!./0#)%# <E3G?C3<E8<7=765<D59M25D3??<E8<G8A3C2<E3
1234567893685:8;;8936<2=>=?89=<@<5236259@3A8?B C;L5=5F9C?@2
=<5C2;52<B=<65<32H`C8?=<@YC7G=6G>3<E572893
C8<=56=2D=<E<E3<987=<=568?;957C4<FC8?=<@G987=6G 6=FC32EC8A=<3L37<59323894E2C;;59<;95G98>2HIE3<34EB
2@2<3>2<E8<C23236259@49=<39=8HIE3G987=6G5:8G9=B 36<=a4G95C6732N 36D=73?@49=<=4=_375686C>L395:24=B
4C?<C98?45>>57=<=32=28E=2<59=48??@=>;59<86<=6JCB TUUfdg^h8E56@b? 882236867c8D?322PQRRTde98M3P
364356<E3>5A3>36<<5822C934562C>3925:FC8?=<@ K=73?3<8?HPQRSQVP8R?<fER5dCiG8E<
PQ 6GL596867eC6M?3@PQRj[d
2<8678972=6<E3:5572<E3@;C94E823HKC4E<34E6=FC32 636<2=6=67C2<9@8678G9=4C?<EC393N @2<=??E8A3<E3=9;95;5B
O57@:3?<3<8?HPQRSSP
D393D=73?@8;;?=48L?3<52=>;?3;957C4<22C4E82JC=7 TUUSVH
>=?M867LC<<39NO57@:3?<3<8?HPQRSSPTUUSVPDE393 373:34<=736<=a48<=56=6FC8?=<@G987=6G3>;E8B
86=738?;957C4<45C?7L3?89G3?@8G9337C;56867<E3 2=_3IE
73:34<2<E8<45C?789=23=6;559E867?=6G867;954322B 732429 55<48C232N3HGHP5X=7=_37J8A59VDE39382<E3
=6GG8A39=23<5D3??BM65D6236259@3::34<2HWC9<E39 <39>2<<579 =
; =A38;;9584EC232>5933?3>36<8?2=6GC?89
=>;3<C248>3:95><E3:84<<E8<45>;3<=<=56245C?7 48C232Hk6<3E234 49=L3;3943;<=56298<E39<E86<5=6:39
L3E3?7<53X8>=63DE3<E3965A=43YC7G32B=6B<98=6=6G <=A3868?@2=2;8862335 :5X=7=_37J8A592P<E373249=;B
45C?7>8<4E<E35;=6=5625:3X;39<2HIE=2=2>C4E N5=?@P;8=6<@P867a2E@V2G=E6<C?>= 2386C>L395:<39>2
=6<E3<987=<=565:?=A32<54MG987=6GZ8@5C6G;39B L395:FC8?=<8<=A3?@7=::3936<23X6=27589<@3
435 =5648C23286C>B
25645C?7YC7G3845D8679343=A38D89728<82<8<3 65<8L?37=::393643:95> >8=62<938> 2:3:63245<29H@3 165<E39
:8=9:59?3896=6G<5C23<E328>349=<39=886749=<=48? <=56=2<E8<<E3FC8?=<@YC7G>36<245>L=63865AA83?9C88?B?
3@382<E33X;39<YC7G32HIE39389365<3D59<E@7=::39B FC8?=<@248?3N;932C>8L?@93J34<=6G4562C>397=2B
36432=6<E3D8@2=6DE=4E236259@<32<=6G867FC8?=<@ ?=M32VD=<E7=8G652<=4=6:59>8<=568L5C<73:34<2P8
YC7G=6G893;39:59>37HK5>35:<E3238935C<?=637=6 M=675:73249=;<=A3868?@2=2?55M=6G56?@8<<E363G8B
I8L?3QH[H
IE345>>57=<@G987=6G867<E3=62;34<=56<987=B <C=8A<38 2;34<25:;957C4<2Hk6>8=62<938>236259@3A8?B
<=56E8A323A393?=>=<8<=562=6<E34C9936<3985:E=GE?@ 3A8?C8P<
=5 6 E373249=;<=A3:C64<=56867<E34562C>39
;95432237:5572867>89M3<23G>36<8<=56HIE393893 <32<2D=<E7=::C3?97L
<=56D5 34?389?@23;898<3=6<D57=2<=64<
:3D39867:3D39\2<867897;957C4<2]93?8<=A3<5<E3 =565:82=6G?333X6;<9 3
32;56736<2HlE3<E39<E35;=6B
9
<4 863::34<=A3?@93;93236<456B
D=73A89=8<=56=6J8A592P6C<9=36<?3A3?2N3HGHP?5D 2C>395;=6=56=2E=GE?@F C32<=568L?38<<E=2<=>3=6
:8<VP456A36=3643;93;898<=562P8675<E394E5=432<E8< E=2<59@H
mnopqrst b56<982<5:236259@3A8?C8<=56<32<2D=<EFC8?=<@=62;34<=56
K36259@<32<=6G
K3;898<32E3756=4N?=M3u7=2?=M3V86773249=;<=A3=6:59>8<=56=6<523;898<3<32<2
v23293;93236<8<=A34562C>392:59822322>36<5:;957C4<8;;38?N?=M=6Gw7=2?=M=6GV
v232<98=637;863?=2<2<52;34=:@8<<9=LC<32PLC<65<?=M=6Gw7=2?=M=6G
g9=36<37<59323894E2C;;59<
W?3X=L?3:5963DP36G=633937P867=665A8<=A3;957C4<2
x>;E82=_322<8<=2<=48?=6:393643:59734=2=56>8M=6GP2C=<8L?33X;39=>36<8?732=G62P86728>;?32=_32
`C8?=<@=62;34<=56
v237:59;822u:8=?56?=63734=2=562=6>86C:84<C9=6G
i95A=732FC8?=<@245938677=8G652<=4=6:59>8<=564564396=6Gyz{z|}~=6563<32<
v232236259@3X;39<=235:E=GE?@<98=637=67=A=7C8?2
h8@C2356?@56359A39@:3D<98=6373X;39<2
i957C4<M65D?37G3P;5<36<=8?;95L?3>2P86748C2328932<932237
I987=<=568?248?32893>C?<=B7=>362=568?867;559?@2C=<37<52<8<=2<=48?868?@232
e34=2=56B>8M=6GL82=2>8@L3FC8?=<8<=A3
g9=36<37<52<86789745>>57=<=32
01 0 234567894
63
9(') , #!$%%.(,,!)$* 0+!.&)':.(4
)!(*($'('';3+)(5)$'( ,$''!(1&':.)!(+4
!"#$%&$'()*+!($' ,''*'-. '.((#$ $%"(;3$.*$&!(0)&*!%(6$.
/('-0&(.%($.'$1%(-.')-(2&,!+!.&)' '(%(6,: $,,)'(# !-.()''(0;7< !!)'$++%()$4
+!'$'(3/%%4.5.!+'$63'$'('()$% *'-.''-!"'1)=-)'((2#&(,'
#%#)!!)'*$')-(0 ,
*'-.3$.0&(.%(,!('!+!'$'( ,!&%'7 0 .)**&()$'(1'/- ''3$.'-(!2&(!
!"+)($%('$.
8-!+!(*$!"6(. ,!"'',)& '-
>?@ABCDÿ
F?G?H@IBJ \QN
KLMNOPQRÿTUUQVWXYZ[ZW\ÿ]OQNWZLM^
ML
K_bR
LQLNQ̀QRÿ̀QRMLUPa
QcU_LZUQ
dX M e ZM f
dXWQgÿTUUQVWXYZ[ZW\
ML
hQMNLR\ÿTMX[\WZUX[ÿ]OQNWZLM^ fLÿWLÿbXMQ[ÿ
hQWOV
\QN
\QN
RQkLVQM hZPV[QÿhXPQcgZ̀Q̀RQMWÿ]OQNWZLM^
gZNUONNZLM
Lÿ̀LYpQUWZjQN ML K_LLNQÿ̀RLPa
ÿÿÿijQRX[[ÿgZ̀Q̀RQMUQÿWQNWNÿ
ÿÿÿMkX[WQRMXWZjQÿ̀LRUQgÿU_LZUQ
ÿÿÿÿÿÿdXWQgÿgZ̀Q̀RQMUQÿ̀RLPÿULMWRL[
ML fLÿWLÿbXMQ[ÿhQWOV
lXWORQÿLÿ̀mZ̀Q̀RQMUQÿ]OQNWZLM^
\QN
K_LLNQÿ̀RLPa
ÿÿÿgQNURZVWZjQÿXMX[\NZNÿ
WQU_MZnOQNÿLRÿPLgZoUXWZLMNÿ
fLÿWLÿbXMQ[ÿ
hQWOV
qr s!"#$%&$'(.+$!'*'*$"('!$)'/('- )*+$"t*$!6'(0,,!'7u /#!3'-"*$"$% ('!4
*$" '-!.+$!'*'($, . !)&*!+!.&)')*4 $)'/('-2&$%('")'!%3*$!6'(0!$!)-3+$)6$0(0$.
+$"78-(!+!(*$!"('!$)'(((&++!',+!.&)'!$!)- .(00!&+3$.#%0$%!#() #!(&&)-$)%$(*
$..#%+*'3*&)-$*$!6'(0!$!)-&++!''- &1'$'($'($.$.#!'((0)-$%%07
0121314516 78
9
9999 99 L9&R
@

&KL*22M79
99

9
9 %99
%
# 9
999
9E@

9&**&*5*Q
99
9 L9&)6 &B'9&/
D
'
&%0118
9

999 99! 9 B999

999

"
9
O 9
9
9
9
9

9
!
9
999
99
9! 9 LA
99
&)
B@
&*8&Q025QG+
9999
9
99
C 6 &-#&@
- &S/*288%9

9

9
#99
9"
$ I9$J' BK T
I
UTAK #
&

9999
9#
9
99 L9&@&
V&@I&R
9&/&@
9S
&/&
$ %&L
H 9&-011MW
99
99
%9
99 9

# 9

9

9

9994

99

9 9
9 L
& @
I9 #&MM&*F1*5*F*1
/&%'
9&CX
79
&@*2MG-9 9
9
&9
&'
&99%9
499@
)%9
9&08&+1+5+12
9
99

99 L
K#99&B
/S&@C*28FT
9D99
$
9$ 999
# 79&T09K11 G#A %
&L99H&7
99
9

9

!
&'9
99
99
99&'
&
&
9&9 S#99)3 '9

9O9

K999
9&*+&0Q5G1

()*+,%9

7499 &T
%$9&)/*220C9

9 9'
9
# 999
%9

999

9#@

9999

9
99$9 7 &@)*2F% 9&Q&0FG50QQ
Q-9499E

9

9
$#9 9
)D99&Q&*5+1

99

999

79 9
&/
$9&H-*2207
9
99
9'99 9-9 @
9
99
9
E '
9

99999 79&C&-Y)9% 9
9&FQ&F2M5M11&M0*
&
L&%0112)9
9O
! 99
9 99
99
999

9799

...&9
9
9
9
99
B
K9&/A9
O*1*11QU*01Q811221FQ*
%&/
&TKD&B9"&D
K
&

9
99
&
9'9 D 011 + - 999'

999

99
(-
/
99
9&0112,
99 @
A
9

)

3
999 9 99
&4 9
!
9&
4 9&
B9"&D&D9
'9&
#
9 /
&0&Q*58G
&H999&B&/)9&H@H

(%
-" 5-96(7+&*8,,
)9$9&@*222-9
9#
&

9
9999

99

9
9
99#9@

A
9

)
#
9/'9
&G&MQ5QQ
9:;:<:=>:? B9"&D
/)9&H@H*22M-99
9
9

99
#99
9
99
9

99

9

O99$@
%9

&@A&/9&B@
C
&DC*22+-9
E 9
9 % 9&**&*1G5*08
9&#99
#9

99
B'9&/011QA
9D99$O9

9 9999&00(*,&0F5G8 @
B%9
9&21&+20F5+2GQ
&H*2*8-9C
H9
-9/9
C
9
&/*22*B99 9
! 9
9
#&I9$J' #"
A
OH-$9
LKR9
(9,&
9
9&/&K
#
&D/
D99&CL*2MF %9
%9
9-9

)/9
K
9%9
C
)9&I9$ B9''9&I9$J'&G*Q5GG8
J' C
9'
&Z&I9
&7
H9
9#9&%011QH$
%-/C*2F80118%
99
#

9
# 99

%-/A
9

&697
'9
&K 99
)3
K999
9&*8&*GG5*G8
%-/*282%9
9
A
99#
#9
S &/7&@*22*7 #

9
9!

799C*8
%9
C
/9 9
9
9A
OH-$9
LKR9
(9,&

K %-/&K9 %9
%9
9-9

)/9
L&CN
79&C011+D99$O 99 B9''9&I9$J'&+*G5+GM
99

9P

9
S#9&&H&C/
%&DS011G/9

99)%9
9
-9
A
9

&*1&
99
9 )3

GF25GQ* K999
9&*+&G5*F
01 0 234567894
63
!"#$%& NM@"@;:*!"#"!#!$%
$!$ $$$"#%$'"$"% %#!$5!$E8%?9!!.
"($"'"!&!##%"#$$ ;+I-;.
!")$*%!+,,-+,. *' >@?AJK;.IK $#!4
&/01& 23&$$4 %($'$" ##$ %&A?
$$ "%4% '5 $6!!!#$!7 9!! $!$.++-;;
8%99$,2:;-22, *A9K9P,$!'%
<&=8'9> ?"$ %&$'$!&78%
3@##!$ % $$ ! $ *!$@A$'I-.;
(A!&#$ !$B$ $ $$'5 *&?>9$0LJ;+@&$%
%%!$ "!$')$*%! %%!G!"'I,-,+
5;;;.CD%E;, *$$H<J<K.9#!T!
&FGH;I.9!$%$$# & (%U4 $$%V45
J$K $3&&!$;2;-;I 9'T!!%6(!"8%9!!:;
8KL*&?>G"$KK;++ 9I;-9I.
?(#&$% !%&' M%#%4 9"#">9#A*H9>$@@;;C;
!>!"+;-+ G"%%!$ %%&$ $' #%4
<&8;2N ($&!! !"$$ !5G"&#$!%$$$)$
O59F($O! $%H'' *%!,;-
3&&$$F4;29F($%@$ 98K9$H&E$8;2;W &
*!$9G@*"#"##;;2-; %($!"E$!$8
<&JPG0A?$93?? %?9!!.;.-,
O23&($%#!A'%O$!$( 9'$@3@4FFB#!$$ 6!
%%!$%&$!! $%%&$##$$ B#!$$'##$$+:.-:.+
+;.2-.,I 9AFX;2Q3&&$R9($O!4
K G;,G"!$$'% $% ''3&&$$F4;29
!$$)$*%!I+;-., F($%@$*!$9G@*"#"
K G3 @>;,G"!$'& ##+2-.+
($G!"'I:Q.R;,-;I. 9$98KI9F($*!$!
K G<H*;2!&!(4 G"F$!&!9?'
$##!$( ($8%9 9$98N(9J9'$
9$,+-;. >3;:I9($ E$$$(!#$(
KK!A9?.G$MA$ G!"'2Q;RI.2,,I
*!"'!9!!;:;+I-;+2 9G0Q9JR;.,Q$R'!!,+H3FG"$
@!>>J"F=;*%!! %J9H%T$"$NB%W($
&#%$"$!$%$ $$%!&4 90!0A9KH89AFX;:+
$*!#$*!"#"!+;;:-;2; 3&$B$#T$!"E8%9!!
@'@3(L3HG.9 I;+,-;+:
F($G!"E$"9!$3>3H! G@$F9% !!
>$ $"!"'' !$5##$$ !"'
@&K;,3$!($% $!"4 G 9!!G!"'+I-.
E)$*%!I,,-I J!A38K 8;2IO(%%!
@ KS$@HFB#!$$ # #!#$*%&($;-I
&!'$&!&)$*%! P&@>3"& KH&$"HA
5;;;.CD%E+ 2G"%B#!$$ "!(4
@A$0>H!A8>!!L $5G"!%& &A!4!&)$
.93&>!"*!$ *%!;+.+-+:,
?'?(#&$J4H!A=PA PA><;+)$ &$'$(4
@A$0>;2,*!$G$'9F($ &$)$*%!.;,:-;I;
%=$$J$#$3G X?9$!A"?G3FI
N(G;2.G">3AG">9$>& ?!T&"!B#!$$ #!#!#$!4
=PA $$$ &$G"&!8%*!"'
NM@"@;229%%!#%!$$'5 ;;:,.,-,2:
G"% '$!$&3"#$2O5>
@A$0QR##9%3>3
H!>$K##;I+-;:+
01234567
819
2 28
91 2 24
5
69
4
!"#$%&&%'%( )"!*+&,-./0"$%&"1-./%&#...+ %.,&2 !%.!%

"./ %.,&2%'"1-"$,.0%$,/,1,+%34"!"./$,&!"1#2!,#2!"10%$,/
"&%&%'%(%/"(%11"$%"."$,025#2,1,+25"./6-.!$,.,6$%!%0!"1%.%3
%!"#$%&!,.!1-/% ($"/!- ,.,60-1$70,/"1%.,&2.$%&"!$,.3
89:;:<=>?@?>@:AB<?><>CD>E=C<>FG9<@99DB9>?BDBH;=BIB?BDJJK?;<>AD;B=IL::>L:M?BB:>
NA?>LKB9:?;MD>=?O=:>=:P
Q,0" R,))%5S%'"$".TUVWUX
0454
o3W3o q2,1,+!"1s%!".0
,6r%0%$% 333333333333 Y
YZ[ \]^_à^]33333333333333333 [b o3W3t r%0!"1R%"$ 33333333333 v
o3W3w $%&&&$"$'%u%."$,. "./r%0!"1
YZY cdadef]^ghfa]i]_jgk^lkgad
mfk^_3333333333333333333 r,,1.+ 333333333333333
o3o3U p"&12q2!,#2! 3333333333
Yn o3W3W $&.+%.!233333333333333 y
Yn o3W3V s%$"11!"$%3333333333333 }
o3o3o %r1" !"1q2!,#2!"1
s%$,/ 333333333333333 YZ} m`da~^_def]^g\]fa^] 333333
o3o3t u!"1.+"./s"+.$-/%p$0"$,. 333
Y[ o3V3U "$%".//,&.$%&"!$,. 3333333
o3o3w r&$x-% ,6u$%'%. 3333333333
Yv o3V3o &&$"$,."./|1"',& 3333333333 b
Yy o3V3t r,1,&|1"',&.$%&"!$,. 3333333 b
o3o3W p0#&!"1z%&- %,&27{&'%.
|-.!$,. 333333333333333 YZ c^]d̀a^]33333333333333333 yn
o3o3V q"&"11%1,6q2!,#2! "./u%.,&2
Yy fff]f 33333333333333333333 yn
p'"1-"$,. 33333333333333 Y}
YZv ]^~g]_jkga^dîg]_`]a^]
^hf 3333333333333333333 Y
7464
o3t3U ."$,02"./q2,1,+233333333 Y
o3t3o "$%q%&!%#$,.-"1$% 3333333 vn .,&/%&$,/%+.%66%!$'% %.,&2$%$"./#&,'/%
o3t3t "$%q%&!%#$,./"#$"$,."./s$-&% .+$6-1.$%&#&%$"$,.,6$%&%-1$5" %.,&2#&,7
.$%&"!$,. 33333333333333 vn 6% ,."10-$-./%&$"./$%6-.!$,."1#&,#%&$% ,6
o3t3w ./'/-"1{66%&%.!% "./"$% $% %.,&2 2$%0 $"$"&%&%#,.)1%6,&$%/"$"3
%.%$! 333333333333333 vv
YZ ]^~g]_jkga^dîg]_`]a^]
ê~fdd3333333333333333333 v
42"6-.!$,."1#&,#%&$25(%0%"."#%.,0%.,.1*%
o3w3U ."$,02"./r%11-1"&|-.!$,.33333 v 0$-&%.$%&"!$,. -!" 0"*.+,& -##&% ,.3
o3w3o %$&,.""1u0%11 33333333333 v} .,$%&%"0#1% %.,&2"/"#$"$,.5"!,00,.12
o3w3t 16"!$,&2u%.$'$2"./u#%!! ,)%&'%//%!&%"%.&%#,.'%.% $,!,./$,. ,6
.,0" 333333333333333 v 0,&%,&1% !,.$".$ $0-1"$,.3."//$,.5$
o3w3w /,&-"1$%q&"!$!"1u2$%0 3333 v -%6-1$,-./%&$"./$%"."$,02"./#2,1,+2,6
o3w3W |-.!$,."1q&,#%&$%/"#$"$,.5s$-&% $% %.% .',1'%/" (%11" $%&6-.!$,."110$"7
u-##&% ,.5"./%1%"%33333333 vb $,.3+,,/%"0#1%,6"6-.!$,."110$"$,.$%
YZy ckf~fkfa 3333333333333333 [
o3W3U -"1$% ,6r%0%$%$!
p#%&%.!% 33333333333333 [
$&%,1/,&0.0"1"0,-.$,6" $0-1- .%%/%/6,&
#%&!%#$,.3.,(.+"),-$$%"."$,02,6$% %.%
R33S"(1% 5R3R%20"..5:>=?;KDJNDB<?>?O??F5|,,/u!%.!%%$u%&%5
{U3U 7U7wwU7Vw 7W¡o5¢u#&.+%&u!%.!%£4-.% s%/"5SSroU
01 0 2345678796
8 254
378796
87 6757574
67
!"!#$ !&)!!+!$%$$*!"*
$ "$ ! !$# $ )!#%!%1*.>$4+<??<0'
%&" !'($)) !!)#$*+)! @*!"$*!)$$!
,#+ !!!$)-*!+" !#%$!$1$&+$))#!*!$
!!#! !)!# )! .!/ !&)$!'5$))% !
$0+1!#+"!!#! *$+$&A$$!+&$!!*$.'*'+*$*0+
$! !)!# #!!)!#%$* !1$!#)!'5$! */
!+11$2!""&)!# $!)#$+)!,#+,$/
!),$)-*' $!!)!#!,$!$ -
3!%$ !&!)$! $$!!)$%!!4'5 !$$ !!* !!
!1%4*!4!"*$1!1 $&*$*!$1!#$1#
!&$*#!###!/ 'B"$$#$!$$!!)$$1$$))/
$!$! *$)!! ! "
"! !&'5$$ !) $!)&/ !!+%$$$1$&+#%$+
!&$+ 6$-$!## !) !!1$$!!#$'5!&$$!
!&,$'7&!&$$1&!$/ !%#$)!,$$!#"$))/
$$ )!# %$)!&$$ !&"!'5$))!$%!
,$#&!!*&'7,$%"&/ $1$&!)!#)'5&!$#$
!&$!&1$!'8!,#+ *!"$$%C$%D$!
$))$*$!#$!$1$!!) $)!# "&!)!$*'9$1$$))
#!)!#$*$)) ! $#!&1$!$#&"&'
#!!)! $#$'9$/
$1&$"$ $+""!41&
!$ $$*$/$#$! EFEGHIJJKLIHMNOJPQRSNJTKOU
'5#$&#&)!$/ IOVWJRLXPYXRJKLIHZNTXPVJ
*!&!$#$4"!!!*
%!*$$1$$&' [\[\]^_àbcdbefghfbdied
5&!!!$#$$%)!#
$!$*!$4 !/6$&*$*#/
!'5$&!&$%$)! 5!%!),$#&!!*&$
$%$$$1&$' !)&!&$+&!)$!$%"
5$$$*!!1$!# &$$#$!&,$'5-
%4*!$!&#!!)&!&$' &!&$ !$ " $&
3!!%:$1$!*$1!11$"!)/ j#&$!!*$+k'l'B%'m$$*!/
)$!!) #$!) + $!%1$!%&m!$!+B%!
#+#$';#$) #! &$$# !%
!#$&$$! # !% $ !%:$1%&$)) "!/
& $$++ $!'5<n'o<o!!% !
$*&+#$!!$*'5 +% "$* $)-&+ #!
!*!#$""!!&21! %$!)=?pq!!<n'o<o#
$$)!$$*)!!+!*"$ .m!$*+<?n=0'5$!)!#$!!)B%r
$+)!+#!$!/,$ "+*&"$!"&
!))!!,#!)'5,$
"$$;<<!!1/ stutvw .='<0
$!$;<='5$!&$! *
!))!!$!+!*#&! "st$$$ &$$#
%$1""!#$)!!'5 "6$ !%: $$#$%&$)) )!#
!1$!#!$&!!#&$/ !#$*1+t'5)$!+stAt+$!#/
)!!,$!+$ !)$ $# CB%)$!D$$,!)
010 23455674389
5 956
4
5 7 5674395 0
! "##$%&'( ($)$(#+$),!$+"),
$(())*"%$ +,+$")*("" !&'!",(!($+(""0
-.+$($(*)$(%#$$#((#,$ !($2+"#("&'!",
!&/","!((%"(,,#0 #$(!+(#+"2+""!("(,,#
"1+0(#$20"(,,#31&&"&4*$+2#$! "$"!",$"1+!$2(
#5($",$)#%(#$$#& .+(*$#&
'!"#"(-"2 6&'&7#($ x!",(!()(#$(!++(
25#$"89:;:<=:>:?@ABCDEFDBAGH389:;:<=AEI #$%"2 +##(*"(#)+($#$%
@ABCDEFDBAGCA4(JKLM&7#$$)()$ ()("&7S$!)O(!+0
$$#(($""*)"$((N$ +((#$(%(((O)(#$%
(%(O($+,+O$"(P$($ ,! Y$([Yx"#!(%&[y
!("Q2" "(#! &7#R2$")()(0 (!++("#$(%(((O$!)
#$($*2$%*5"O2+(( )(#$%2$#5Y$(&[z*
25 !"$2#!$#$(# !$ ($O$*$%)(,#$%#$2$5
S,)#% $2$&7#$$" $)R$2+"37(%&_&J4&'(
$*$+$2((%&T$(U"$1&&"&!(%2 (!((!+!(( .+(","#(,
+"$$+(,!$+!$"$2 $""(%+) (!++&/"*$($((!",+
1&&"&#+"#+#$)#)(#$$(0 $$#"(%($",#)$(#()$
()2)(#$(!++(( $" #$$%$!)$!%$$(*Y2$5[
((&'($(()$))S(!$"$%,+#0 $!)$$#)&N$###$(!+2"(0
(O(#(%$,JVW"W())($ #(!($",!$2$#5%+"*+#$)$($
$+$%,W&X$%$(!(#$(()$))$" (#$,$"&7#"0#(#!"
+,+$$%$)#)(#$Y$Z[ ,"!((%"$"(#+"("$((
P$)L&
\]H%^ 3_&_4 x!",#$(!+(O(!+0
+($$#!)$"$%$($#$,#
\($((( $"^(#$%$( *3$$"$"4O++$!("")($(
)(#$(!++((&'($(()5$ ,)(#$(( *&'+21#R12(
7#R$$)*$+,++,+!2, )"$#(! $Y%$$[Y
$ à $O+(($.+("2 $#+(#$
$#+))$"(($)$3
X#(_&_&b4&

cdcdcefghijkklmjinkomfpqfoklmji
{|}{~}|}
rgsfptk

7#R"+(%#(2+($$!2$"
)+2("$(, !"$"
*$(!)$)$(%#$$#((#,
!#+"2!$+"&'(!)$!0
"!",(!(O!",#$
(!+(3#$"!",(%$"%#$(
"$4O$"!",$"1+!$*$% S$!),!
3u(%OJvw_4&'!"$(+""$ A9FEG<=A+"2$*$%" ",(!(&'CG?C9:>?::?
(!$#(+$($"*$($( Z$#"(%(&Z"#"(2% $()R"&
!",!)$,2S,$))(" $#"(%($#!!+(" &x$$"!O
)*,$(%+O
*$+$(&N$#,!"$$#($" $"$)$(#$0*,)($(&
00 0 12345676859
7
14392676859
7 6
5646463 956

!ÿ#$%&ÿ'()(ÿ
*+,%-ÿ%./ÿ%.ÿ+%/(/(

1;<

+%/(/(

0123405ÿ2462567387946ÿ#:&
=>?@A@A BCDEFGHIJKLMFNOPFKMHPQJLMRJQNLHI KGJSJKGHQ bOQYJQDcJJKJLKGUPKGJDKUPQULQULJCWHKKJQUYUMPDKFHPFJPKLUd
HNTHPDKUPKVKMIOWMXULJCJUKJQDJLMJDHNCUMLJQFHICULMDHPD KMHP_eGJQMNNJLJPFJKGLJDGHWQMDQJKJLIMPJQfEKGJFHPFJPKLUKMHP
UYUMPDKUFHPDKUPKZDKUPQULQ[DKMIOWODXMPKGMDFUDJ\]^DOFLHDJ_ QMNNJLJPFJfJKcJJPKGJDKUPQULQUPQKGJMPKJLCHWUKJQgh^ ZHL
`LJaOJPFEHNbOQYIJPKDMPcGMFGKGJFHICULMDHPDKMIOWODMD ih^[CHMPK_jklmjCCJLQMNNJLJPFJWMIJPZKGLJDGHWQ[_
KGUPnKGJDKUPQULQ_SUPELJCWMFUKMHPDHNJUFGMPKJPd HNQMNNJLJPFJKGLJDGHWQ_BIHQJLPUCCWMFUKMHPMDMP
DMKEWJRJWULJCLJDJPKJQ_eGJCJLFJPKUYJHNKMIJDKGJ IJUDOLMPYDJPDHLEKLUQJHNNLJWUKMHPDGMCD_wPKGMDKECJ
LJDCHPDJMDoYLJUKJLKGUPnFUPfJCWHKKJQUDMP`MY_i_i_ HNJtCJLMIJPKKGJQOLUKMHPHNURJLEfLMJNKHPJFHOWQ
eGMDVdDGUCJQFOLRJMDFUWWJQUCDEFGHIJKLMFNOPFKMHP fJfUWUPFJQUYUMPDKURULEMPYDHOPQCLJDDOLJWJRJWKH
ZpHLMPYX\qri[_eGJQMNNJLJPFJKGLJDGHWQcUDKUsJPUD EMJWQUFHPDKUPKCJLFJCKMHPHNWHOQPJDD_VMIMWULWEXKGJ
KGJQMNNJLJPFJfJKcJJPKGJh]UPQgh^CHMPKDMPKJLCHd QOLUKMHPHNUxUDGHNWMYGKFHOWQfJKLUQJQHNNUYUMPDK
WUKJQHPKGJNOPFKMHP_eGJIJKGHQHNFHPDKUPKDKMIOWM MKDCGHKHIJKLMFMPKJPDMKEKHFLJUKJUFHPDKUPKCJLFJMRJQ
fJULDUDKLHPYLJDJIfWUPFJKHFOLLJPKKJFGPMaOJDHN fLMYGKPJDD_`HLRJLEfLMJNKHPJDHLfLMJNxUDGJDXKGJLJ
CUMLJQFHICULMDHPXcMKGKcHJtFJCKMHPD_uPJCHMPK MDDOIIUKMHPHNKGJMPKJPDMKEHRJLKMIJMPKGJPJLRHOD
HNQMNNJLJPFJMDKGUKKGJIJKGHQcUDYJULJQKHcULQ DEDKJIXDHKGUKMPFLJUDMPYQOLUKMHPFUPfJfUWUPFJQ
MPKJLRUWJDKMIUKMHPXLUKGJLKGUPKJDKMPYNHLDKUKMDKMFUWWE UYUMPDKQJFLJUDMPYCGEDMFUWMPKJPDMKEKHFLJUKJUFHPd
DMYPMvFUPKQMNNJLJPFJD_eGUKMDXKGJKJFGPMaOJJDKMIUKJQ DKUPKCJLFJCKMHP_eGJDJIJKGHQDGURJCLHRJPODJNOW
CHMPKDHPKGJCDEFGHIJKLMFNOPFKMHPZihXh]XUPQgh^[ MPOPQJLDKUPQMPYKGJCGEDMHWHYMFUWLJDCHPDJHNQMNNJLd
UPQLJDJULFGJLDcJLJPHKFHPFJLPJQcMKGDKUKMDKMFUW JPKDJPDJDKHKGJKJICHLUWCLHCJLKMJDHNDKMIOWMXNHL
DMYPMvFUPFJHNQMNNJLJPFJKJDKD_BWDHXULUPYJHNFHId JtUICWJXGHcKGJUOQMKHLEUPQRMDOUWDEDKJIDMPKJYLUKJ
CULMDHPDKMIOWMcJLJKJDKJQUYUMPDKKGJDKUPQULQUPQPHK JPJLYEHRJLKMIJ_
bODKUDMPYWJCUMLJQFHICULMDHPHNCLHQOFKD_ BQbODKIJPKIJKGHQDGURJPHKCLHRJPDHODJNOWNHL
eGJKGMLQIUbHLIJKGHQMPFWUDDMFUWCDEFGHCGEDMFD UDDJDDMPYDJPDHLEJaOMRUWJPFJMPUCCWMJQNHHQKJDKMPYX
cUDKGJIJKGHQHNUQbODKIJPKHLURJLUYJJLLHL_eGJ UWKGHOYGUQbODKIJPKMDHPJcUEHNKLEMPYKHHCKMIMyJ
DOfbJFKcUDYMRJPFHPKLHWHRJLURULMUfWJDKMIOWODWMsJ UPMPYLJQMJPKWJRJWZzJLPUPQJyUPQlUcWJDDX\qqq{
UWMYGKHLUKHPJUPQUDsJQKHIUKFGUDKUPQULQMP SUKKJDUPQlUcWJDDX\q|h[_}UPYfHLPUPQFHdcHLsJLD
fLMYGKPJDDHLWHOQPJDD_eGJIJKGHQFHOWQfJODJQKH JICWHEJQUPUQbODKIJPKIJKGHQKHDKOQEMPQMRMQOUW
QJKJLIMPJQMNNJLJPFJKGLJDGHWQDfUDJQHPKGJRULMUfMWd CLJNJLJPFJDZ}UPYfHLPX\q||{}UPYfHLPUPQpLUQQHFsX
MKEHNKGJDOfbJFKHRJLIUPEUKKJICKDUKIUKFGMPYXNHL \q|q[_BQQMPYxURHLDHLMPYLJQMJPKDoKHKUDKJnUKKGJ
JtUICWJXODMPYKGJDKUPQULQQJRMUKMHPUDUIJUDOLJ fJPFGKHCMDUFHIIHPcUEHNMPMKMUWWENHLIOWUKMPY
010 23455674389
5 956
4
5 7 5674395 0
!"#$"!#%&&'&"()$&#" *)%")"#+"#)$')))#"!1)")$)6
*"'+))",!&"))'%&"*) )')S2UVW#+SXYXWT0*))X!*)'))
&)"+)-)!*)""'"$&#")*" "'XW!*))$))')'))-"##%"1"#)
&)&)!.)$#/0)1)-*) $"2#)*)*#T/0)1)-!)"'!)0))
&)*""#!)!'*)%**%!#"2"% 2)1)2)0))')!2)#"'#')!'
!"''0!)#%)*'!3)$*))')$")"' ')"-5)1)'$'0!**)!)&"+'!)
&)##0*)))#"2"))3!&)'!')))1!) )!&"!'))*"#')0""0)$'6
"4"2#)!&#'#5&)*$)3!1"6 !'$!&#!')'!%-0!*"'),')'$"2
#)'%"4&)'")&)0*"")0!*$)6 WZ5"#!'+$*))'%'!'""#+"1)
!'+ 0)$'!'-)"*0!*!*"")!!),')'
S5)1)'-[\]^-[\ZVTN**)$##0!'+)#"!'*!
*)#_
78789:;<=>?@<?AB<@?>CDAEFGC>H<C>I?
Ùabc#+Ùc#+bd#+a SVeT
J1)%)$#)*'!3)$)'%&)")&)'*"
2))'*)!)"#!"!'$"!'+"#)&)") 0*))c0"*)*"")!!),')'"'a0"
*)!')'!%$)'"!'/!!"##%('0'"*) "!'"#!%'"'))&!')2%*)'!$
K&)*$!'+#)!&#!-L*)))!*!+*#% &)")&)''*)0-*)$'!'$&)"
)$M!)'!')')!&#))'"!'%!)# "!+*#!')!'"#+Q#+#0!**)),')')3"#
')""!'N*!!!''""))#!() *)#)$*)#!')"$'!'N*!0"!''"
*)&)*$'"'!&#!-0*))*)))'"!' *)R)*')!"'#+$'!'0*!*0"""!+*
$&"'%"!!'))"%+!1)"$)3)'%' #!')!'")&!#+#S)')1)#+*%!"#
$*)'&2)$!&))"*#)1)#!4+)'+) !')'!%T
*"'""'"O"!'+"#)*"1)&"'%)P')$ P')$*)&)!&"'*"")!!$"
*)&&&'!)!$%"%**%!"#$'6 0)$'!'!*"!"'"&&"))#"!'6
!'-"3"'!"!1))#"!'*!2)0))'*)))!1) *!*"")),"'!'+!!1)#%")#)")0*!#)
!')'!%$")'"!'"'*)*%!"#!')'!%$ *)#+$'!')'N*)0)$'!'0!*"'
*)!&#N*!!"'*)0"%$)!2!'+" ),')'#)*"'')M"#"0$!&!'!*!'+)'-
)Q)')1)!'*)0-"!'+*) !)-#"+)"'#"+)*%!"#!')")"))3!)
!'Q$'!'$")'%%)& 1)! &"!'"!'"'"'!'"#!')")!'*))'"6
%'"&!"'+) !'#)1)#P*)'!'"*")'))#)!
N*))*'!3)$&"+'!))!&"!'+)0$ *("'&)")0))$'*"1)"0)
)"#!)))!'0*!*24)0#2)"() $'!'),')'+)")*"'')Sf)!)#&"'-[\^[g
$"!'")"'"4"2#)!&#R),"&#)-"26 f(0!h-[\^[g5)1)'-[\]^TJ&"!'$
4)0#2)"()"4"#!+*')'!#! 0)$'!'0!*!$$))'),')'!*0'!'
))&)*"#$"2!+*""&"!'!&#N*) R!+Ve
)*'!3)0"&!M)*"*)),)!&)')'6 f"'%)'%%)&*0"'),')'#)*"
##)*)!&#"'*)24))')!'+ ')N*!*0"&)!1))')+%)#"!'*!*"
S')!)T'&2)!'!")*)!' &"%*"1)""!1)1"#)$"'+"'!&)'!'+
"!$*)))!1)!')'!!)N*!$*))!&6 "0!)"'+)$)')+%!'*))'1!'&)'N*)"'+)
#0"0!)"2!+*"*)"'"-!0#2) $&*)#)'')"'#)")*)$"!')
"!+')"'&2)0!)"#"+)"*)"!'+$*) "!2#)')!1)[WWiN*!)))'1)[W#+
"'""'!$')6*!"2!+*-"'&2)')6*! '!$')')+%-""!$[W2!##!'')N*)
"#"+)J'!&"'2)1"!'!'555)1)'. %'"&!"'+)$*)1!"#)')$*))%)
#"2"%"/"1"0"*"*)#')$' !$$))'#)1)#$#!+*)')+%!)3"##%2"N*
0"'),"#%!'"#*))!2)#"#)$ ),')'#)*"'')*"1))#+!"#!+'!M"')
R)*').#+)#"!'*!0")-")#') $)'%%)&*"")')"2""'+)$
*#+0!'"#!')"$"*!'0!*)!2)#-!') *%!"#)')+%#)1)#
01 0 2345678796
8 254
378796
87 6757574
67
.//&" ÿ
5"+&+ÿ,-+2 0" ,
ÿ7ÿ4 ÿ1ÿ4
*(+,--+ÿ !"#ÿ
ÿ1ÿ2ÿ$ÿÿ
6&,"
ÿ3ÿ4
ÿ !"#ÿ$ÿ%&'&&(ÿ )
5"+&+ÿ,-+2 ./
0"/&," ÿ
0-ÿ*(+,--+ÿ !"#ÿ
-ÿÿ1ÿÿ-ÿ$ÿ8ÿ-ÿ2
ÿ7ÿ4 6&,"
ÿ1ÿ4
ÿ3ÿ4
0-ÿÿ !"#ÿ$ÿ
9:;<=<> ?@ABCDEFGHI@FBJK@FBFHLMBLLHNOFPBQEOMH@P@CRCBOHBCHNOF@FBRBFBCOHBSIDDBCBFHGECTBLUVFOM@RWM@RKM@HHNBBJK@FBFH
XBG@YBLHNBLM@KB@DOZ[\]^_`[a^bcU
dORFIHESBBLHIYOHI@FOLOHBLHYBHN@SOFSHNB qNBFLBHHIFRHNBLBFLOHI@FLBQEOMPHNBD@MM@AIFRCBMOi
CBLEMHIFRD@CY @DHNBK@ABCDEFGHI@FD@CYBSOF HI@FLNIKLLN@EMSN@MSr
IFHBCM@GeIFROFSTOMISLfLHBY IFgHBTBFLhHNIFeIFRU M@ESFBLLsXCIRNHFBLLstM@Rubv stM@Rubw kwUxl
?@ABCDEFGHI@FBJK@FBFHLABCBKCBSIGHOXMBDC@YTOCi
I@ELBJKBCIYBFHLUj@CBJOYKMBPIFOGC@LLiY@SOMIHf OFS
YOHGNIFRBJKBCIYBFHPLBKOCOHBLGOMIFRDEFGHI@FLABCB
SBCITBSD@CHA@G@FHIFEOkBURUPXCIRNHFBLLOFSM@ESi bvM@RkuL@EFSlyk]G@FLHOFHlsbwM@RkuMIRNHl kwUzl
FBLLlUmFBG@FHIFEEYAOLHNBFLGOMBSOLODEFGHI@F@D yk]G@FLHOFHl
HNB@HNBCAIHN@EHELIFRFEYXBCLUj@CBJOYKMBPOLEXi
nBGHA@EMSXBH@MSH@OSnELHHNBXCIRNHFBLL@DOMIRNH OFS
L@IHYOHGNBSHNBM@ESFBLL@DOH@FBkoJBSXfHNB M@RkuL@EFSlsbw{bwM@RkuMIRNHlyk]G@FLHOFHl kwU|l
BJKBCIYBFHBClUpNBBJK@FBFHIFHNBYOHGNIFRBJKBCIi
YBFHG@EMSXBOGGECOHBMfKCBSIGHBSDC@YHNBCOHI@L@D L@HNOHKM@HHIFRODEFGHI@F@DM@RL@EFSIFHBFLIHfOL
HNBBJK@FBFHLIFHNBHA@LBKOCOHBLGOMIFRBJKBCIYBFHLU ODEFGHI@F@DM@RMIRNHIFHBFLIHfA@EMSRITBOLHCOIRNH
010 23455674389
5 956
4
5 7 5674395 0
!"#$%&$'()*!"# .#,",+#5,,#6##5
+,-,#./0+1'2324#5#6"5## 6#*#-,#,-,#5/0+#5
"5,7,#5"##.,#,-58,#( ]##,1'23S4#^,&,_8#-5+J
5()#"6,"#5,#*6,# 5
,-,Q,(X,8#- ,#**#",1*#J
9:9:;<=>?>@ABCDEF?BGBHC 7,-#58#7"5*#75###,+,-8#,1
*#-+,8#,#76*"J
,58,-/`##5[#'2Wa4(
I,,#,*,,7#**8J
5##"8.,# 58",6#*
#5,*,K*6 ,*+5#-( 9:9:bcde>=>fghiB=CACjkBD=lmn=>GBH
L6,#1###".M*+N,*#5
,#55*5#"8.,# - oAHf?>DHC
N,*(08 8""O,5#
#,#*5.-#,,## V76"*#5 ,6"*#,
#,# ,*P# -*-*#,# 8,-8 ,*#.,+#(),#,#"85
,##78""-".M*+N,J "8.,68#8#*#,##*"5.
*#5,""6"*.-*".M* R5###58#-6"*# #,#,-
# 5"8.,,8,,*#7, #,7 (L#,#+ -*-*#,#J
8""(I.+"-1566,*#7*!" #. ,5#.#5 -J
*"5 ,5"*566,,8#*76"*1 7*# ,*()#8J-,.*6"*
#",,7##J55*#J 5,+56,8 #68##*#58#J
7#QQ8#7"58##5#RN5J,#7 8#*#-!"+#6"*"55*,.
*#7#QQ*#7,-,#7 ,5"*5566, Q*6p-8O".,#,#()
-*-*#6"*/# ,6"*#5#7 [*#O[Q*!"##+*J
6"*1,*+-4( -6#p-8O".,#,#*##6"*6
LN55#,7"8"5. ,J ".,#**,#15*#*#1#5
68#7"58##5 ,6 8#N8"8 ,#/`7,'2S3q0,-,1'2234(
,*#7*!"Q8 *#7,-*# L,+,6!"## ,5.-
/L5,1'2ST4()8#7"58#*#8 V5,1# ,7 -716,5*,6
*#85#*!"#*##.6,",# *,*#,6#,+#5,*,
*#8#",81Q8#",86 -*# */V5,1'2a'4(),#7+.-
!"##",#**/718#1#1
*(4()## ,6,#.+68#",8# rs/r8#Nt4u/vwt4 /%(S4
,*!"#8,-,#Q,5,58",
8#",58 ,+#*#/U#,S4( ,r,1r8#N 8#N8#,1
I 6#,,8#5"*+*5( #5v**,##*,#6J
)- 5"#,*Q7,-6 8#N8#(Zp-8Q*1v#!"#- ,J
,# #585#7,#5, ,#5*#*#6p-8O
*1."6JM"6-7,*,*"#,,#7 ".,#*8 N(0*#++.576#
/V,.#"81'2W%4( 8*"# ,,*,1 ,#",,J
I ,.8 ##*#7,-*#7#+5## 7#,# #,#.#,#5
*X*,Y76"*(Z5,**# #p-8O".,#.57,#(0,#J
55#185 ,5"*5#*J #,,58,#87,#,*,
-8/[*V,51'2W\4(U#7,-*##,#5- *8*#/U",#(1'2WTq[*V,51
#55,#5"# 5,-7##." '2WS4(Z# 67**,#16"*
80+#,#575*,6,#J 6,8#0J#5*",+1##K# ,1
+*#7#58#7"58#/U#"1'23S4( # ,#5#,K#p,,7
]+#,7"8#-Q56*#*"5 #",#6,#7+/X7(%(T4(
01 0 2345678796
8 254
378796
87 6757574
67
!"#$%"&'"&ÿ )*ÿ+,"-#"*.ÿ/0&')&

7ÿ9ÿ73!:ÿ214ÿ%ÿ;<ÿ=ÿ1>
73!:ÿ9ÿ?@@
<ÿ9ÿ?@@
56""'&"$$ÿ7"$8)&$"

ABCDEFÿIF~Kÿÿ{ÿÿÿ}E~ICÿIJDF

ABCDEFÿHAIJKÿÿLIMNADÿFDOBHFIOEA

1)&"&'*!')&ÿ23##3)#!*4
PQRSTUVTRQWSXYÿ[\]S^WU]ÿ_QàWbTRV`cdUYWSeÿ

rmstgksmÿutmolmknÿgvÿwpxqwywz

{ÿ|ÿ}{~Eÿÿÿ}ÿÿ
ÿ|ÿÿ{~Eÿ|ÿ

ÿÿIÿÿCÿ|DÿÿÿIJÿFDBCADIOAJLBIÿFCEÿCÿD

fghÿjgklmknopnqgkÿ
¡¢£¡¢¢¤¢¥¢¦¥§ ¨ ¨¡¥ª£ ¤¯¨ ¡¢¤®°£
¡¢£¢¡¨ ¡¢© ¡ª©£¢££« ¢±²³´µ¡£¢¶¢¢¡¬¢¨ ¨¡£
¬£¢ ¤¡¢£® ¢ ¨¨¤£ ¢£ ¡ £ · ¤¯®
£¢¤¢¯¡¡£¶®£¡¢¯ ÁÂÁÂÃÄÅÆÅÇÇÈÇÉÊËÄÉÌÍÎÊÏÎÌÉÐÍÉ
¸¢¥¤¥¯¡¤£¥¡£ ÅÑÒÓÈÑÉÊÆÌÔÕÅÇÖÅ×ÐÊÑ
¡£¥¢£¶¸¨¡¢¡¢¥¤¡¥¹©¡£
¢¡®§º»»º¼®¨¤£ ¢£¤¥¯£¢¤¡½¡¢¢£¤¢ ¨¢ ¡¢ £¤£¢£¥£
¡¢¡¡ ¡¡¾¤£¡¡ ¢ Ø¡
¢¡¿¥¡£¥À¡¢¡¾¤£¤¨¢¡¢ ¢£ ¢£¡¢¡¡¡ ¡¢££¡¥
£¯£¥¡¢¢¡¾¤¡¢®°£¢¸¥ ¶

¯¡¤¡¢£®¡¨£
£¢¤¥£¶¢£¡¢¡
¢¢¤¢¡¢¤¡¢¡¢£¢®
012 345678954
9 77954
4674 75 6 0
!"#$%&'((&)""!*+'&,' QRSTUVWXYZVU[\]Z^_XX̀aZ
,""((&,)!"')%""!*(!(!" VU[bcUdW_XU^XefV^Wg
(!),"-.',"'*""!*+"' !'/
(!" ')"0,&,"#%('!'&&&"%1,"
"%,&) ",!(!","-2%0'!("*%(%*"'& hijiklmnopqrnmsturvwpxpyr
!"'!%1,""')0%)"')%!""!*+'&/
,' ('!'&&&"'!"%$ 2'&3-4-2%!"%&)
0'",!0%'""'((&'" )!05% 9('&z)""!5'" %5,')"('&'
00,0 &+&"!0('6'+!0(,)" ' %!(!"!,!""'"-2'"
')%!' !'5" $%%'"!/ !(!"'! %&&00!'"5!,("
6'+!"'!&7&*',"(!&0"-8!%!"%/ ',{P|}P&&"&,"!) '&'*!)'&&'&&)
&)"'!"0&'! 0'*$'*")!"5# ''",)-2%"&&"'!0)~)(%&'&&&"
$%%"'!"0'75,"!)/%! :"7/&7&&"<!'%!%' ,!":!+&&"<')
0('!" (!),!"-9'&5") %("*/ %*%'+'&"(' ','$7-$&&")/
%(%*""&'!'!*)!0("*%(%*"'& !'!0 %",!!,)5(%&,0#05!'
,"#,' '&","))"!"/ %'",)"!,,!')0'7'$%
"!*%'5" (!),%'!'!""'"', ""!* !+"-(!'%(%'",)
5!)&+&"-9%!'!0'*(" 0'7"'$%%,")6,)+!0
"0&'!*- %0,%')'"0&,&"'!&+))
2%!0')!%"%'(!")+)'" %%'!/&7&''! '!%(5- &&,"!'/
!0' %"!,,!'), %6'/ %""!,,!""%$ >5-3-}-2'"&&"
+!"""#"%*%'+"!56, % ',)'! )()&*(!'5!(!"#
'('&*)"-2%+",'&')'&"""'! ,0'7'$%'%%!')"%'!,/
)","")&*!6*#'""('!'%'(!"'!)+) "$ &&"!00 "5'&5,"-
&!')+",'&(!( 5!'&&*:;%'(!43<# 2%'"!(!&&"0'7'$%%(!/
')=,!+'&,' :;%'(!44<->!,!%! 0'!*'" !+"+!'5'(!"*'( -
!0' ""!*, %!')!"%,&) '7",!!'"0!0&,&"'!!&'")
5'"=" %"""",%'"?&)" %"5'("0,&'%'" !+"')")%
:4@@@<!%0(!%"+ABCDEFFGFHIJKLJMNOFC '""5'&" %%5%!(!""5!"%
:?&)"#3PP4<- !'-
fVg̀QR ,""') 9"!*+'&,'

0%)"("*%(%*""') ," "*%(%*"'&",)* ='0(&"
""!*+'&,' $%'&+&"% 8 !'"&, 2%!"%&)"#'
"0,&,")) %!"%&) +"5'#6'+!
0'",!0 0('",)"#)&,
0%)"
$%'&+&' ' 8!%!"%&)"# 8!"5
%'5 ,"/'&/
(!+) )!
%'"% 9'&5+')! 9'&5'!,"*
!&'"%($ ,0!'&!"("" '")"!(+''&*""
(%*"'&"* !)!"'&"!0
')""!* )!%!"%&)"
!"("
%'"%0'%5 ),"0(!),!"# ),"55!)"
!&'"%($ !')/!&'"%(" 0'%!(0z
$"0,&
01 0 2345678796
8 254
378796
87 6757574
67
f
cbb
cb
e
bbb bc
b bb cb c
QRSRTUVWXXYÿ[W\UW]XRÿ^R_`RSTâ
g A)('(!$4#/(%&$ -(H"%('!/#/8('#$4#/ #"

$#0%0%#&'1H6I!4(:#4#)&$()!#,/#'%!&'!&*#%4#/#/8('#
('*!#:#()&'%($#)).)(('*#B%($#)).)(0#0%&*#) 0!164#
!!"!#$%&'()*(+&', -(-.',&-/0(0&))(123 6E@!-!+##%('*./(/&#$#0%&'(#(!!$&(%#*(!*&/#!
#0&%4#)&./5673%(!%#8.*!5693%&,#/&'()(--##'%'#:#!%#" +&%4()',O"%#/&'()564#6F@!-8&%%##$#0%&'* '%1
/&'(%&',&':(&.!8('$4#*#'*&',! #'$(0!.)(%#*#$#0% >h$4#/(%&$ -%4#H"%('!/#/8('# )-($%=#$#0%3/ *"
!%.$%.#!5;9<-($&()'#:#$4*(%=/0('&%(!%#(--##'%!%#" #)#*(-%#%4#!%.$%.# -4*0!&'3%4#:&!.()#$#0%0%#&'1
/&'(%&',&'%(!%#8.*!1>!!"!#$%&'()*(+&', -(%(!%# 6('!/#/8('#!#,/#'%!(#!=/8)&D#*8=%4#$=)&'*#!('*
8.*123$'&?#*#0&%4#)&./523#0&%4#)&()$#))!%4(%/(=*&-" #B%($#)).)(('*&'%($#)).)() 0!8=%4#4#(:=)&'#!$''#$%"
-##'%&(%#&'% %(!%##$#0%$#))!5@3%(!%##$#0%$#))!36A &',%4#/1i#'#%&$())=:(&(8)#!#,/#'%!&'$).*#%4#8(#)! -
%(!%#0#593(B'!-/0&/(=%(!%#'#:#!/(C&',!='(0%&$ !#,/#'%!jj3jk3('*k('*%4##B%($#)).)() 0$''#$%&',
$'%($%+&%4#$#0%$#))!1 !#,/#'%!kj('*kjj3/(C&',%4#!#!#$%&'!$('*&*(%#!-(
#$#0%0$C#%1
64.,4,#'#%&$#!#($43%4#'(%.#('*%=0#! -
%(!%##$#0%0%#&'!4(:#'+8##'$4(($%#&D#*1 4(:#(8.%lNNmllN(/&' ($&*!('*(!4%#B%($#)"
;!+##%38&%%#3('*./(/&%(!%#!3%+ -(/&)&#! - ).)(O"%#/&'.!7($4/(':('*7#(.$4(/03FNNH1
#$#0%0%#&'!(#-.'$%&'()3%4#6E@!-!+##% 64#%+ -(/&)&#!$('#B&!%!&*#8=!&*#&'%(!%#8.*!3
('*./(/&('*%4#6F@!-8&%%#%(!%#!164#!##$#0" 8.%(##B0#!!#*&'*&--##'%$#))!h.,&%(3FNNK164#
%0%#&'!4(:#!#:#'%('!/#/8('#!#,/#'%!$'" -(/&)= -6F@!$'%(&'!(8.%nN($%&:#4./(':(&"
'#$%#*8=&'%($#)).)(('*#B%($#)).)() 0!4#'$# ('%!+&%4lL&'%($%,#'#!$.#'%)=C'+'7($4/(':
GH6I!J1;&,.#F1K!4+!%4#((',#/#'% -( ('*7#(.$4(/03FNNH1o&--##'%6F@!/(=8#$"
H6I +&%4&%!,#'#%&$())=:(&(8)#!#,/#'%!3+4&$4 #B0#!!#*&'%4#!(/#$#))!164&!/(=#B0)(&'+4=
&!()! %4#!%.$%.# -%4#-(/&)= - *#$#0%! / !%8&%%#%(!%#!.8!%('$#!(#!&/&)(&'p.()&%=('*
('*%4#:&!.()#$#0%34*0!&'164#6E@0%#&'! *&-?$.)%%*&--##'%&(%#164#'./8#('*:(&(8&)&%= -
4(:#(8.%LMN(/&' ($&*!('*()(,##B%($#)).)( %4&!-(/&)=/(=8##!0'!&8)#-%4#(8&)&%= -/(/"
O"%#/&'.!3!/#%&/#!#-##*% (!(G:#'.!P=" /()!% #($%% (+&*#(',# -/ )#$.)(!%.$%.#!
%(0*/(&'J(-%#%4#4=0%4#%&$()0$C#%!-/#*8= (/ ',%4#:(&.!8&%%#!.8!%('$#!164#46F@lL:(&"
%4#0(&#**&/#-/! -%4#!##$#0%!164#6F@! ('%4(!8##'&*#'%&?#*(!%4##$#0%-/ )#$.)#!
012 345678954
9 77954
4674 75 6 0
!"#$%&'()*+',"()!"#$%&'()"- ;)"";&&$',?,"&(&"!)""#"(;;&&$',&)#1
#,./ 0121!)(!$%&'()'%-&(3'&")"' !()&")?S90?'#&&")""!&()"%%>+#(<#,
4"#"&'%%$*",&&"5*%'#,#"67"+&&'(# >"+!?8@@I-6'(%,")<"!!%"+"#&
'#8.9:)"!(#'*%";()&''#")'&",'#"#'&'<'&$ &(&"&"F&)%"3')"%%$&(4&(;&"
<"*""#',"#&'=", >;""&%6?8@@ABC'+ "&%6? +'#'4#%;();&'#&"()%<'&$6
8@@9-6 "&&"*,&"+"%<")"(#&'#",'#!"'%1
"D.(+!)'"(#%$&)""!"!&',"'#'# 'K",&)&)"(#'&'#4(;*+!#,4)((<"(#&"
&3((+*'#&'(#?;()+'#4"&")(,'+")6/#",'+") &(#4"6"&(#4"'#(&+((&#';()+);"6
'&"D.DED.8(+*'#&'(#&&'"#'&'<"&( "!!"));"'(<")",3'&+%%(#"1!",
4%&+&"#,&;#&'(##++'&&")""!1 =%';()+ !!'%%"6""")<"&&'%";#&'(#*&
&()6"(&"),'+")'D.8ED.9(+*'#&'(# ,(#(&(#&'#&&"*,6N#&")!")",+(#4&"=%1
&&;#&'(#&"3""&)""!&()6"++'#, ';()+ !!'%%"?"!"'%%$(#&";)(#&#,",4"(;
3""&)""!&())""F!)"",'#,';;")"#&&&")""!1 &"&(#4")"%'4&%$%)4")+)((+1!",;#4'1
&()"%%6>(&&"D.#,&"8.)"G1!)(&"'# ;()+!!'%%"?(;&"#+()")",,''#(%()6""+%%
(!%",)""!&()G.-)"(%;&()$#,<'% *&&(#1!",&)&)"(#&'#;)(+&3(&(;()&&"
)""!&()6"G1!)(&"'#'#'#&)"%%%)+""#4") *,"?(#&"<")4"M)<',(#?DUIU-6")")"
(#'&'#4(;&)""*#'&?('&",&(&")""!1 (<")#,)",(#"',"(;&"#&")'()&(#4"?
&()'#',"&""%%+"+*)#"6H&'+%&'(#(;&"&&" 44"&'#4#<")4"(;"<")%#,)",&&"*,
)""!&() '6"6?*'#,'#4&(&"I1J-%",&("!)1 '#&"#()+%,%&;#4';()+ !!'%%" J'%%")#,
&'(#(;&"G1!)(&"'#*#'&?3'#&"#&'<&" >)&(Q?DUUD-6M%(#4&"',"(;&"&(#4"&")"
(&")"#K$+"$&"+3'&'#&""%%?'#41 )""<")%!)%%"%4)((<"*(&&3(1&'),(;&"
,"(;+!%'=","<"#&6L(&*%$?G1!)(&"'#*#'& 3$*Q;)(+&"&'!&(&")((&?%%",&";(%'&"
+$&'<&","#$%&"$%"?%",'#4&(!)(,&'(# !!'%%"6T4)((<"(#&'#"<")%#,)",&&"
(;$%'MJ#,E()!(!(%'!"?!)(,'#4 *,6/&")!"'%'K",&)&)")"*(&"<"#%)4"
'#('&(%&)'!(!&"N9-H4'&?8@@0-6>(&MJ *&&(#1!",*+!))#4",'##'#<")&",1R(#&"
#,N9";)&")&'<&'(#(;'#&)"%%%)+"1 *Q(;&"&(#4"?&"')+<%%&"!!'%%"6"$
#'+&'<&'(#()'#&'<&'(#(;'(###"% (#&'#"<")%#,)",&&"*,'#&"(&")4)((<"
'#&""%%+"+*)#"6"""<"#&%",&(%'+ ()+(&1%'Q"=)"&&))(#,&"+6&"*,)"
'#OF())"%""?3'')"P')",;()*'#,'#4(; %(%(&",(#&"(;&!%&"V&*"'#,3")"&"
#")(&)#+'&&")<"'%" !Q"&-&(&""%%+"+1 ),()*(#$!)&(;&"!%&"&(!?#'+!()&#&*&
*)#"#,)"%""(;#")(&)#+'&&")+(%"%"'#&( (;&"#(<")%((Q",)";()"#'#4&&"6")((&(;&"
&"$#!"&(&'+%&"&"('&",&&"#")<"6 &(#4"#,!!")!)&(;&"&)(&)"%("#'&'<"&(
H%&#,()&&"+"#'+!!")&(3()Q &&"6W)"P"#$(#&(;&&"*,(3&&!"(!%"
+()",')"&%$(#'(###"%?)&")&#<'G.6 3'&'4")&&""#'&'<'&$&"#,&(!("+()"&&"
H(,'+"#&")'#4&""%%')"!(#'*%";()"%%+"+1 *,>)&(Q"&%6?DUUX-6
*)#"!(&"#&'%#4" #'(#'E"%"&)'%4),'"#&- W(),';;")"#&!')(;#")<"'##")<&"&"&(#4"&(
('&",3'&%'+'#OF6R)'('(###"% +Q"(#&&3'&&""&)&)"6'+$"F!%'#
<"*""#!)(!(",;()+",'&'#4%&$&&"6)(&(# '#!)&3$&""#"(;&&"')"'&#&&(,'""?
;()()&&"+$"#&")&&")""!&()"%%#,&"# &)+?#,4'#4?'#(#&)&&(&""#"(;+"%%
&'+%&"'(###"%&";+'%$(;',1 Y"';;"#*?DUUD-6";#4';()+!!'%%")"'##")1
"#'&'<"'(###"% MHN-()!(&'+(#,1 <&",*$&"(),&$+!#'*)#"(;&";'%
&#"##"%>+#(<#,>"+!?8@@IBS #")<" )#'%#")<"RNN-?3''&#+"41
(#"'(L"&"&%6?8@@IBH4'&?8@@0-6T<',"#" 4"&?)("&""),)+6'')'&()(&"
!('#&&(&"'#<(%<"+"#&(;+"+*")(;&"&)#'"#& &%%$!")+'&&",+(#'&()'#4(;+#&&"#")<"
)""!&()!(&"#&'%;+'%$'#()&)#,&'(#?!"'=1 '+!%",)'#4)4")$(#&"+',,%"")S'+#&
%%$+"+*")(;&"!(%$$&'Q',#"$,'"";+'%$ "&%6?DU0A-6"4%((!)$#4"%#")<" )#'%
(;)""!&() CS?(#+",;)(+&"$#,)(+"'# #")<"NZ-"#,*)#"&(&")")(;&"&(#4"#,
3'&"$3")"=)&',"#&'=",-N'+)"&%6?8@@0-6 &"<4#")<" )#'%Z-&(&";)!(&")'())"
.""#&3()Q%(44"&",&&""#'&'<'&$&( (#&"&(#4")((&6"4)"&")!")='%!"&)(%
01 2 34567898
79 3654898
798 78 68 685 78
!"#$%!!$ A7:: $$*7#!6$!(:
&' !!()%#!*+,-./01(*+,-23456 g%$$h(!7$$*i$7($
7!$$!$87!$%"( :$(7k:#!7:&')_3($($$7k
6#7*$!(9$(:% $$!$j $:%$%&e'l3(j
#7;($$($$ #7:%$%&_'l3&D;:7(D*
74<=:%!*$!($87!; +,-?34A$$ $($#7$!:
#77:"!! $!$$*$(:%$;87!!6 $
&>!!#$*+,?@34 0>!$!7$(:$$4A$$ $$:9
)!"%!6$:%%$7* :$(( m#!$6:h6i(7
$$%(:!7!$%$( $$:!$$$: 7!!
7$ $$7$$%!:(7!9 $7%$\$4h)"6ih:6i!"$
#$$%$4)!"$$(7:( &0#(;*+,,j34A$%%$#!79
$*%!7#($*( :($%$!$:7!(#!\$B9
#%$(:7%!6$($#" "%%!$$:&!65$37$$*
$$!%%6(#%%$4A ($%%$$$7%$#g%$*
$7##$$$!"6#!7::6%! =:%!*"(]7!6 7$#$$:
!#(B"%%&C:#7 ( &cd'6(e$*+,-234c(!$71$*
D*+,,@34E$!" $7!!6$$6 (*$:!$
$$%$$:$!"$64 (!7#$&G$7\!4*
5!$$:$%$ ($$: +,?@34'6:!$!7(#7:$%$$$$%9
87(*$=$"$##7;(9 $#!7:&)(l!9)!:*+,,+/
F(;7#B;$6$:($ )7#*.nn234
$$%$*77!!6$%&'G76*
+,2234
HIJIJKLMNOPOQROSNTUVWopLSNLNTUV
HIJIHKLMNOPOQROSNTUVWXYLZTNTOM LVpqTrNYQOsVNOQLRNTUVM
[7$%%7!87!$"%%$($ A$$$$;:%7!%%9
$#$7#7$6&G$7\*+,?-3 $!$"%!!!$$$$:!!*$9
7$$:;$787!$$7] $6(%*(:=7 $45(%
:$%7%$$4A$!$$!$87!9 (]($($ $%$"$$7(
$$;*$!6*$7*(4[7$$ ($$$:7!4e $%%6
"%%$(1#7%7(:9 $$6$6$:$!#$:
!#$*:$!6:!!*$#*( #$/$7$87$!6$4E:
7::4^:: $!$$$:7!(6$!$ !#!6(17$(:!"!$:7!*
#!7:$%($45$#6 $:9 $%!!6:!*!*(:!$$$4
!!6 (7(:%!=!$$$4A$ l!#67!:#]$*
($7$$(!;4A:!!$ $$9 7$\$$$(%4c76$$!6(%
!!67$(($$($$$;$ :!"!$!#*$;7%9
$7$$7!:9D( $$$=%( #(\:"4E#!!67;
$($($&_7$\()$$!*+,,+/ $(7: 7$!"*7$##7
`;!$$(a;!!#*+,-b34A!$$!7 ;(F(;(%$#9
$87!$%!6$7]($!! 0>!;!!%(7$$"$!(4
$$$$&cd'6(e$*+,-234f;"* 5(% $$!6(:$( $$:9
6($:6$=%$(":9 7!7$:(!!(
::!$*:\#:877$7! #7*=:%!*;$!7$B;("
%!$$6"!74 =((#77#:&DF*+,?,/
012 345678954
9 77954
4674 75 6 2
!"##$%&'()*+',(,+*-+*((.*( H *+'=.(+=(.*(=(,+,*()(4
**.(,+*/,00',*.11.,.-,(+(/+%2+/3 '=+-,;(*+=',==((.*(*(+*)+/1.(,.00
/)()*(,-0*,*+(*+.(0+(+00'.*,.(4 ,),6,(++-.*@,5,(.(,+*%B)*.*+0(,++=
,5+,10*.(,0*(,-0.(,+ ().'.1(.(,+,*0** A,,.'*+*,*0***/(()..A.0+4
+6*(.',*+-.**',*.11.*7,*0-..' (.(,++=*+*(.*('.0+7,%%/)()*+*
2.01!"89$% ,()(/+*+0(,+*,*,A,-+0.+(.(,+$%
:+(),-1+(.(',*+3.+-1.,';14 E,-,0.0()-,;(,*0**6,((().A,-+0.A,4
,-(*+(.*(.'.1(.(,+%<+(.(,+*+=>.<0+ ,(.*('.0+%B)5.01.((,*().(.00=+
.+()(.*(.(60+/().'.1(,5030?+=/),) 0.**,.0(.*(A.0,(,**)+/(),*,),6,(+1.((
1/.(/.*();(-;.-10?/+0'(.@ +--+0.00'-,;(*11**,+7.'
++()(.*(A.0,(,*%B)*/.(.=(*.0(.'.1(.4 .(+*)@!"89$%J-.=++'*()*,(.(,+*.
(,+.(.*(*+.'C+6,((%D.((.*(**/( ,-1+(.(,'(-,,5()+3.00.11.0+=()K.4
.=(A,,+.,'.'(.*(*6,((.=(*+* 3+*.')+/().6.0.'%F+;.-10,=,(
7.'E),@ !"8!$%F,5G%8*)+/*() 63.5*.'/,*()*+**+=.,'*.61.4
*1+*(++(.(,+*+=>.<0.=(',==( (,.00-.*@'6*/(**=+-*5.%B)*5.()*
.'.1(.(,++',(,+*%H6+3().'.1(,5+(.4 *3*.'.0+0?.'',5,(*+/10.*.((.*(/),0
(,+ (),*.*.0((.*(%H(().'.1(,5+4 '.*,5(),(*,(+=/).(+0'6.+6L4
(.(,+ (),*0,((0++(.*(%0+/().'.1(4 (,+.60030+=*+**7M./0**!"88$%E+-+=
,5+(.(,+(),*.*+I6,(((.*(().(,* ()*-,;(,),6,(,+==(*0,@(),),6,(,++=
*(+5*(/)/.(,(*0=,*1*('%D.(.(.@ 6,((**6*/(**.11.(+*,',()(.0
+.++=()=+A.0,(,*'1',51+/).( 3+***(-7M./0**!"8"$/),0+()**).*
).*1'',(%B),**)+0'.0(**+3.0.(,+ (),),6,(,++=6,((**6*.0(.-+0,@0'
/+@*(+()'=++(+00,5+.(0.*(+*,'4 (+1,1).0-).,*-*.(()1(+*()-*03*
,5()==(*+=(.*(.'.1(.(,+%+()()*+03( 7N+O.'.(+*)@!"PQ$%
.'()(.*(-+00*()-*03*.0,,(**+ B)..=/;1(,+*(+()1.((+=,),4
*1+**% 6,(,+/))1.'',(,30.(,+*),1* *+-(,-*
RSTUVUW B.*(.'/.((.*(+=>.<0=+00+/,5',==( 1+('%0+/().'.1(,5030*+I6,(((.*(,*1+('

.'.1(,5+(.(,+*%B)@*)+/*().'.1(,571(.(4 .),5.-.;,-- /,()/.(%7=+- 7!"##$
-($+(.(,+*%B.*(,(*,(.)*.-,,--.(.) +1,5)(!"## 6()H-,.X*)+0+5,.0H**+,.(,+
.'.1(,5030%H6+3().'.1(,5030*,.*'*.0((.*(,* 1,('/,()1-,**,+$%
01 1 2345678796
8 254
378796
87 6757574
67
!"#$%&!'##( ##,##)+ '#')
))#'!##"##!# #*$%D+ )$,#.
#*$+$,'#,#)#' )G$##'#! $'' +'
##)'))%&+(-+#$. ##*$%0)'# $-,#
#'##(#/0!!1+ - )G$##H$#*$,$ (#
2334567%-2383,9%:+.7+# +$,'#(#G$#$#*G$#.
) !"!# ##);<=+##,. # $#/1+ -23A39%1#7+# +
##,(%:!!'##( ,$ ),# # $,!'.
,!!>##%0##)( $, #G$##%:## G$#
$##*$ +#'$ " . #(!!#"%:!,#!
# /?"$#-23@A9# "!#. $#"+##"+#-#!+#
#(,##"'$ # $"H#/+H$9#*$%0+#
,'!#"#)# ))/B"- +# $)#"(! + %
23829%0 )# +# <##!-#"+#)#"
+ ) #+# +#(#"7 +# +7.
$"%CB###++(. #"##/IE,,!J9%: #'!'#"
#!7,! # #" ))$'')+#-#"
(/D$7%-23A85;$'!%-23AA9% #"$")
:# !*'# ),>#)) ) ##,##%0 ##)), ##*$
$)#)%0#)!'. )'-'#!+# %K,#L
##(#!''#+#*$/0!! +#*$)$ $-)*'-
1+ -2334567%-2383,9%: ) $$! $'' #),# + +#
!"###*$ )+)E( ,,(%D$+ #$ #
#"#"-$'#(#" ##) #*$-##''#)))
#"#"!% ##,##",#G$! +
6#!-7+'' #*$ $'. #/D#D$'-233A9%:#
' #+)' ))# *'#)'E(
$#'F6#"$4%8 + ) #(#$ ) + #%
MNOPQPR ;#*$ $'' #%:STUVWXYTS+ ##*$%:^]fgVWXYTS + # )

'#(,# )G$##/ZSST[\]^\ST_9#*$ +# '# $-$#" + $#"
`%̀`Àa;'/_bcX^T_9`%̀`4da;'/eWTY ,# #$$'' (/)1+ /23A39-
\]^\ST_9)+#"'#+%;#*$ $'' ## '!#"23A3-,!0#h!"#0 ##-
+,!$,# + +# ' '#+#'##9%
012 345678954
9 77954
4674 75 6 22
!"#$$%&%'%(!)!(*% .-896-?29->0<N-9C3?>29SbU>5H32738-30=/68>08
203?4/9>./9/799C/82?89307/399C/7>07/09639->0
+%%*'( ./9/79/.?:@>89=/>=4/ABC6/8C>4.89/89838N/4438
639-0E85>6?-99/60/883?>1/9C6/8C>4.?>9C344>N.-5D
,-./-0.-1-.234.-55/6/07/8-09389/8/08-9-1-9:/;-89< 5/6/09-39->0-09>L9389/6MP8/08-9-1/Y30.L0>09389/6M
=369-724364:5>6?-99/67>@=>20.8ABC/?/89/;3@D P-08/08-9-1/YE6>2=8P_3N4/88<STàYAG>09389/68C31/
=4/>59C-8-89C/E/0/9-7344:-0C/6-9/.-08/08-9-1D 3@>.-K739->0-09C/BcJV]U`9389/6/7/=9>630.8C>N
-9:9>7>@=>20.87>093-0-0E9C/52079->034E6>2= 38-@=4/d/0./4-30=399/60>5-0C/6-9307/Ad30:>9C/6
FGFHIIJ9:=-K/.?:7/693-036>@39-79C->26/37>@D ?-99/682?89307/8827C38e2-0-0/348>8C>NN-./136-3D
=>20.8ABC-89389/L?4-0.0/88MN38=6-@36-4:892.-/. 9->0Pf>R>@2R3-/934A<STTUY<?290>0/8>.63@39-738
28-0E9C/7>@=>20.=C/0:49C->26/3<>6-E-0344:7344/. OBH30.O]ÔA
=C/0:49C->736?3@-./>6OBHPQ43R/84//<STUVWX>;< ]/7/09892.-/8C31/-./09-K/.C:=/68/08-9-1/E6>2=8
STUVYAZ2/9>9C/=>9/09-349>;-7-9:>5OBH38N/44 >5L82=/69389/68M30.7>2098>5=3=-443/30.9389/
38-989/0./07:9>E-1/>5530>.>6<@>6/6/7/09 ?2.836/7>66/439/.N-9C9389/8/08-9-1-9:30.6/8=>0D
892.-/8C31/28/.9C/7>@=>20.[D\D=6>=:49C->2637-4 8-1/0/88Pd-44/630.Q369>8C2R<STTSYAZ2/9>9C/
PO]ÔYNC-7C-8C-EC4:7>66/439/.N-9COBH6/8=>08/ /0C307/.96-E/@-034-00/6139->0-0827C-0.-1-.2348
P_3N4/88<STàYABC/-689627926/836/8C>N0-0X-EAVATA N-9C3C-EC/6=3=-443/./08-9:<-9-8=/6C3=80>9826=6-8D
BC/@-0-@2@./9/793?4/7>07/09639->08P9C6/8C>4.8Y -0E9C3936/439->08C-=?/9N//0O]Ô8/08-9-1-9:30.
>59C/8/7>@=>20.8<OBH30.O]Ô<5>44>N3?-@>.34 8>@/4-0E2349379-4/8/0839->08827C389C/8/08-9-1-9:
ghijkjl OBH30.O]Ô./9/79->09C6/8C>4.8Pmnopqr\nmsY30. O6/88YAG>9/9C39OBH E-1/83?/99/68/=3639->0>59389/6

=/67/-1/.-09/08-9:639-0E8Ptuvwpqr\nmsY>5aAaaaSd OBH 30.0>09389/6E6>2=8</8=/7-344:N-9C9C/=/67/-1/.-09/08-9:
30.aAaaax[d O]ÔP56>@ _3N4/88PSTàY<?:=/6@-88->0 639-0E8A
>5y05>6@39->0]/96-/134_-@-9/.Py]_Y30.^;5>6.z0-1/68-9:
01 2 34567898
79 3654898
798 78 68 685 78
!"#$%&$$'$$( )&##$&$10$0)4/???;%#&$#$)$'(
#)*+,*))---.)$ -'#-% $)-#'$.#/$0))-%$%.
))---.-$))'-/)''$-/ 0#-0)!C%#&$#-$ )
$)0))'0)-'-12$)3/456578###!/ $)$'-% ).0-'''#'(
45697:$$$ 2$)3/4559;!8<$/&. $.$'0$)F$)-#'$.
)'$$#-)$#< )&#<$ 0<.!@$#) #'$.))$.$)
'$$#-)&%$--#))1=$#!/ '$%#.9y>9&-$#'##)-'%#&$(
>??9/)#)@'-$)- A$-B$)/4554;!C #)1A-$)-/>??4;!C&-$#'##)-$)
'$$-<//-)) '&))-)$ F)&$'$#$-)$'$)!C#'(
&)#.-0).)&) ))$.0$)( $.$)0$B'&. -$)-)-$-/
)-#)# '-)-$.-%).%$# )&&'#)#.))'- <-&-/'/
&$3$#-3*+,*))---.)0$-'$--( &&$.1=$$/4554;!
-#$)0)1=$#!/>??9;!" 0-##. }#-3)$'0$)$&-D 0-(
-&0$D-%-)0$))<$-))-- #-#'##)/#'$.$'0$)$$$'##)!
-$'&0 )')*CE<-##)< C.$)#$)-.#-&-
)&&-F$)00$))-')1)-'.0$'- #-)0~.$$0#' -&!C
-$))/$-)----;$G$) -#-.#'$.).)& &---)'(
1H<#))/4565;!C-)-##)$)&$%$#0$-( -#''-)-'-)$$
'-0#/ 0-%0<<IJKJL)) $0#'&-)%$0#$#)'-'!,$
-0$'/$G$)&.' $ 0$)$)).)& $-#.$%$
)/-)-&-#$)-'$#)$& < &%/)#'3-%&.)$.#'$.
)00$))-! $0#'&&.0$-D)))$-%
$&$)).)& &%!C#'$.).)&-)
-&&$& &%/<$!"'&&-B$.
''$)<#< )$)$D$)
MNOPQRSTUVRQWXYVZ[T\T]V $&#'$.$'0$)).0))$%)&##
0))%)-.'$-$-$&0#-$<.-
RQW^_Q`S[TQZTabUc\\ #'$.#)!C-)-))&-&))#($0-$-%
defeghijklmnjiopqrrsrjtusivkwli --%) -)/#-%---#<-'(
))&##/ $$ 0$- &)
G$0$'0-$#-!@)$.0##(
C#'$.$'0$)$#' -<)&##0$( $) <$'--##))
-)0-#-&$.-%-)#'-.!C-) )&##/'## )&-/ )'$0#-))$))$.
$&#'-&.)$)&0$'-'- #.)-) -)<-))#'-)'))&##
%-)&%/-#)&)#.)&##0$( --D'-))1./4554;!
'%-$$))')G<-%$% C&'-)&)$$'0-$<<##
)'##.$'-'--.))$.$%)! $)/)$-%<- -)'$.&-#.
x$$'$-)'$/#'$.))) 4/???%)$#'--&&&#)/ -)'(
)$#$-)'-)))---.!C$$ $.$ 2'3 "F##*$-->??
)$#&-##-$'0$)')-) 12'3 "F#/4554;!C-)&.)-%##$%)
.$&-#30$$-%-&') %&-#.-&%&!"z9?)
<->?yz?$.D'-#-<-'{G|- $'0$.0)$'--&)!C$'0$)$
&')#.$1A-%!>!4?;!,'-)'-#--) =(0$-'0# $'0$)/#-3-$$'0$)
-'$))$'$'##/F0)-% -)#$'0$&#'#)!C.)'
$'0$)'&-'#)-&#-!C&-. --'-%)$)&&$)%&)''
#'$.$'0$'##)#-)-)-0-#-& .-$'###$ F$'###$#0) )$
.') -F-#'$.#)! ($&-#)/#-3-$&-#.C>+$'0$)!
"$&-'#&0#-D'-'$-) --00-)')/$$$&4??
&-##-)$'0$)) $D$)-&' $--#-.1A-$)-/>??4;<-)$% -$%'-
012 345678954
9 77954
4674 78
!"#$%&!'()*++,-+-.-,-/*012,234/3),-.,-)3+5/2/6)78,-/37*,3+*),-+-.9*,9:-35;:2,-+*;:-)1,45*72*.<-9/*)176.7-//:37-+3*</:3-,/3)8*)93)37)605*32
*.<-9/*)13;8/:3.860=>4*,/)-9/+.3-?8,(/:3*.<-9/*)13;8/:3.860;-++8,/*/:376.7/:)*6(:+0-..*;3,8,(+/:39)87)8<*)0;.-/3=@!&#!A$%&!'B8-()-0*<*.<-9/*)13;8/:3.860=
98298.8-5B2C3,C)8/89/3)08,-/8*,*<*.<-9/*)1)393;/*)93..5*)2*.<-9/*)1)393;/*)93..+5+92+6;;*)/8,(93..+57927-+-.93..+5>4-4*,76,C.3+=DEFG#H%&!'7-+8993..+/1;3+8,/:3
*.<-9/*)176.7=I.<-9/*)1)393;/*)+JKL+3,C-4*,+/*(.*03)6.8J(.L/*0-M39*,/-9/N8/:-;89-.C3,C)8/3+*<08/)-.JOL-,C/6</3CJPL93..+2/:3*6/;6/,36)*,+<)*0/:376.7/*
:8(:3)+/)69/6)3+=Q)*++R9*,,39/8*,+J+*038,:878/*)1LN8/:8,.-13)+-)30-C371;3)8(.*03)6.-)93..+JSL2+:*)/-4*,93..+JTL2-,C()-,6.393..+JUL-+N3..-+)396))3,/9*..-/3)-.
-4*,+J)9L=020696*6+.-13)9*,/-8,8,(*.<-9/*)1)393;/*)93..98.8-5*32*.<-9/*)13;8/:3.860=

01 2 34567898
79 3654898
798 78 68 685 78
!"!#$$%!&!#'# UVWVUXYZ[\]^_^àbY``
(&)*)+,)-#.%/#01! ! !
$ 2$ 13314%5678 2! C&%/2/&#1!%!;#2
$$%!)9!2!&4!#$/1/#'/&: F;#1$#$ ;// %!$/1/#
!#,%!!#314##3;!"1/
"1/2!<3##!&/1213#! #!#&% 2/1213#)c#!
$/#2#$#)9!!1#!81#8 ##1: !#3 <!/!#!1!!$ /
1##/138 #!%#31"1//2! ;#!
#!&$#3#!!$
1!/ !;2#$/$/1/#'= /) F;##$#2 ; 881$4!8 #
%2#$//)L!!12
5>>?,)
-!1///$1!#$##$/! 33/!/B1
/ 1 /
B
$ #$#!$$ !2
#$/#@:13/13#!#) #%/#!#!&$#F; /12#!###3!:
A!!&13#/#! 1;! ";###1% %;)E #!#$3/
@:3!#%!#81!! 1;!B2$# $ //2#34 ###//2#1$ 88!:
//'E32/)
#1 # !2/2/121/#)-#!#C-D!121/1 5>OOdE32 !I !5>?J,)-/
CED81!! 1;#;#!1!!/#) /$!<$3/#;!$
!11

$/$!
C!!F<G H !I HH!#1##!# #'81#!/2##8!%,% $
1//%1$/##!&;/21& !8! 3! $ 1$3!##!$
#$$%!3!/1# *J$K#/ !3##3!!#/1;2$1:
! 1!3!/#&!;/#8! !#//2,33#$#!"!&)-#
!; <! !#/#!!!#$/#)-# &!'
# ! $3/"1!31## #!B2$1#1 #!#/&2# #///1! #% !<!/
1!1 % #!$/1/#1CED3 5>?*,) 2#$ ! !!!/#!#2#2#$'QB!
#1! !!##!#!#1!1## C#$3/$!#!1!1!;!1 !2!
&13!1!!/'(#!*JJ5,)-1/: $3!1#!!/#$//!&!#/
1$!# /#3! !8 F8!&1/! #$
1!!/81#;# # 4!!1///%2 8// /)-4 #3 #$3/%;&R1
/!&!#3!1#)-41
!1$$%!3!/1!&) ! #!#!#3#!!$3$ /2
L!M/# #!!#3/3!# #8 !#
'N 5>?O,)P1131//<3###!/2 8/43!& ##)G8#8 //8#$3/ !
!23133!)Q131//#8 </)9! %$ #/#!#/# !
#$3!3R1#$#&/$/) 33)D!1!&!!## #1 #F;8//
1;/2%/14#
S$/# /#!$ 3!;/33!& !#/3##&F;;
&!#'(#!*JJ5,)S!# 3: 213#'E32 !I/!5 /#3/1:
#!%2 1;!!#&$!# #1/%2#8 //8!& !> ?J,)c!
</!&
/12%/%)T8;$ ##$81$: 1!%!#$//%1$#1/)-
3/1%2113# !$/3/ 33//%/!!/#$//# #.##03 !!12
$/1/#!1!;#/2$ !2$/1/# 1!%#1/$##8!##!!1%$ %/2
1!#$/ 8 213#)-##/ %2;//F;##1 #;!// !$ //)-!#
1$%!/1M/2'E /!1/) # #$3/$#/1! !$#/%/!&#!#:
5>>>,)-%!1&!B#3!#3!# !'#I3>,)e!!&#!&# $ 1#
1## 2!!#!.10! $###!"##4#!3!/!:
M/223)K8#8 2/1: !& !2#!#2 !/2##F;)G
! 33#%3231/3!1&!! ;//#!/1;2$ 2 /#;#$/:
$1!#$)S1 11!<3/!82#$: 21#%##$ %/$
!#1!&!M/28!1!1!! &$!/#$/#1 #$!/'T /3!*JJ?,)
!1##)C!/13#8&#: 9!$
/#1$3! 1 #1!1!!# M/#;#/2# 31#!#/!!#/#$//#
$/)
!1#/!&3! 2)
012 345678954
9 77954
4674 78
!"#$%$&'&() ""*&+ ($" .("A)&(1#0#"" !(#"9!&2
((&"$""(#+(&$,#%&+-.$-#/!#& (#.&A)&(# .("-(&%.!4c&(8$-"1
&+#(#& &0 -&-"0""&01&2!""% "#$& &+ .%$(,(!&$-&.%
0""&0 ('3.("415665*47&0)(1 +&(&(/B.#!(&$1."9#&+&(/$2
#$-"%$&(#& &+8"#& 0#&.0""&02 -"&0##&+ -(0"""&0(&"%
#/&0## &&"9$!#$ +&( '](# "41defg*4.&+_(%(2
((&"$"":;,$"")&".$&+"#<.#%#& (#/`&($#"%#/!&$-&.%4;!(##!"<.#&
$&.%0#("#(&.%4=8-!&(4>& &0"2 #0(!&!(#& # -(&%.!8!%
"&0?3(# 0#"&"%#/ &-#!%.4 (&"%&($##$.$ %!"!&!(#&4
@" "-#!%(&.4;(0#"" h&$-&.%-("&0#((&"%(."#,"9
!"(#$-(#& &+)&"#"A)&(( &!&(#.&-(!#)%A)&(1"&./&$
-(!#)%9((&"$""4;#,#%&+8"#&2 .$$#& &+++!&+#$#"(!&$-&.%#
#%.!%A)&(-(!-#& '$(&+$""*#!&$2 "09-&##"#94;#,#%&+(&"%"9#
$&"9-(!#!%9B.%/.!0#(0 +&(#$#/A)&(#$-!#%#!.%+.((#
9%#++(#(&$# /"+(&$ (&$# h-(i4;--(&!._&%&(.#`j$."#-"
$&.4C&0""&0(# &&"."9 &+(&"%j)#%!&+-&#"&(9!&2
(<.#(%+&(((&"$""4;0""&0($9 (#.#&4
#$-&(-(&+-(!-#& %.(#/ &($"2 ;(&"%(#/"9)(#"&0## %
#/1.#$9.--"$%9&($!#$ !(&#%#)#%."'k0""41deelmC)"41
#&#/%+&($"&(9)".#&4 deff*4C&$#%#)#%."0# &(0# &($"
&+$""(."&%!&$+$#"#
&+#$#"($""#/!&$-&.%4;##!&%##&
DEFEGHIJKLMNOPQRSTUMUVUMPKSWQXRLUYL !$# ""%-!#b!&$#1&--&%&/("&2
&(&"##"#9&$""4C-!#b!&$#
ZSNT[UK #&-(#&""9%b%!&%##& # 0#!
#%#)#%."$""(&"%$&( 0&2
;&"+!&(9##)#9&+.$%&(#2 %(%%)##&&)-&-."#& $ !&!(2
$"#($(,"4\.(#"#9&%!$9-& #& 'a$&&("41deifma$&&(1degd*4h&$$&
&%&()(9"&0!&!(#&#"".(- -!#b!&$##!".% ###)#9&+&"2
##)#9&+ ("9""#(.$"$&+!$2 "&0#/!&$-&.%&+-&#"#$-&(!# +&&%:
#!""9#4]9#$-&(A)&(!&$-&.%( %(&&1!&$-&&+&(#'n9&!,#%
%!"#-(-(#""#& (/1.!."+.(2 3.!$-1deff*m!#&"1!&$$& (-!&$2
!&##/!&$-&.%"#,9"$(!-1!/ -&#$9('o"&#%o#""#1defd*m)2
&(,.,2"#,!&$-&.%1&-&##$-"&9% ("$""(!%2!#+9!#%#$-&(#%#(9
/&%&(#^#& /4C&$+&&%A)&(( A)&('a$&&("41deifm3( %"41defe*m
) $&(-&1"#,$&89-9(^#!&$2 %#!9"1"!#!!(#92-(&%.!'k0""41
-&.%&!!.(#""---(4\($""&(/#! deep*m(#$9"$#1b-&#"/#'a$&&(
$&"!."( &&++!#)#$."#/&"+!2 %c&(((1degi*m#&.9("%9%1(-&#"+&(
&(94;)((9&+(-(&$!&$-&.% $"9A)&('a$&&("41degi*m%!()&1(2
(-&#"+&(!#(.1("1$#1%-#2"#,(&2 -#$#%&(('o"&#%q##1degf1.
$(..""9-&# -(2-($#""#& (/4 "&k0""41deel*4a&(9-""%(
!&(1"!&&"!&$-&.%"#,&"(&"9 $.0(!-"$$(&$0
%0 #(!&!(#&(!-(-(&.2 %#++(&"+!&(9<.#-$%#$9 &
%1&"&./0$9#,&+"!&&"_$""91̀ -&#"&+&(!-"#&&"/($& ""A2
#!&(&-&!$#!".!-9(^#1# )&(4a"&1-"#0#&-!#b!&$#$9
# &)(9++!#)&%&($&"!."4 -&&(B.%/&+-(#!."(&%&(1.$9+.!2
a%/(# A)&(((!#&.$#! #& -(+!"90""& $&&(A)&(4$,"#"
!$#!" #%#b%# -(&%.!1% &8!".%#-"#+(&$-(#!#-#& ."
!$#!" &%&(0 $""%+(&$&" &%&(# <.#& #,9!&$-&&+""
01 2 34567898
79 3654898
798 78 68 685 78
!"#$%&'#()'# *&!+ c] ="2344A6$?##)'*'
!)!*' #*#"#!' -&!)##''#&'#& *!!*#"!
# !($ #"'08 e2345fd8234gA6$%&
%&,!!& '#&')' !#(#+ #*(, ]')'F! h'*')#
*',###(!!$%&''! F*"!##]''()!#0c-!234446$
- ($."''*#&'&!',!! ' >#,,'#"'!#'*##' )!#'
#/"#!'*,)'##&',!# ")!#$%& !(#*!',,''
01 234556 #&)-'"*#7)# ' h')'*)#, '( !(,'
#/"#!-01 8234946$8'!(7)'+ F*"!#$
,##& !#("#'":*###(' >)##&*,, )(*&!(#7#
*##!(!!'*#'&-#&#))! #&##&' **)#*&,'*!()'+
*"!'!!(#(!( "##&'!!' , '('2#&'/"##' ',# ,
##(08 ;, 234<46$=-'2# h'''*)'! "#*#''
!!#&)'!, , (#&$>'-+ ,!!0_'"234g96$;'",'&' *,,!+
#&&#'*!!#'#"''#!##(2 "2!))'##&)'*)#"!
1 034556')'##&##&?''*#0@*# ,!'#-#&*#'01&#'### !$234gg6
A$A$36!!#&' "#A<BC<D ', ( ")#&"'*#&''#$=-'2
'#$%& "##&'#,#&E#& #&*&,''!!(",!'##&"#
*&#*',##-!!"+ #&)' , (, !-#&#*&+
#'('"!#,"!$."*&#&)'!,- " *!:'$`'*!F*#)'!*&!!
#'##&)&(*!#,"!" *F',( )'!,$i'#2#&",''#!*#+
*&', #')&($%&##!G**#'# '!'$8'!( ##,)# #'*!F*#
'#- &&!(*''!#-#&*',# ''#&',)!F*#$j7,+
)'', *2-#&#,"!"'# **"#' )!c"G*#'k', #*2''#2
5<D #&'**',#$%&"&#'+ ,"](2'!*](2#(2')"!2 "#2#
*!#, #''*#&'&!, (# -&*&l- ', ]' #&'! "'$j
&&$ *,),###&"'#'*!+
F*#"##&h'''*-'!#&#
, (#&'!*#''',#&"'*
HIJIJKLMNOPQRSTSUVWXNQYTSYQRZ[VTU\V ,&#'!* '"*&'#$%&"#&(
!' !&(*0', !&("
)'",F7#2!#',)'##'#)'+
>*#'###!,# !#(##"&#+ ","*& 1&!m$<6 *!'!,*
#(*&2#&,!!)'"-#& ''*$%&,*!#"'*, #,(+
', ']!(-''/"!#$87)',# #'"##&"#'$_ !,*''**!"
'#F*#&-#&##&",',!+ )+-('#,!!*,-#&-#'
''))!*!!/"#!'2,!( ,!!!]!!$%&7,)!'#&/"#
-#&"))'"0^' _"*&,)2345C6$ -&#&'#&)'",'(*#'*']-
=-'2#&)'*!!'#!# #,'*!,#$j#&' ))'*&##&
($`#-]- ,!!"#*#*:"'") )'!, )')#&#'*#'
#&,2*!! #)++#&+)&,2 )*F* ,2*#&(, (')'#!*]
!(#( -'a#&#)("'#"b )*F*'*)#'#()' '!#'")*,)"
0c-!234556$%&F*"!#('!**# 0j,'234536$=-'2"*& ##,)#''"*
'-&(, (*!*!##,!!" #(#,#&#'#, !!$
,"!#)!*&*', #01 23454d^#(234436 m#&!2#&' *'! ',#
#)'##'")'!,,!!',)'!, ,-']''#F! "#/"!#(*#+
'!!!$`"',!! !!,# '',!!$i'7,)!2%!A$A&-
#& !#(# !(#*!!('*E, (*,)+ )'*#*!*')#(#, '''**+
#*,)!7',7#"'0c# !$23443d ",')'"*#'!!(', #&7)'*
012 345678954
9 77954
4674 78
!"#$%&'("#)*)*&'+* +"#'('.'#%0&'#+*A%6'1#%$&/$%06%0-'J3"+'&/+'*%
,-.$&/".%0"!"#$%&'("#/'*% ,%$&"#%.%0)*/*(#"-1*2 4/.'+%)A%6'*"+'F#-/&)$A%#%$&'#72-&&A/*4/00."&
31/$) 31/$) 2'*-FC$/'.&0)!/*&/.$&"#*1'$/C$&"'.%20'&A'1%.:
34''&56%./00%7+%0&"08 9#"4.56%./00%7+"0%**'*8 '0/*&&"$0%**/F)&A'%#"+%%*%*1'$/C$2'##)7$/&#-*7"#
,#-/&)7.".:$/&#-*
;""!)7.-&&)
,#-/&)7.".:$/&#-*
;""!)
"&A'#F#-/&J\.&A%&$%*'&A'#'/**"+'-&/0/&)/.A%6/.(
<-&&) 1%.'0/*&**/+10)'*&/+%&'&A'('.'#%05"6'#%008F#-/&)
=#''. =#''. /.&'.*/&)JE/FF'#'.&1%#&*"F&A'4A''0+%)%110)+"#'
,0"#%0 ,0"#%0 "#0'**&"!/FF'#'.&6%#/'&%04/.'*%.!*0/(A&0)!/FF'#'.&
>/.&) ?""07+/.&) 6'#*/".*+%)'6"06'F"#!/FF'#'.&4/.'&)1'*7'J(J7F"#
@'#2%07$%+1A"#%$'"-* ?%#%4%)7%./*' *1%#S0/.(4/.'*J
5"&A'#8 B./+%0
9-#.&
3-0C!/$
D-22'# ]^_^àbcdefgchijkglmkefmnopqhlehefgcr
%E'*$#/1&/6'%&&#/2-&'*!'#/6'!6/%1#/.$/10'*"F.".:"6'#0%1%.!
%110/$%2/0/&)&"$".*-+'#1#"!-$&*
sftebkmubllkmnnfgcrhcqvmimhnm
2,%$&"#%.%0)*/*(#"-1*!'#/6'!F#"+ #%&/.(*"F%#"+%$"+:
1"-.!*".GHI%&&#/2-&'0/*& B./+1"#&%.&"1'#%&/.($A%#%$&'#/*&/$"F&A'M%6"#
*'.*'*/*&A'/#&'.!'.$)&"%!%1&"#&"2'$"+'-.#':
*1".*/6'&"*&/+-0/4A/$A%#'*&%20'/.*1%$'%.!&/+'J
%.!/.&-/&/"."F&A'1%.'00'%!'#*!-#/.(&#%/./.(J KA/*/*1'#A%1*+"*&"26/"-*F"#"0F%$&/"./.'6'#):
KA'*'$".!*)*&'+ /*2%*'!".%$%&'("#/L%&/"."F !%)0/F'J;A'.".''.&'#*&A'A"+'"F%F#/'.!74'
&"2%$$"M%6"#*!'#/6'!F#"+%F%$&"#%.%0)*/*"FA-.: "F&'.."&/$'&A'$A%#%$&'#/*&/$%#"+%"F&A'A"-*'w&A'
!#'!*"F"!"#&'#+*%.!%#"+%&/$$"+1"-.!*5?/6/00' #'*/!-%0*+'00*"F&A'/#$""S/.(%.!$0'%./.(71'#:
%.!N%40'**7GOPIQR'0&'+%%.!3"-&A4/$S7GOPI8J *".%0$%#'1#"!-$&*7"F2%2/'*"#*+"S'#*7"F1'&*
=/6'.&A'!/FF'#'.&%11#"%$A'*%.!1#"!-$&%#'%*7&A' "#1'#F-+'*JKA'*'"!"#**''+ &"$A%#%$&'#/L'%.!
%(#''+'.&/**-#1#/*/.(0)1%#%00'0JKA'&'#+*F"#&A' 1'#+'%&'%A"-*'/./&*$%#1'&*%.!!#%1'#/'*JBF&'#
&"2%$$"4"#S4'#'!'#/6'!F#"+&A'B3K>0/*&"F"!"# *'6'#%0+/.-&'*7&A'*'%#"+%*("0%#('0)-.."&/$'!2)
$A%#%$&'#."&'*&A%&$".&%/.GHI!'*$#/1&"#*JKA/*0/*& %6/*/&"#JKA'*'.*'"F*+'00A%*%!%1&'!JKA'#'/*."
1#"6/!'*%-*'F-0*&%#&/.(1"/.&F"#"!"#!'*$#/1&/". .'4/.F"#+%&/".$"+/.(/.7*"%&&'.&/".%.!*'.*"#)
5E#%6./'S*7GOPT82-&/&/*F%#F#"+'UA%-*&/6'%.!$".: F-.$&/".&-#./."&A'#!/#'$&/".*J\.*+'0070/S'&%*&'%.!
&%/.*2"&A('.'#%0%.!*1'$/C$&'#+*J &A'#+-0&/6%#/: &A'&A'#+%0*'.*'*7%!%1&%&/".$%.2'1#"F"-.!5?%/.
%&'%.%0)*'*"FF#%(#%.$'+%&'#/%0*A%6')/'0!'!*)*: %.!Z.('.7GOIO8J
&'+*4/&A*/+/0%#$%&'("#/'*5VTW85X%#L"%.!3&%.&".7 KA'*'.*'"F*+'00%0*"*A"4*+/U&-#'/.&'#%$:
TWWI8J &/".*J !"#*"F!/FF'#'.&x-%0/&/'*&'.!&"+%*S"#
&A'#&'#+/."0"()*)*&'+*F"#%#"+%&/$M%6"#* *-11#'**".'%."&A'#7+-$A0/S'+/U&-#'*-11#'**/".
A%6'2''.!'6'0"1'!F"#*1'$/C$/.!-*&#/'*JKA/*.%#: /.&%*&'JKA/*/*A"4+"*&%/#F#'*A'.'#*4"#S72)%
#"4*&A'1#"20'+ *"+'4A%&%.!+%S'*&A'&%*S"F 1#"$'**"F"!"#$"-.&'#%$&/".6/%/.&'.*/&)*-11#'*:
!'6'0"1/.(%.!"!"#$0%**/C$%&/".*)*&'++"#'+%.: */".JKA''FF'$&$%.'%*/0)2'*''./.&4"$"+1".'.&
%('%20'J .'1"1-0%#*)*&'+/**A"4./.,/(JTJGGF"# +/U&-#'*4A'#'&A'"!"#*%#'6'#)!/FF'#'.&%.!'%*/0)
4/.'%#"+%7%##%.('!/.%4A''0F"#+%&4/&AA/'#%#$A/: *'1%#%&'!1'#$'1&-%00)70/S'0%6'.!'#"/0%.!1)#/!/.'
$%0*&#-$&-#'5<"20''&%0J7GOPY8JB*/+/0%#%11#"%$A 5?%/.%.!E#'U0'#7GOYH8J,/(-#'TJGT*A"4*1)#/:
4%*&%S'.4/&A%$/#$-0%#%##%.('+'.&"F2''#M%6"# !/.'[0%6'.!'#+/U&-#'*7'*&/+%&'*"F&A'/.&'.*/&)"F
&'#+*5>'/0(%%#!'&%0J7GOPT8JKA'"-&'#&'#+*#'1#': &A'1)#/!/.'$"+1".'.&%&!/FF'#'.&0'6'0*"F0%6'.:
*'.&F%/#0)*1'$/C$%#"+%."&'*JZ%$A"-&'#&'#+A%* !'#7%.!'*&/+%&'*"F&A'0%6'.!'#/.&'.*/&)%&!/F:
%.%**"$/%&'!#'$/1'F"#%M%6"#*&%.!%#!&"%$&%* F'#'.&0'6'0*"F1)#/!/.'5F#"+N%40'**7GOYY8J !"#
%1#"&"&)1'[*&%.!%#!F"#&#%/./.(4/.'1%.'0/*&*JKA' /.&'.*/&)!'$#'%*'*%*%F-.$&/"."F&A'$".$'.&#%:
*)*&'+A%*'+2'!!'!$%&'("#)*&#-$&-#'&A%&+%S'*/& &/"."F&A'"&A'#$"+1".'.&J3-$A/.&'.*/&)/.&'#%$:
'%*)&"-*'J\.&'#/"#&'#+*%$&%*+"#'('.'#%0$%&': &/".*%#'+"*&0/S'0)$"++"./.%00$"+10'UF""!
("#/'**-2*-+/.(&A'+"#'*1'$/C$"-&'#&'#+*JKA' M%6"#*J
01 2 34567898
79 3654898
798 78 68 685 78
!"#$%"&'(")*)+% ,$#"##"!-!- $5+#+#%)"#%$#)3.6.17,#!.%"+#"'),$#+$5+#

.$%%$!!"#$%"."#".+#)+.) !"+#/+#0."+-$#/ +#%)"#-4! !+$#-!"'3"3#18#$%9$:'+"'1;<=>?@A
."')*)+%12!!#+#%)"#%$#.$%3#!)4'+ .$5#+)*$,B!!9$:'1
.$!+#")+3#$05.0:*#'"),#$%%C+5#)53/ .$%3'C,$$0)$#:4#"-)'F!)%+$."!-
3#))$!"')$$..5#) !$',".+$!18-5#D1<E)$) !."#".+#$4#)4#"'%!5+)$,#3"+0+")+!-)1
"+$/.$%3$!!+$0$#%C+5#$,4"!''!"!0.!/ 3!$%!"$,"0"3+"+$!"!0#'")3#)!+
!"%"'0*01)$0$#.$%3$!!+)"#0)+!-5)/ %3$#+"!+.$!)0#"+$!),$#)!)$#*+)+!-"!0"
":'A11A+*0$!$+)% +$:'!0 !+$"!! -$$0#")$!*)!)$#*+)+))$5'0:0$! !"!
$# !)3"#":'%C+5#1B0"3+!-+!$)+$$! $0$#/,#!4#$!%!+1)+!-"-"!)++:".F-#$5!0
.$%3$!!+%"F)+$+#$!)+"!0$5+;G"'))A $,"%:!+$0$#)''"'+#+M5"'+*"!0 !+!)+*
<=>?@1) )"!$'0"!"'*+."')+#"+-*5)0:* 3#$6'$,"+4#):!-+)+01B,+#")$#+3#$0
)$%3#,5%#)1H!+#*!-+$"!"'*I".$%3+/ +$',".+$#*)*)+%:.$%)%%5!+$"+4#)
+$#J),#"-#"!.A)$%.$%3$!!+)%"*:#"0'*0)/ "%:!+!+:5'0!-A'))#)3$!)4+$+$)"#$/
+!-5)0 !+.$%3'C%C+5#"!0$+#)%"* %"+.) ,+*$..5# !++)+3#$05.+A"!0%$#
:$:).5#012,+!$) ),"+-50+$$!$,+ #)3$!)4+$$+#L"4$#)$#"#$%")3#)!+05+$+
F!$!.$%3$!!+)A+$+#.$%3$!!+)%"*)% #'"),#$%)533#))$!,,.+1)%"F)+)+!- !
+$%#-A"''$!-+%+$:%$##"0'*0!+601 ",".+$#*A,$#C"%3'A3$+!+"''*+#$5:')$%5!'))
K"++#!)$,"0"3+"+$!+$+)+#$!-)+.$%3$!!+$, ."#)+"F!+$ !)5#+"+++)+"#")$0$#,#$#
"L"4$#$4#+%%"*C3'"! !3"#+*)$% "+'")+!5+#"'!+):".F-#$5!0)%''1
012 345657845797

(("!"%$;"!%(""((
!&E ((""(#!!7
C(!C"%>!(("!($)
*($($&="""!%
(""%>"(">"$(C>7
($"(""#)
"!(""((+K"!&7899B3&L>""
MN(D+899O3"(%"%
!PQG ""(%#"!)
($7">""% *(
!!"%>!D( R+$&HS3&
TUVWXWYZVWY Y
[\]\^_àbcdcefghijekefdjedcl
mnoepceqle
r #"% (("!!%("("
$(#"!"""!("#"*!!"
"!D&E(("!!%(")
C"!%""!(!"
""!!&E%"("!!(()
!"#! """"!>%QR"(!
"$%&'("% !"#"" () ""!"$(##%"(
*$%+,--./-01.23"(" +s"6"*!78998t6"*!"67899u3&
$%"" ( *!"#+245./ K"% ""$(#>)
-01.23+6"*!789::*$3&
!" >"!#>7
7%&E%(!")!"")
;"!%"("!$("!&< #"(!$$$"$(7
!*!"(&= )"!"""7"7
>"!7""!"( "*""7!"(%"%+")(>3)
(""(;"!%(""(( ) !7(!>"!
#"!($&?"$!76">"@!( "(!>">7"" ("
+89AB3*""%7$) &(!" """)
C!*>"!!%!"$("! >$*"">(%7*("
$C!($7"!>"%%) (("!!%("(!"")("!!
"( "#"($" "!!("&(!"7%"
$ *"D&E*!>> %($ &E(
">;"!#>"($!!%* "(""v%$ $(
"" ( >"&F*#7"() "($(&
"!($*( E$"( (#
>$"(!"!%$ ""("!""!((("&K(
"%>!($&?"$!7" ((J"#"""%
(""!%+GHIG8H3$( >"!#"N >"!"(
"!!( !*"("%+6"*!7 !"#(("!#$)
899H3&?7""!J"#( ) #&% ""%>"!
"%(("!($">! C>$"$!!""""("!
01 1 2345678796
8 254
378796
87 6757574
67
!"#$%
!&'!(!#!&)&%*%)+%
,**-!&.))/#)#!&!&
!"#$%+,0)&!**!&)&
1!&&) )*%'2%3456789:;<
/%1%!0%!&#%&+!#2,0)&!**)
=>9?@A6B89:;</%1%!0%
!&#%&+!#2,1!&&) &3
C,#%))/#)#!&#0)&!**!&<
#'%1!&&) &%#$&+#
!#+$& !"%*%0%*3C,#%
))/#)#!&#1!&&) )*%'2%<
#'%0)&!**)%#$&+#!#+
$& !"%*%0%*D, E)-*%++
DFGHIJ-!#'/% !++!&,#'%
K+21'& !1L1!%#2J3
D#)+#%J&%0%M(%+!&&%0)#!&.#)+#%($+DN)( )& bcdcbefghijkjlimnkopmfnqihrh

)&O%**%P)&#<FGQHJ3L#'%+%/)/!**)%)%&#R$+#
#)+#%+%&+2.)&+<($# !.'#(% %)11$)#%*2 jstfprphufphih
1*)++!M%)+.)&+,#'%/%1%/#!&,1'!*!/%//% C0)!%#2,+/%1!)*!T%&%0%%&!&.+, #'%#)1#!*%
($&DE)-*%++)&L#%0%&+<FGHHJ3S0%&#'%#)+#%($ + )#+%&+2+2+#% +1)&(%(+%0%'!+#*.!1)**2
!#+%*,+%% +.)&!T%#/0!%#!.% !&)*)11%++# !&+P!&)&#'%%/!#'%*!)*#!++$%+3N/$/+%+,
#'%)* !*!%$3U!.% !&)*M(%+)+1%&)$&#'% &1!1%/#!&<%+/%1!)**2#'+%!&$1%(21'% !1)*+<!#
#)+#%($, !&.)1')*!1%V*!P%+#$1#$%DW'!#%'%) ')+*&.(%%&#'$.'##')#,%%&%0%%&!&.+)%#'%
%#)*3<FGHXJ</++!(*2%&')&1!&.#'%!)11%++##'% *!P%*2+%&++3v%&%)**2<#'%&%0%M(%+!&0*0%!&
%"#%&)*%&0!& %&#3
U'%%1& !1!/)1#,#!.% !&)*Y)0+&#'% &1V11*!)1+%+&
/#!&)%+ )**!) %#%)&+*-*21&$1#!&.
,)&Y)0!&$+#2!+.-!&.3Z)(&!"!% )% %!)%#%0( %+3w)&2,#'%1'% %+#'%#!1+%&+)#!&+
!+)#!.% !&)*+#!$*$+)&#'%1)(&)#%(%0%).% P&-&)+U)&2)+ /%1!)*,) !*2,%1%/#/#%!&+
($+!&%++[+)<(%%<+/)P*!&.-!&%+<%#13[) $&#+ &%*+DL!*0%%#)*3<\yyH1J3U'
+
!%&#x% %/#K#%&#!)*DUxKJ1')&V
#'$.%+)*%+-*-!%3K$##!&.)+!%Z\<-%1)& 1')&&%*+)&1&+!+#,,$)%+++%/ #%!&+, 1)#!&
1!)#%+$($&!#+3S)1'
)+P)($##'%%1& !1!/)1#,!&!0!$)*+/!1%+ +$($&!#1&#)!&+)*&./%/#!%-! #'+!"+%1#!&+#')#
#'%!$+%!&0)!$+/$1#+3]&#'%^&!#%L#)#%+<+V 1++#'%1%** % ()&%)&%)1'1 &#)!&+)+!&.*%
1)**%%#'&!1,+)%%"/%!%&1!&.)/%!,)/! /%%.!&3!.!&)**2!+10%%!&z
.-#'$%#)1&#!&$!&.!&Y$",! !.)#!&, #%1%/#+<)-!%0)!%#2,#'%+%,$&1#!+&)/*'1!'*)/ 'V
/%/*%+, 1$*#$%+-!#''#+/!121$!+!&%+)&) ')0%(%%&,$&!&0)!$+.)&+)& )&2!,,%%*+V )&&
.-!&.#%&#-)*%++&%/'(!1)& %)0%&V %&#1%**+DK)#)/$#!)&%#)*3<\yy{=|%&P)#)1')*) )&
#$$+!&!&.&#'%/)#, )&2C %!1)&+3L)*%+, w&#%**<\yyQJ3U'%M+#1'% %1%/#!0%UxK#(%
+)*+)')0%+$/)++%#'%+)*%+,P%#1'$/+!&1%FGG\3 1'))1#%!T%-)+#'%UxK|F<)+V1)**%0)&!**!
_%-/.) +,%+%)1'')0%)%-'*%&%-1)#V %1%/##')#!++%&+!#!0%#1)/+)!1!&)+-%**)+)1!!1
%.!%+,1'% %+#'%#!1Y)0)&#+<+$1')+`#!&.*%a /O<'%)#<)& %1')&!1)*+#!$*)#!&3&% % (%
1 /$&+3
012 345657845797

$!$!$! "$"!"$5&
!"#"$%&'" W% * (!!$!
($" %!!%" $ 6!%(# !
!!$" ! %" %+/!477.&)$$" !
%!!"$!&)$ "*" $$$!"$%"$"<!!$
!$!*!# ""% %$! % 6# "+X"47N-Q
"%#!! $%#%%"*( P* /472Q;6"47==.&
!!)"+&,--.& "$!5$!*#
""""!*/01 #!6 6 !Y"*"$
""!%" !*(! "+8%!<&472Q;6"47==.*"
%+2.&"$" ( !% ! $ %("
34" 2" % %!!!"$&
!"$!"$!"$ ;" " %("(<
# $" "$5# %"# !$+)!&
+!6""$$!"&.+!&,--7.&8"% 47=.!%##< "!
!($!%! #! """%! !$" (
""%%%!9%" !$!&Z# ! "$"
("%"&:!5$"$" !%"!" 6*$
""$ !!%$!( "$"+P*&472.&:#%!,&4R
(!%#""43" * 6 "9%" %!
"""%" $$!(( #$"$"% $$!
!%#"""+)( "!" 6""%!$$*
&,--.&:!!* $!"! "%"$$$!!$" ! ! !
"$"05&+,--2.;! $$$!!#%!(+!" ;
&+,--. 3<" +,--=.& 477N.&;6(! *!
$! %$!"!$
9%"Q!!"%(% #!
>?@?ABCDEFGHIJKDHLM +; P*47=Q/!477.&
"6 6<
!% $$$!* U"%
)"% "#! "" !!$!!9%!&)"
!! !$$$!"$%%" (! "!$$(% %*(
"$" !"$$$!$$! !("< "$(!# !9%! (*
$$$! ##!"$%%"6#!& !$ (!&
$" "$"*!*! O !%(# 6# "
(*4$$&(%4--$ !!%!!"<
$$! !!!( %! !# !S5("%#*
%!%";"$!"%!+"% #!# !S*&!!#"!
0$!N.&O$%! !"$""%*!! !$"(*!!# $$$!
(%!#$ !!*!$$! < !S* !
!+/! P*4774QP*47R.& %(Y"<%%*!#+P*
(%!"!"!" % 47R.&$!" !!
*$$$!"$%!## !#%( ![%(Y"\
%!&;%*"$"$$!!##! (%"!![(Y"\
!""!(! %!($#"!S5&
4-!#!+P*47R.&;S!$ )%! 9%! "
!*% $$" T$! "!! #!*" !!
U#+0$!.&V!!!%" 9%!< +/! P*4774.&
$!%"$! "!& U"%% %"< "(%!
00 1 2345678796
8 254
378796
87 6757574
67
!" #$$%&'%"'$'()*%""(#%'!+)$, #$$%&'%"%5'%"**)!$#)!'5'%)$$%-)2'%",'$

)-.!$"(/0"$)!$'#'.!'1$%-)2'%"345 #((!*$ *"--"2+$ 8%5*'0$'*,%5 !!%'%*"-0"%"(!#
%5 55%"(%5 *)!6$#-"$%!'%%5 *5!"*#$$%&'7 9*52:000!$345 2'%%! ((*%$-"! 0!"")*#("!%5"$
%"%5'%$*"!!2'%#8%555#%'!+%'1 "(0)%$0*$3 8%52"8#%'!+%'13;!"-<!$*"%%'#=%6$"9>??@:
45 $+-."2$'%%5 ('!!5%#-"$%!'% %5 8%57$$$" 8%50!-$$"3
#$*!.,'%2'$%A2$5,%5 /0!*$(!"- '#M)!05+,>?HIiB"- %%"7M)j&'#B',>?Hki
000!.)!,BCD,-)$%'!#,'#$""3A/0!* B"-- %%"7M)j&'#K !'#&,>??I:3B"-02%2+
8%5$0*$$)$%$%5'%%5! '! '6'!%+"(!!7 '"$-*#6#)'2$*'#%*%-'+"#"!*"-0")#$,
%'%6 E'6"! /0!*$'#"%'22!!%'%"$'! 0!$)-'.2+(!"-%5 '.2%+"("#"!'%$%"$%-)2'%
%5 $'- 3=%)#$"($+!$%*%!'*%"-/7 %5 %!-'2!6 .!'*5$%5 '$'2*'6%+9J"%+
%)!$'#0"%%'%"8%5#((!%!!%'%$!'0# %'23,>?lH:345!$'!!%'%6 *"-0"%%"-'+
$F)*$'! $)$%6 "(%5 0"$$.2%+"(-)2%7 *"--""#"!'%$'#E'6"!*"-0")#$3L6'!%+"(
02 !*0%"!- *5'$-$("!"!'2*5-*'2!!%'%" 552+!'*%6 $)2()!*"-0")#$5'6 . #%m#
9G'82$$'#=%6$,>?H?,>??I:3J!*%- '$)!7 "%5!!!%'%6 $0*$'#(""#E'6"!$,$)*5'$*"-7
- %"(F)'2%'%6 #((!*$8'$'%%-0%#' 0")#$(!"-5"!$!'#$5,-)$%'!#,'#%5 2'*!-'%"!+
#$*!0%6 $%)#+.+B2(('#K +-'9>??D:)$' 9%'!7#)*:('*%"!(!"-""$'#!2'%#6%'7
6'!%+"(!!%'%E'6"!-'%!'2$345+(")# 6#* .2$9n **)$,DII@:3A%5'"2'#*'-'2#5+#
("!#((!*$2'%- 9$5"!%6!$)$2""$%: '! "%5! /'-02$"("%5!*"--"E'6"!$%5'%'!
'#.)!6!$)$%2$$'%"$'-"%5 !!7 !!%'%6 9<!$*"%%'#=8'7B'-0.22,DIII:3
%'%$%$%#3L2/*"("!*'!."'%"8'$#62"0# B'!."'%","!%5 0!*0%""(#$$"26#BCD,
.+K'!0!'#M*J'29>??N:'#6"26##$*!07 6"26$'%!)2+-)2%-"#'2$%-)2)$3o'##%"%"%5
%"!$("!*""2,%'$%,%!-'29.%,.)!,)-.:, %'*%2 $%-)2'%""(- *5'"!*0%"!$,BCD'*%$"
'#%'*%2O- *5'"!*0%"0!"0!%$3 ."%5%!-'2!*0%"!$9J $$!! %'23,DIII:'#
)$%'%"!+!*0%"!$9B5'#!'$51'!%'23,DII?:3p"%5
PQRQSTUVWXYXXZU[UZ\W]W^_[UZ̀^_ "(%5$ *5-*'2$$'%"$6"26 %5 &+- *'!7
."*'5+#!'$,85*5*'*"6!%BCD%"*'!."*
[âbVWcZd[eb`èZ^f '*#3;"!%5 $$ "(%'$%,%5 $%-)2'%"8%5BCD
'00'!$%"6"26 %5 /%!'*22)2'!'5+#!'$ &+-
45 %!-'2E'6"!$$$'2$"'((*%(""#E'6"! '#%5 %!'$%!*0%"!0"%%'294n<:- *5'$-
"%5!8'+$3A6$)*5.$%-)2'$g'B2 9<JqDG>:"($")!!*0%"!*22$9B5'#!'$51'!%'23,
*'. !!%'%6 '%55*"*%!'%"$9h! '# DII?:345$$*"$$%%8%5%5 5'*- %"($")!
h 25'!#,>?H?:3B'!."#"/# $'0"%%!!%'% %'$% .+BCD'#$)00!$$""($8 %$$9B"8'!%,
%5 '$'2*'6%+,'$'!-'+"!'**"-0")#$9B' >??HiK 8$" %'23,DII?:345 !"2"("**0%"!$
012 345657845797
2
4)R)"(
!"#$ %%%&# /($R$((R
'())()$* )"S0)$,-B1&#4
+!" ,--%&.)/ /?((#4?
.0 # #1&#2 )($##$
(.)"2") )"/"(#T
2")3 $(// .)"$804+ ,-A-&?(U
/. #1$)( )".R($/(
2"))"#4. R(") ".
/(2( ")#V8)
"2"#522"2)467)) ""?DD/"
)($ ,--,&WU8)".)(
/))( W()$(("$
.#8)$" R/("R
$"($"2""2 )"#V7")"
/"2"/) /X"
5$,-- 93+!"$,--:9* )$")
+!"$,--,&#*+!",--,& D/"($,--1&#
)"<"""/) 4()
:;,%&#4).)") )"$"R(
""=>""$/= $"?
")2/)$" "( S"
"2"")"( "#$,--:&#
))# 4))(2"2
+"$) ("2)"
"("$() "2&$()(((
")"/2))"(? $("2"
$""$ "" "$,-AC9++
7""#$ %%@&#+" "#$,-A,&#=""(
/)")")"./#+" "#)))"
")"/".$ ((
/))"( "2+"#$,--1&#8
)" 5$,-AB$,-AC&# ))""(
4)"")".$ 6+!"$,-- 9
()"($/)($)$ 4)D/"$,--B&#4()
()$ 3D/""#$,--C90/"R
$)") 5/ Y"$,--A&(((R
3)$,--B$,--C&#8 "() ""3
"()2$) "("2)"#
/"(/ 4)"
)" D?(/""$ %%-96$ 767&"2/)R
%%C&# ())
" Z""RR"#$ %%,&#
S(""767/""R/
EFGFGHIJKLMNOMPQ [(2
R/"" 0$,--,&#8)
"2"(/
4)")" /)./"2 3
("(""# "#$ %% &#32")"22
01 2 34567898
79 3654898
798 78 68 685 78
!"#$$%&$%#$'$#"% FGHGIJKLMNNOPQMRLK
$%&%($%"#)*$%%#"%+
$!%$%&%#$%&"%$! 9%)S$(!)$%$%) $+
$%&%('$%,-%#$ ./0012* !# )$%$%
%($!*$%".3)$%#$'$##$!!+ )$
%)* %$%'$%)$"4"%$*$) ! $!"$%&#$!!%$%))
)$&))5#'"% %#)$&)$'(6* !%% *$)$%"".3*%
$',)'*$%&2.3)$%)%# %($$%$%%#&
#!$ $%#$"$'
$%"$%%%("'$#&7"+ ,/2%""" ,;2$%*$)!!%#
$%8!!&$% $%&% "%#. ?)%*)*)%%(, -$'$%#T(%;<<12.
9%)$"% $%#$'$##$!!%$%$%&% $%$% !)$ )
)
S$#$!!%%+
%$)&) )(
$%$$'(6*,:$) .;<<=2. @)#$@#7$8"
>%#$'$#*$))$&)6*%# 7%"8 ( )$. )"@)(
!!$%&% $$%.?"#$+ 3)+#$ !$%*$)!+
$%*$)$%&%@%%#@$#" !$%$(!%
!$!!))%#% %$&) .3)%$%$'$(@$)#$!)%$
$% #"$%"$$%#%$$5$%. "$%)#)#$%&$%&,U"V./00EWT*
:$&/.;A)*%$%$%$%&%("% ./00=/00A2.X%'$
"#$$%$&) )""%*$)$" )$!(%$%"@@($! )!+
$"!@'&)*$%.B) )"% $%(5$%&"$#$"$#C$#$%$% )
)%&!%$%!%%$%%%$%$%$+ )$%&*+V%*%"% #"%#
$%'@*%$"%# C% );+%+Y+%,T@%./00A2.
!%$%!)'#,D$%#.;<E12.
Z[\]^]_` 9'& $a$%%$('!$%&%($% @(defg%#fgghi"$'(.:D$%#.,;<E12@(

*$%*$)0A00&b!### %%$$#"% )+ "$$%!)9$%c$(!U%&(%#j$$.
$'$%&$%.c""V%#**$%&$%#$#
012 3456789
5
667

!"#$$#""%"%#" hijiklmnopmqrsrtuvqopumwoxtqn
#&&!"%'$#((&()*+,&#"%
"%((& (#&" "%-&# #./()0 &$#D$D&@#" &!"%()%(%)#$$"&B
1(2"%3&0"%!"#$$#"&# #"-&"%#"#""" "#"#&
&#!#"#$$)45#6$"#$*07889:*;-""%&" #&#"@&$2%"$)%)(B#"@4y&-2#
$"&#$(""#$ "%&"!"#$0#!#$$ c0]Â`0]Âd?<-&(%)"#$*0]^>>?<-&(%)#
-&&"&#"0!#2"%## $"&#$ =#0]Â8:*C%(#""& &-$""%#""")&#"B
"#""!-$#"4<=$-&#5#6$0788>?/"@ 2%6#D-"^8z#"@")*C%#"06% &#!
"#$*0788A:*,"%$#$$"&#"-&$"&2-"!B ##!(${-"#D-""%-!!#" 2-"#B
"&)06%%$"&#$"#""!-$#"- & "&)#$#"&)"")&#"2(&-2@&#$$
#2"""0"%"&!'!"#$$+"&(&"* &#"2 [#@&"&2"%0"%&$""$@ &
C%&)##$)"%-$D#&-$""2-% "&#"D"6"%"6!#$"*
D"6"%"6 #"*,"%#B y6@&0"%&%#@D!#)"%&"-%6B
"#D$%&&#!#"#$$)!%D)##$ 2%#!" (|"#"{-#$"0"#D$)
$-0"%"## &"&##$$#"&)B 6"0"%(& &*!(&B
#"0(D$)-"(""$(1#"(&-"* "#"")0(#$$)#!2-"&#-!&0
,"0"%&!#)D!"#$"%)"!$#2" "!#""&D-"!@$#"$$#"&)#"
!#$$$"&#$(""#$*C%&#$"%(D$") "'"#"0+(#&"-$#&$)&"&##$$)(&@&*
#"%& !"#$$#""%#"!#)D#"&- }"&##$!$$(&$)$#$3#"(&B
"#"0D-""%"$$"&@&#$* @##"#" &! "%&#$#@")*<-&(%)#
6&&4]^>>0]Â8:""%#""%&-!B
(-0"%)$D-")&#"#"&#$0"&D-""\-2B
EFGHIJKLMNOPQJRSTUOVWXTKSVQYKLOTU ! " '"#"+!#2"-*C%$$-$!B
#"D)(%2"%"&$%-"-&2"#"20
6%%(&%D""%&"&##$(##2 @$#"$
Z#!-$"B!#$1(&0"%-$! !#"&#$# "@$)-" "%@$#"$[#@&
#-&(&"%#""%#" &! &) !(&*
!#$")!#)[-\-2!"#(&(" "%&D&@#""%#"%#&%"#"#-(B
&!#"%&*C%&-2%-&1(&06$#&#D-" (&#($##""#"#%#&#"2@$#"$
"%(#&2 $&#"#"0$&0#& [#@&"")*b/)6"#$*4]^>`:1#!
#!"%#@1("#"#D-"6%#"#B &#"2 &"#"#&#""&D-" &-"\-
"!#)#!(#)#"%&*C%&-2%&(#" "%#"@#&#-&*}#"2 &($##"
(#&2&"%&-2%#"-&#$B-&& &B &#""&D-"&##"% &-($##"
""#"#[#@&0###"#DD-$"-( &#""&D-"&##-&"&#"
$#2""2&#" "%1(&4/"@ &#*~%#2%#(#"&#"
"#$*0]^^^:*_&#!#22 &2 "% &"#$ @$#"$6&""*b/)6 "#$*"&B
&"1-((&""%#"""%#""%!&22 "% (&""%#@ &#()%$2#$ "#
#""%&"(&("#&'&#$+(&(B (("#(%)#$"&#"* !$#& "
"#"\-"! &(D#4/!#$$ 6# - &D$#D&&)\-[#@&#"@#&)2$@B
"#$*0788`:*,"&#"D"6&)!#$B $ -&##")4a&2#<;#$0
"#"%&(D$-&#$-D"&#"%#@D ]Â^:*/-&B%# &-"[#@&&#"26%$
&@6D);$6%4788`:0/!#$$#a&"" \-6"% %2% # $@$%6 $6& &-"
4788`:0#b&%#2#c2$4788d:*C%B &#"2*
-"%#"$$66$$-"%"&#" %&"&##$!$$(&!""0#!!|B
"%#"%#@D!"%#@$)"-##&!" 2"%#"6"%#4;$6%0788`:
&$@#"" e"#"f&0[#@&f&&"#"4%!B #&#&%##6$$*C% "(
"%:0#$&f[#@&*g"%&"&B!#$")"&#B -( "%(|&f"#"(#&2*(#&"#!
"#&-"%&@6(#(&!" %# &-" &#2#"&#6D&&)$-B
#D@* "4-&%6 ":##!6%#"
01 2 34567898
79 3654898
798 78 68 685 78
!"" 'P$,##!3,,-./0R12K4)$)$#%,&
#$%&'()#!*+),-./0/123$## !)#))%,!)),##)8!#)B
$#!%&#$%&!-%%&4 )L,&!)""#$##)#$#4
%!'56!7&-./00123# #!!)%#$#,!25#!B
8!#$##!4!# )4!#!)""#,#44))"
4!9#)),)&"# 4!# )&$,23$##
!!2:)#6###%#&!# %&&,,,!#!$4,)##$)!
"!##,,),)6##-###$ !)#))###" #,,#'7) ,2-
#,,"&##,,"8)'3,,! .//=12E#!&-##&4L,))!)!
;#-<==>12:#4),!4,))&" #4##))8,)!!
!#!##$"!##!&,#% "!9#$##'P%%!*)-
)4)"),)))#""2 <==/12K,#,)#-8)#%J#6B
:!",,?4)4,)##8 ,&)'#&)144!9#)?B
$)4),%))###)42: ##!B!#$##';#
)#)@",!?4)#'38# ,2-<==R12:6-##,##$
,2-.//>12:4",!,)#& ),)##!,)&B#4)L))
!)$!$""####$# ,""#%#),,&2
#2A?4)$) K"#)!))##4##
B"#$##!,)C)!#- #%J#)#!6,#)@
"?4,-&!)8##$## 44#Q!))#'8!S,$!
""#2:)@"##))8,))##$ 56-.//N123$%&!##$##
%&"#$##)#4!)!%& "##Q#$%&#,)#'56,2-./0/1-
))),#$##)#"!#!4))#" ""))#"4!%&
",&,!#$)#$#$),,)! ()#!*+),'./0/1!,#)#,%,B
)#""';#-.///D38# )"8,),##)###)#)&#)#
,2-.//012 #%&* 4&,2'./TT12O$8-$#%J#
E#)#""#)4,&)@B )#!6,)#)&)#!
)"#$##)#!#9#,,"#)F 4)) ))4#-#))L
G8)!"H,)&I"""#" "#$##)##'56,2-.//=-
%#8)#$#,,)!#44## .//UDP$,##!A,6-.//<12V!Q#B
!###")))!#,)-J#,)6#$ -,!)&##,&%#"
#$,!'38#,2-.///12K%"%) #4##)")-!,&)#!##)
))#!)#8)!)L! ,)B!,, "#$##,,2
)8)&)%))###%)",? :)#L!)#%!)4,))#"$&#
'#3,,!;#-<==>DM!G,- )$)##&8,)#-4),,&!#)4B
<==N12:)##,!)#)#4,) )8,&##)$) ,)4,)%#"4,?
#)"L#$!)86)?4)B "!#!-#,!%!!2E,##B
"#9!4))')22-@81# ###))#%#)),)#"8)#
%!!)&'3,,!;#-<==>12 #&)#""#)$))#
+,,2'<===1#$!!)#,!# #)!"$)%#2W#4!#&
"!$!!$#%J#,!# #H!4I#"))4##)#)
)-%,&$#$#2 ##)%,&,&,,$%,#4#
O$8-)#!&-#$## )")%&4)8)#$)# 8),B
%&#%#,!!#$#%#8!'P%%! %,%,,'P$,##!A,6-.//<12K,,!
*)-<==/12 ###""",,#$B
:#))#$))#)#! ###,!%8)$!$)) ,###4#
$)))2E#!&-),Q##)?# %)##)))#,%,)#,,####%
$8,! #)%!)#,)!H)!)I ,!2:#""#!)###!,)
#,8,#!)8!" ),#4" A4/2
012 3456789
5
667

!"#$ '0()*](3/(32(/B^+3()*+()32371+*4(*00*4('2,-/;
(3('+>12(*2/1(>.*+/*H*9,9BU3++*+327V3+('/)-GB
%&''()*+,+'-./'012(*+34(1'2/5*(&**26'7381(1*/ KLL%W3O/9(*/4326'7-83(*'+36*81'+3(*4)1815-+29%)*+*
3+*16.'+(32(120''7/9:2*1/()*12(*+34(1'2'04)*6; 3+*0'8G+*6*71*/12=3+1'-/4-8(-+*/B/-4)3//(3+4)>
14381++1(3(1'2&1()<3='+/327()*/*4'273+**00*4(/12 4'+2B,)**B.12*3..8*B/-,3+B3275**+9E>/(*63(14/(-7;
<3='++3(12,/43-/*75>4)32,*/12=1/-383..*3+324*9 1*/'0(+>12,('&3/)'-(4)1815-+2&1()7100*+*2(
?2>'2*&)')3/4'6.3+*7<3(/'73('43+5'23(*7 (3/(12,+12/*/)3=*/)'&2/'6**00*4(0'+/&**(H6'/(
/'73&188+*4',21@*()3(()*(12,8*16.3+(*75>43+; .+'2'-24*7OB/'-+B327.*+)3.//38(HE1@*+327_3++1/B
5'271'A17*&18838(*+()*<3='+538324*123.+'7-4(B KLMWQE(*=*2/327X3&8*//BKLMÒ9I'87/(16-81.+';
-/-388>('1(/7*(+16*2(&)*2()*43+5'23(1'21/2'( =17*3(*6.'+3+>5-(.'(*2(12)151(1'2'0.*..*+5-+2B
.+*/*2(9C83(/'731/-/-388>(''/&**(9D*43+5'23(*7 3/G2'&2('632>)351(-a/'0*()214+*/(3-+32(/9E124*
4)36.3,2*1/-/-388>=*+>.''+&12*9
E*=*+38./>4)'.)>/1438/(-71*/)3=**A3612*7 4H438.3/+31F14*175
21/81.17/'8-58*B()*R27132+*6*7>'0,)**
-((*+O)3//'6*6*+1(9E'-+()12,//(16-;
12(*+34(1'2/'0(+1,*612381++1(3(1'20+'6 4)*61438/ 83(*/381=3+><
&1()(3/(*327&1()'7'+.*+4*.(1'29?/126'/(835; 35-/*7'+38(1/'/&B&) 14)63>.+'=17*/'6*+*81*0('
-*/9%)*4'65123(1'2'003((>B/'-+B
'+3('+>./>4)'.)>/14/B()*/*/(-71*/)3=*0'4-/*7'2 4'87B327/&**(/ -,,*/(/4)188*7>',-+(3/3,''7
.*+4*1=*712(*2/1(>4)32,*/12/12,8*4)*61438//16.8* 4)'14*9?4-8123+>.
61A(-+*/9%)*F+/(&'+G*+/('*A3612**00*4(/'04)*6; 4)-(2*>/&1())'(4-+++31*4/(1&' 4*'038(*+23(12,4''8B/&**(
-87/**6('034181(3(*()*/*
14381++1(3(1'2'2'8034(1'20'-276-(-3812)151(1'2'0 12(*+34(1'2/9
/6*885>43+5'271'A17*12()*2'/*HI312327J-+.)>B
KLMNO9%)1/'44-+/*=*2()'-,)()*'2/*('0()*/(12,
0+'643+5'271'A17*1/7*83>*7/'6*&)3(4'6.3+*7 bc#d"#$ef g
('()*'2/*('0/6*88/*2/3(1'2/9E124*632>/6*88/
38/')3=*321++1(3(1=*4'6.'2*2(HD'(>*(389BKLPMQ 8>B8*(-/4'2/17*+()**00*4(/'03..*3+324*'2
%-4G*+BKLPKOB1(1/.+'5358*()3(/'6*'0()1/12)151; C<132=3'8+.
(1'21/34'66'2*=*2(12*=*+>73><3='+.*+4*.(1'29 <3='+12*(+*4+*3.4((11''229%) *81(*+3(-+*4'24*+212,4'8'+h
R03.*+/'2)377*4+*3/*7/*2/1(1=1(>('23/381++1(3; +*/*3+4)*+/3+*4/1 /i-1(**A(*2/1=*32712(*+*/(*7
(1'2()*538324*'03+'63(14<3='+.*+4*.(1'261,)(5* (16*/4'2(+3714('+>H1'*92,*97(
3-( )3(1(1/4'6.8*A3273(
/)10(*71203='+'0()*'8034('+>4'6.'2*2(/9R01++1(3; j*63G*2'3((*6.()*+*('.=+1'23BX3 27X3&8*//BKLLMO9
=17*34'6.+*)*2/1=*
(1'21/+*7-4*7B()*2()*12)151('+>*00*4(/'01++1(3(1'2 +*=1*&9
&'-8738/'5*+*7-4*79
D'*/4)1815-+263/G(3/(*/12()*6'-()B()*&3> *(1*_- 632/3+*3=1/-388>7+1=*2/.*41*/9R2632>/'41;
/&1()63(-+*4-8123+>3+(/B()*=1/-38.+*/*2(3(1'2
()3(43+5'271'A17*/(12,63/G//6*8812()*2'/*S '030
T3+(13812)151(1'2'0(3/(*+*/.'2/*/)3/5**20'-27 34(*+1'/('171 4/
9
/3/16.'+(32(3/1(/<3='+327(*A(-+*4)3+;
?4'66'2F2712,1/()3(&)*20''7/3+*
0'88'&12,.+*(+*3(6*2('0'+38(1//-*/&1()43./31412B 6'+*7**.8>4 '8'+*7B()*>&188'5(312)1,)*++3(12,/0'+
.3+(14-83+8>12)151(1'2'0/'-+32751((*+(3/(*/HU3++*+ <3='+12(*2/1(>H
327V3+('/)-GBKLLWQX3&8*//327E(*=*2/BKLMYQ U3-(@BKLLNO9l00**9,49(B/' D-5'/**(389BKLMNQk*882*+327
X3&8*//*(389BKLMWQT+*/4'((*(389BKLLZQT+*/4'(( /1(>327<3='+17*2(1F403(41''823 '+*70''7/'2<3='+12(*2;
327E(*=*2/'2BKLLWB5-(/**38/'I'&3+(BKLMPO9 HKLLZO9J1/4'8'+*70''7/'+<3+*7 ='+
1/4-//*712E(188632
/3+*8*//*00*4(1=*8>
['(*()3(43./314127*/*2/1(1@3(1'2(3G*//*=*+38612; 17*2(1F*7HD-5'/**(389BKLMNO9_' &*=*+B()*.3((*+2
-(*/('7*=*8'.B19*9B1(7*.*27/-.'237*83>5*(&**2 '0+*/-8(/1/61A*7327124'2/1/(*2(1
(+*3(6*2(327(*/(/(16-81H\+**2BKLMLO9E-4)3(*6; H/**D*8&14)*BmNNYO9:24*3,312B8*2( )1/81(*+3(-+*
.'+38,3.&'-87)3=*'44-++*7('=3+>12,7*,+**/ (1'2/63>7+1=*()*.3((*+2/'012<-*24*92J' 3 +*73//'413;
12.+*(+*3(6*2(*A.*+16*2(/&1()(3/(32(/9?8/'B/124* HmNNKO0'-27()3(6'+*+*7&12*7*/4+1.('+++'/&* (*(389
43./3141212)151(1'21/6'/(+*81358>'5/*+=*70'+/-5; -/*75>3.32*8&)*23&)1(*&12*&3/12(*2(1'2388+>*
/(324*//'6*(16*/+*.'+(*73/.3+(1388>1++1(3(1=*B()* 61/4'8'+*7+*79
12)151('+>*00*4(/**212.+*(+*3(6*2(/(-71*/63> ?2*A36.8*'0=1/-3812<-*24*/'20''7.*+4*.;
5*7-*('7*/*2/1(1@3(1'2('321++1(3(1=*4'6.'2*2( (1'24325*0'-2712()*81(*+3(-+*'2.*+4*.(1'2'0
01 2 34567898
79 3654898
798 78 68 685 78
!"#$!#%!#&'#(%(%)%%%#*# defegeMNeQ
% +%%!##%+ ,*%
%%! -. , #) ((%!$%/01 S)!/h&i&/i#%!/'&b&!+i#%!/D&-jj?&'+3#!
/!+#%2#3%43#*%%5&6,%%/##*% *%#!+$+(!) #!#$*%$&i*%$
(%$%(#!#$!#%!#+%!) ((%!$% Sa%!%/%m/&kD&k/kl9@&
47!")!!+83!%/9:;<=>3/9:?;5&>!%+ ^%+%4%+9&5:/6 n9&]$# "%!%#$!+!&b!ò&'&
+%$(#%(!%#%+ +%%!##% ^%!/((&-<@l-@!;+&)a%! 7*"&a(!"%/
-. !#*%)((%!$%4$5#!"/ S%/m&D&!+p%#%/o&m&9:n;&>*%(%$B$!#
3#*%%/!+0&6,%%/,*%!3$3%% #D$ %#*!%`>*%B* (+&m3!i*%$
%%+/+$!#!%+(%+47*( S%/"m&D&
/-/<:l@;&
%#&/9::@5&A*%!#%#%+!#*%+,#*$+/ #)/3p#%+#%%*/o&m&!+7%/7&9:n;&a(%$B$!1
+%`>*%#( +&i*%$
+$!#! -. +( a%!%/-/9nl-@&
##$*!$%(%!$%/%3##*#! S'% %/m&D&/q%!#/8&!+ 8/S&r&9:;?&
+!%B!++B$3##,,&>*%%$* a3/-%;%/9
!#(%$B$!&7%$%(#3!+'#
<kl9;<&
%(*C%#*#*3!%$###*%%!%)%%!1 S!+%%!/r&h& /o3!+/m&7&!+73/D&9:n?&D!%(*!
#3#!)%+&D%!E)F%$1 !+3)#!$%7% %$#%#%+##"%!(!&i!+!
#%G+%$(#%(!%# )%3)F%$##3 S!+m%3!!/s&9
7*" !+7*$"/@;/-9;l---&
:n<&S"%)$+%!(%$%(#!&b!`D&
)& i&i#%%##%!+'&7&p%+!4%+&5/6!+)
7%$%(#!&bb&7$*(*$m3+"%!#!+'%3%%!#&
S$+%$/s%,t/((&-9@l-:?&
S+!/h&9:n:&o$#!!+#!!!3)%##%
HIJKLMNOPQRLMQ )3+!*3!3!"((%&a$!+&m3!8%!#
r%%$*/?n/<k@l<<-&
Sa>' 9:?:&a#!+++%B!#!#%%#!"#%!1
S!(#!#!,%+"%)%! %! (%1 Sa %3#!#%!+(+3$#&b!S!!3^
!!((%$#!#*%3!$#!#*%%!% '#%>'/7a*#!++%(+*&S%
/(&-&
$!a$%# >%#!"!+
#*3"*,*$*,%)#!3+#&T!+%#!+!"#*% S /s&!+o,%/6&>&9::-&u3##%!+_3!##1
(*"$($%%#*%%!%*%(3#% #%%3#!*"*1!#%!# ,%%#%!%!+,%%#%!%
!#$$3!##*%#%! 3!$#!!+*, ^$ 2#3%&i*%$a%!%/9n/-<@l-@:&
%!#!!#%$#&>*%*#$3!+%(!!!" "*% !/S&S&!+^%3$*(/v&h&-jjn&>#%%$%(#
%!%#*+%!#*%+$(!%($*(*$/ ^#*3&/o&'&9
! % S!!3r%%,s3##!/-n/k?:l<9<&
#*%#%#$#3+ %#!*()%#,%%!#1 i&i#%%##% !:+n?'&7 &6# ##%%%$*&b!`D&
&p%+! 4%+&5/6!+)
33!+%(!%&7$*(*$#*!!"/#*%!/ s% 7%$%(#!&bqS/>#!" !+ a%!"&S$+%$/
!#F3#)3#%#*+%! #%#!"/)3# ^#,t /((&-l9?&
%,%! 3!$#!#*##%#!*( ##*%"%!%#$):n#:&^
* 3/o&'&9 ##%##%$$*!%#%+
!"UVWXVYZ[\&>*3)%(!#%,#*# 7(#*3$&a$%!$%/-j@/#%# #
# *%)##%3)#!$%;1s1
:k<l:k@&
$!%!*!$%#*%$!#)3#!%! "3(# â#&/'
*3/o&'&/i!/A&a&/i%%!+/i&>&/v!/o&
#*%(+3$#+%%(%!#$%!#&]!%3!+31 a#!% /o&D&/a#%%!/m&i&!+a#,F/m&S&9:n<&
#$%"3%3%+#*(+3$#+%%(% m3!#* !+3 "3##%!++3 !#%&
!##%!+*(+3$###%#&S#B#"!$%3$* ^#*3/o&'&/83/$q!'%
%S% +$S$#!/-kj/;nj&
&^&!+'%/b&m&9::<&7>ic7r]7
##%%!#%%3#"%3&^3##*%% ,! #!+^% #!"`S!# /($*(*$!+%2%%$#&7*"
*/@;/99;@l99n9&
*!%2#%_3%#̀Ea%!+%UVWXVYZ[\##%#&G>* ^#*3/o&'& /'3(*/i&o&!+i%%!+/i&>&9:n?&
(($*+!#"%3#)!"+%%(%3!+%1 a,%%###%+
#!+!"#*%%#!*()%#,%%!!"%+%!# ;j:l;9k& 3#%si&7*" !+^%*/-9/
($%)%!+%! %(!%&b#%#*% ^#%a
!+7
#*/D&i&9:@<&S#"%!$ !+&p+7$%!"
$"!"/-k/9-<l9-n&
%! (%$#)%!+(%*(#*%#%#!" ^%+%/o&
!+!##*%% #*% )3+!"!++%1 a$%!#'& 9:;9&^(*$(($*%###%&S%$!
!"/!#*%,+%%%!"!%%!"#*!#*% ^!"*/S&p&//^
#/<: <-9l<k9&
$*/v&v&/+%v/i&/^%*!/m&'&!+
#$!(##%!(% %c!*(#*% 7 %!"/h&8&9::j&a%! #3+%,#*3$%1#
2#3%&i*%$a%!%/9@/<<nl<@;&
#%#!"&
0121314516 78
9

!"#"
!$!
% A(.E4) 2L"7FD6
-"D!
&'()*+
!,
-./,

!"/-(0"1
."/!498=
"-22!
%"
!0-3) A ",(B
,6JD9! +CK,B
3)4451-
!3!
67)
%"
36H/6 ;("N,A3==) 7

"2
225898 !
3!
!5G885988E
9,-.5:
!76"')7 A >77,>"3D%
-
@
26
!
"!"
%7;
-!"6 "

!"67

48
73!
!<3-=95= "
276A
!35E95=E
9
%>:83
" !2

A
D
:C"C)G)
2!7@

67>?2
6!%6-22@A6@ %%
"!

%2!2
(7363"

A7;B, =59E
9"A(*"C
"6/A- A
D
:O"C69:G-".DC?
!
22
""7 !"@!
" 736>D
-3)J3GG-
!3!
67

D
76!
"!!
%

,7" )
%"
HA!,-

2
"(7363"
8=E9= A
77"6J!
2
D6
7
9
-3"9!2:*43! 2%!6(7363"
449=
FD 622
%
;54EG5 A
77";G@2

!
"
2@
9
-3:
9!2:*":

%
L!!7
%!6&'9
!::":
9:56!26
!D
"!
%!6
C
"60"1&!
7."A2>D

!

A
!38=E984 C"22859G5
9"H3GH"D"'27!@ A
%O94827!
7!

!7".D
4988 %"72!2
M6!26
!
9!,C"-?/-D
%7
66 GG948
!""!2'-!
7"!%@ A@P
+(>"A
H38)2
%@

AGE4E94
72%!6A
!35E954
9-9-
%A3"3!
A@P
+(>""+3=K"
==6
%
!"6
!"
FD2@ "2%
7
?""
?""
!2
"6
!
"!" 6
%"
% !2
M6!26
!84595
2!."I
6"7!5849E=8 A"
;J
AA->"
9799
-3*A/"J/) 2A 9!!
==G"
!
72!
D"
%!6
J3!,(*
<@*J6J"67
"!" 62
!!2"&&'A

72!@
H==E) !
7
"
D
""
77!

"2

622
!

2."
26
!
""226
!

2!2@ I
6"7!4G9=4

A9
%6E5954 A9(4K!
!

'J
"!
9%"J/"*+
!,
()=A
!' 2!
D"

6QA
!38G4984
A
!"36JA/A9!/ A9() ""

7AK "

9,).9!,3"
3E!

R
6A
!35
7"2
7 ! 6 !2
!
."
5498=
!"
%88E59E58 A
JH3DJ-"C)8!
D"
A%//!%

7

677%
?"!
"
?

"
%
!
&'/A%"/*0"1 A
!3=49=
9
!
67)"K7!
-!"
!;B, JA!
! ;>/(
%
3J"
224E9=8 !./.==4) !
6"26
%67
A
H344J
77

D
67'#;
$ 9-D
(7."3!
!4/559/5
3!
!E0E. 614G94 JJ
;;%"9
-3===
A
H34),
'*6"
"
L@ ) %
%7'&%
7 "
!
3!
!=58G4984= ";;!
!585985
A
H3"J?483!2"D
7 J
!(==8)
2!72!2
!

"!!
",
%-7;B, 2!
D"FD."I
6"7!E
-!"673!
!5484985 5498G
A
H3">%)GK7!6"2
" J

(@3
A)A&KS6
!
%7"

6&'A7770"1 "A>===6!26
!"
@
K7!
")&&&/!,7<
D
6;B, %
!D
"!

!
" 6! @
22498 "
7!%
!
%
A
!3E
A
H3"26AC=&!
4498
!!2
D"

7""

; J(-"9!2:*48) !2!
6
8EE9E4
7!6
7
!2
M6!26
!
A
?(9*6A--">".
(==E GEE9EEG
A

72"!"
!D67 J
K,693C"NBGE-

!22
0)/1!7
6"
!2
7"26!26
!2
7!
!%6M) 2
!=G49=

-!6
%
!3!"
D
!G8G4948
01 1 2345678796
8 254
378796
87 6757574
67
!"#"$!%&'(%)*('(+,-*#(*. 1*((+41F:B(+"&'-)&%(*%&(+%$+ ,D%*'"
!(#$+$+")'%+/+01%$+2%+-3 %+-!&-"$&2%+-4("%G$*=:<HE;<=<
4*($#"&5(-67"(8)'%+9(+(,9'(&&9#*$+2(*4(*&$+ 1*((+41FTB(+"&$+"$M$"(#(*!(#$+,D%*'"
##:;<= "$&2%+-4("%G$*=FF==;F=F
4*)22(*3>?)*@A*+-*,,BC9+-(* 1*((+41F@%#%$!$+ (+$$X%$+%+--((+$$X%.
?%+-D*EF/+*%+%%&*$2('$+%&$')&%. $++"(+2)(#*-)!(-MM*$(,(S#)*( %&D
$+,*' -*) G&%$&(0!"'(*$!*(#+( ,*' !+!(+*%$+N()*!$(+!(((*IEE=;EF
%+'$!%+-+*'%&")'%+"$&2%+-4("%G$*HI 1*((+41H7"((,,(!,('#(*%)*(%+-!+!(+*%.
E:;F: $++"(#(*!($G(-$+(+$%+-K)%&$,!%*M+%$+
*%G+$(JCFHA-*K)%&$09('%+$!%&&2(+(*%(-')&. @"('$!%&9(+(E<=:;<:I
$-$'(+$+%&#*L&( %*( %M&(9!$(+!(HFE;FI 1*((+41%+-1(&"%*-4F9%&% %+*%&$**$%+
)M(@N@%*-(&&CO%+-B%&&(*AFI>,,(!, @"('$!%&9(+(<H:;HE
!&*%+ %+-P%G*%+ +$-(+$L!%$+#(*!($G(-P%G* 1*((+41%+-%D&( 877"(#!"#"$!
$+(+$%+-"(-+$!K)%&$,,*)$.P%G*(-M(G(*%2( %+- , '%(+*!"('*(!(#$+$+"(+(%+-')"
!%J(?)*+%&,Q-9!$(+!(<:==;=<: /+07O1(!"(&&B4%*")J%+-?4
>+2(+7FH7"((*!(#$+,A-*C!%-('$!N(DR*J 9+D5(-69'(&&%+-7%($+8(%&"%+-$(%(%G(+
>+2(+7%+-,%,,'%++@:CM&)(?)-2'(+ , N(DR*JNR##H=:;H:=
A-*/+(+$?)*+%&,>S#(*$'(+%&!"&2:F 1*((+41C&G%*(X.((G(B*%G$+1%+-CJ$*%G@
H=;HT HII:@"('("($%+-%(0>G$-(+!(,$+-(#(+-(+#*.
>+2(+7%+- 4E=+2(*' '('*,*-* !( $+2,(+%$+$+(+$"$&2%+-4("%G$*FT
D$"%+-D$")G(*M%&-(!*$#*?)*+%&,>S#(*$'(+%& :HT;:=E
!"&2IIHH;HHE 1*((*@C9*)!)*%&*2%+$X%$+,"(&,%!* .
>#J(>B!@&)*(97%+-%D&( 87HIIF>,,(!, ('/+07O1(!"(&&B4%*")J%+-?4
+%%&!!&)$+%+-*%&!+%!+#(*!(#$+,'(%&. 9+D5(-69'(&&%+-7%($+8(%&"%+-$(%(%G(+
&$!%(,*''(%&%&Q-U)%&$%+-*(,(*(+!(HI N(DR*J##T:;F
==;=E 1*)"J%B%+-9( &(4?4)*+$+2')" +-*'(
Q%*M'%+C/%+-8(&&(J%+1EFU)%+$%$G(%+%&( , /+07O1(!"(&&B4%*")J%+-?4
LM(*##)&%$+$+*%!"*-%'#%+$+(*G( %+-,)+2$,*' 9+D5(-69'(&&%+-7%($+8(%&"%+-$(%(%G(+
#%#$&&%(C'(*$!%+?)*+%&,C+%':=:I;:H N(DR*JNR##TT:;TFH
Q(!"+(*17TT5*%+&%$+*$2FTI6>&('(+ , 1)$+%*-?.[%+2M*+B%+-(D$B?FT7"(
!"#"$!>8C-&(*5*%+688D( %+->1 $'(!)*(,%*$+2(+!$+D$+()#+*(#(%(-$+2(.
4*$+25(-68&$+("%*%+-3$++N(DR*J $+C'(*$!%+?)*+%&,>+&2%+-O$$!)&)*(=E
Q$*(($+9HII8D"(&,%!* (''%J( (+( , F<;F
!(+N%)*(<=H;HF 1D%*+(>%+-8('%++8:7"(('#*%&#(*!(#$+
Q$"(*V4)&+4%+-NM&(C@<"$&2$!%& ,'(+"&?)*+%&,9(+*9)-$(I==;<II
,%!*!+*$M)$+2"(G%*$%M$&$,(+*% ( '(+0 1D%*+(>%+-8('%++8T*L&$+2-(!*$M("(
(&%$+"$#M(D((+ %&$G%*PD*%(%+-('#*%&#(*. (+*!"%*%!(*$$! ,% $'#&('-(&'(+"&&)$+
!(#$+,2)%* $')&$Q-U)%&$%+-*(,(*(+!( ?)*+%&,9(+*9)-$(=;<F
:::;T< 8%&&B4%*")JB@%$+39%+-9(G(+?@
QSC=H7"(*(&%$+"$#M(D((+!"('$!%&!+$)$+ E:7@%(M&$+-+( %+-"(%(,!%,,($+(N%)*(
%+-%(*!((-$+2 ,"(N%$+%&C!%-(',9!$(+!( H:=<<H;<<=
V9CF:;HI 8%&#(*+4HIIFB(!"%+$' %+- !+(K)(+!( ,
Q*%+JC%+-4*%'?FF7%(.'(&&$+(*%!$+ %*( *(*+%%&'(&&$+20@'#)%$+%&P)$--+%'$!M(*G%.
%%+%+--*%+-(#(+-(+@"('$!%&9(+(=<<: $+ %+-#!"#"$!%&M(*G%$+@"('(+(I
Q*%+JC)!"(+W%+-B$X(9?9F%9*%DM(** ;F
-*M)+*(-!&*(+"%+!( "( D((+( , )!*( 8%*#(*9?%+-B!%+$(&B=@%*M+%(-D%(*&(S.
&)$+@"('$!%&9(+(<=E $!+07('#(*%)*(%+-@AH &(G(&$+P)(+!(+-(!*$#$G(
Q*%+JCB$X(9?%+-@%*(*FMC+C ( '(+ *%$+2?)*+%&,Q-9!$(+!(:FF=;FF
,M$+%*'$S)*($+(*%!$+ ,*+$+( D(((+(*@"('$!%& 8(*+%+-(X9O%+-%D&( 87C'("-,%-\).
9(+(<TH;T=H '(+,*#*(,(*(+!(&(G(&,!%#%$!$+$+&$K)$-%+- &$-
Q*%+JCG%+-(*W&%%)DN?%+-9!"$,,(*($+8 ,- (' %'+2#%+(&$ ,D("+$!2*)#Q-
N?=4"#(*!(#)%&%+-!+!(#)%&,%!* $+P). U)%&$%+-*(,(*(+!(I<;<
(+!(%(.-*%+-%(.%($+(*%!$+(*!(#$+%+- 8(D +8&&D-7@"%+-*%9%+-8*?HII
!"#"$!:<=<=;=:< 1)%*&,%!*%+-*$2('$+%&$+(*%!$+ $+%'-(&
Q*%+JC3( (&N%+-9"%,,(*1I7"((+"%+!('(+ !%*M+%(-M(G(*%2(@"('(+*(*!(#$+H<;IE
, D((+( M *%DM(**-*$ $+*)!$+-(#(+-(+ 8MM(7T:BA%J("5(-6(G$%"%+0A*"(B%(*
@"('$!%&9(+(::ET;:EE Q*'(%+-D(*,%@''+D(%&">!!&($%$!%&%+-
1&-($+>49(+%$+Y(*!(#$+Q$,">-$$+ @$G$&@&&$(*4JN(DR*JNRTH(-$$+
4*JZ@&()M&$"$+2%!$L!1*G(@C 8*+(?8%(?%+-%D&( 87HIIH7)*M$-$%%'(%.
1&-($+>45(-6HII8%+-MJ,(*!(#$+4&%!JD(&& )*(,%&$G%*#*($+*(%!$+ D$"%*$+2(+)M%+!(
)M&$"(*/+!B%&-(+BC @"('$!%&9(+(HET:=;T:
0121314516 78
9
! 8@$9O$!&!$""
/.!
.!$
""

#
$#!"
""%&!$'
"%&!$'1)22
()*+ 8@$9(0
$#!"
.$
."
$#
&
9
+!
", "

# &M"
""> $$"
#".!

&!
&$$"""
-$$""./0&

$!"
'
!".727++)7+
.!
.!112)1 8@$9+?$-
3/%%"""
3& 4395 ,
"6 9
& ?&HI%
"-!
$
.0!&!'

" &97**+"!"

""$#&$. 172)1*
&&,581578#
&!"
"" 8@$9!"-$.
#!
& $M

/
!"

!
#"9"
$0!&.
#'!! ;$".&
$
$$
P?
E $".#!*

*172+)72( 172)117
:!
9+*;
$
#"-"&$"&!< 8@$9%$A%%7.!
$
!$,
&
#=&&"
.!"
> $$"
#$$?$$&
"&"".!$?
!
$
.+HI
9
"$> ##$
(21)(( )+*
:$"&0'
"@!AB+-$ "
8@$9 %
%:'&"!
#
$!"

#

C$:
$
#'
.'" "!.3/55 HIO$#!"
"
71)1+ Q3
!
#"" 3""
$'.&
&
5$$A"A'>AA:%$##
9D$$8
O$#!"
" 7(" ":
7**%
$"
,"@$-!
&
"

&<3 0
!"
#
'&$$"'!!'
A.

&"
#"!.?
E $".#!7 7)77

2)27 8@$9L$$""2'
.
"
5>"

A82##!"
#!!<
$!&!$"3/>$$%:
"F"

"" &.
$
.G>-
2( HI %!"F"
&"
# ?$-

7)7 %
&
0&!%&!$'
!".B"
%
5 :'(%
"
$#!"
.."&3/? 7()7
B8'$-HI9
,
$
.
#"'&$$ 8@$98%>!
&&
!&<
B$.9@4
A94 7*2)71 !$#
=-
3/0!;
5@& 45;(J&&/0>!" HI?$-
'!!',$!$!K
!$AA9@4
A 0&!%&!$'
!".B"
71)++
5&J<5:

%
98 ";!9 8@$9$

L:+%
&

#"<
.7**1
"
$!$

#" &K ""< !$/'"-".$!"$"-& &"
"-""$
! $."""-"."
.$"
!< :
$
#'
.'" 22)(
,&'!!777)772 8@$9'!$$!"!"!$<
5
F:90(A""

# @"<
#"")

&M" :
$
#?
'!!
""..
$
.>-
771()12 )+
5
F:90>"
A82>"" < 8@$9'"-0##!"
#
$!&!$

-$,.$"<"
"""& $"
& ""

"".
$
.>-
172)
.
$
.>-
12(()(1 8@$9'"-0;
,. &"
8,,"97**3& !"

#
-$
$#!"
."& $
$!&!$""# !"

#$
!
#"& $"

,"
$!
!""

" !-@" !"
G.!
.!1(7)(

# !
#"
!"-$%&!$'1 8@$9'"-0M"
#
$!&<
+2)+2 !$""3/L8B'%;8!>
8L8-&
>0?!LB >B0!HI!"

#%
& $M'&$$"
&
#&$$$&"!!".#
"#. 0!&!0 "$'.. 7()1*

&M" %&!$'+17 8@$9'"-0*##!,"@
8LB$$!
M1 !"

#!
&< "!"

#
$""/9
.
$
!$ !<

",.

&M" %&!$'( " $& $!"
3/>LL:;
HI
7)7+ %&!$3""
"9

"!$AA
8A9
;70,$"."
!&","@ 9@4
A94 ()7+
$"

&M" %&!$'(*1)2 8@$96@$$,1%$!$&"
#

8-:8@$9##!"
#!
$

""""
$#!"
!"

#".$-

N &"
#@"&
!$ $"?
0!&.
#O "$&
$
.O"
$.
$
.?$$1
E $".#!71)7 *)+
8@$9(( $"
#&M" "" 8@$99
:L+0"!.
#!

$#!"
'
.
!77()71( $""
9%&!$'71()*1
8@$9(-!#
$,"
8@$9;
F'A:2##!"
#
,""@"""&M" :
$
#%
& "-
$!!
""
.
$#!"
."""

.
$
!$.!
$
.121)2( =-
"C!"
&@
$$.
$.
8@$9*0!
&

###"&"
#
!
&!$ %&!$'*2()2
"-"."
"""
# .$"
!,& 8@$9
&%:%:
"
2
%%&!$'27()72+ D"

"
$#
$<!-
!
$
01 2 34567898
79 3654898
798 78 68 685 78
!"#!!$%&'(!) 6)+3'2'$3#+G'N'0.1C'M!)8
*+,-+./01' %!"&!#")"'N!%&%$
2 +3'4'+5+6'7'+2$8$+9'5'$7!+ G!<+00+10C/10>'
6'0..:';8%#<$%'7"8 6+Q'7'$G!"E+2'6'0..0'4##+
*%*"$+.+:1/=>' O"$$O"$#!#'QJ4'['Y!+
2 +3'4'+*<+?'5'+(!#)+@'A'$@"B+5' 9'2'?%+2'6'G!"E$7'G'* T$'U+*)$
,--='68)8""8$8) 43!$?'9<+P FE+##',-=/,CC'
)"'(!)*+C-+0D=/0.:' 6+Q'7'$*#&+@'6'0.D,'4O"$$O"
2 +3'4'+*!E+*'+*)%!+7'+2)+7'+F&+3'+ $<!#8!'(!)*+1+
(!#)+@'$G+G',--:'6$ ../0-D'
)'(!)*,.+,=/CC' 6+2'+2&+?'+M<+N'\'$5$$+R'0..C'
2+('G'$2 +3'4'0..0'4)H"8& N 8"$ O"88$O<)BB
'(!)*+0>+,,=/,CD' !O%"8X$"8I&%'Q
28I& +7'(',--.'( !")!$! 7"8R$*+C+,,1/,C1'
)!'28I& $5'!#JKK '8I& ' 6+Y'+G!+R'$?"O"$"+?',--0'4!8
)KL!L)!'!)< $'GV2&"&+1.+C-./C,-'
2!&+5'2'0.1='G!)%+*$M$'A! 6E B+3'9'0.10'4! $#8
N"O!+P FE' "&'5)7"8N%!&%+D:+CD1/:-='
2"O+6'G'+2 +3'4'+2<+M'$5+4'M' 6"%+P'7'+A)+6'N'+2%+4'3'$3)+
,--='Q$I8)H)&0HCH$ M'0..:'*"$%8!O<%#H
0HHCH8)"88""8'7"8 !#$#%#!O%03HP69##%'
5&""$R$(!)%+=C+DC,=/DC,1' M"#7"8G!)%+,0.+.,C/.C='
6+G'+3+7'+*+4'$G"E+2'G'0...' 6"#!%+('$(+A'*'0.D-'4$8J
()O # $ 8 $' (+ .>+ Q$#$<''N!%&%$G!<+,:+
10C/1,C' >-0/>-='
6+5'G'+5+4'M'$3!E+7'0.D1'M888 6"#!%+('+(+A'*'$G!"E+2'6'0.11'6""
&S"#E&&!)8&S"'N# 8$8'*%N+0+,-:/,00'
#$!0.:!5(*P6&+P ;+ P&$ !+4'A'0..='*<% R<'ME%
25+.KD1' 5+6"E+AQ'
6+5'G'+G$+7'+?&+9'G'$5+4'M' PS$+6'*'0.D>'4!"<8!&"#I
0..0'*%!!$JM)8) $H# #<8!#!))"
"<'7"8*%*"$+>T:U+,-=/,,=' 8!#'(!)*+00+,DC/,.C'
6+9'?'$2 +3'4'0.D='5$S") PO+5'('+5$+9'5'+G"!+7'+2!+M'7'+
#)B8%'5##+>+0-C/00:' *!)$+7';'$*+N'6'0.D1'6$I8
6G$+9'2'0.DC'57P?HK&%<& $$B$%)8 ))&%'5)
'N#VN%!#!%+C:+11/DC' 7"8M&%$[""+CDT,U+0:C/0:>'
6G$+9'2'0.D1'4#%!#!%$!G$ ;]6!%+6'$Q!+9'0.D>'^))#J
W"'(!)*+0,+C,C/CC,' Q)#8!$&)88"O'QJF'
6G"%+?'3'0.>>'6&"$)8!8 @ )"$6'9'@T$'U+^))J5G4'
$")!$8$#$")!$'7" 6?EE+P FE+PF+##'1=/.C'
8MX#)N%!&%+1,+D>./D1C' N&O+9'6'0.DD'9!#8#$
6G"%+?'3'$G!"E+2'6'0.1C'Q "$#88$O"'QJ7'*)+
O)" !$88W"'N!%&%$ ?'5'G!+9'6'N&O$;'9!$T$'U+
G!<+0-+00-0/00->' R$5#$P"'5$)+P FE+PF+
6G"%+?'3'$*!E+4'9'0.10'4$ ##'C=C/C1-'
8,>)#"$8)'N#VN%!#!%+00+ N&O+9'6'$G$$E+@'*'0.D.'5$O")#8H
,,D/,C,' 8!EO!'R$_"%$N8+
6("+*'4'$2 +3'4',--1'5)#8 0+:1/=,'
)"$<+)# N&O+9'6'$?"E%+A'2'0.>:'2O%#H
$B8'N!%&%$G!<+.,+ $"8<"&!%##8)E'?%
>=D/>>:' *5O+,>+==/>,'
66"+7'N'+?<+@'Y'+2%+4'3'$3)+M'0.D0' N#"+5'+N+5'6'+*%+Y'6'$[ !+
N%#!'7"8!(!)*%+ [',--C'4)49N!$O%$J6!)
(!)())"+C-./C00' 8)#"'P"9< KP"+:+
6&$+6'('+9$+?'*'$A%OE+@'5'0.D,' =,./=C.'
98$$8OZ<)&%%)' N+4'+Q!"S+F'$F#<!+Y',--1'6!J5
7"8!5)*%8G &(!)+:-+ 8!&E!)#"$'7"8!
00./0,D' 5)5$)%8?)&%+=1+D1C/D1D'
6)+3'2'0.10'M888##$" N+N'$N+5'6'0.D0'*#I)0+DH
%8"'N#$N%!#!%+0-+ J4!)#!#)%$'(!)*+>+
0=/0D' D1/.C'
0121314516 77
89
8 9 !"9#
.-8!9?7BB=/!F9 9,!

$%9 &! !&'!(%9
99)
"'!* %99!9 1>"!DG9! 9
:! 99!%88
**+** *C=+*=
89!,(--. 9
-/ 0 1
.-?-3ND K:8!! %$

2
329!!&4' 9! 3' ,!99'!199%
,& 9 9
.7BBCDG9! 999999'!
# '91)$9%

, 8!,!52! 9,! 199
%9%'2! ?'!

1
5'
-99 , '91)$9%
, (% , E9'!%&
,&77+B*
678 ;
3 5E2
95((! 1%9
8%
69-9 11-6:!2.-6;
9<$ 911919! < 9 9!
=$%9 >'9911!&
3%99!&! ,9&'!91!, (%9
997*
9! 9" &
!1
;1
3?'!
1. !& *C*+*C
997*+7 9"9.56;
9<$88' ,'%9!9#
8!9?)
"'! &%
,
!'# D1191!
%9
!! 89!9 J

1!9!9 " ,9'! ,%99!& 8&%%& **C8+*=B
@'
911) A'
& 8!919!99B 9"9.56;
9<$D%991
*C+*=8 !999@'9
!99 1!
%9
!!
8!9? 9"9?=D1191!
8%&
,&:9%" !*8*+8=
%9
!! 9>"! 1!9@'9 9"9.5:39!.('! D)!9&58',%.
1!9@'9'9!1%
8%&
,&:9%" != 6;
9<$7BB5 !9! 1%9919
9
+7 !
%9
'
(%9
99**CB=+C*
8!9?9"9?88&%%&
!99 9"9?(( O:'!39?! 2
&1
,
9'

9!99 1%9!
!!
1!&%!9%
(%9
99*8C*+8=*
0 ,9!9#9
% 1!9@'9&1%
' 9"9=/%9&%%&
;8&%
,
8%&
,&:9%" !8B8+87C 9" 9;8C=*+
8!9?; (29
E7BBB99!999 9"9=(!

&"
1'249 "9
!
!! 2&
9%&99%

91!
'9" 2! 9
9! %3?'!
1
8/89!9!(%9
997= DG9! 9
8&%
,&=7B+7B
7*+7C8 9"987$%9'!! ,
&19!&9!
8!9?5
99%96*9! 59! 8&%

, 7+*
29;99!
%9
!! 999!'!9 9"99
9!D<=
9
(%9
99*+CBC 9,!&
91!099!9'
'?'!
1
8!9??%9)! ?7BBC/!F9 DG9! 9
8&%
,&=C*+C
9! #D1191 119!99
!9, 9'! , 9"9?8!9?:395=$%9@'
9G'!9(%9
997**+*CB 19!9! 92&!69! ,- "
99 '97BB8)
"!(%9 !&$9%
,&$&
! 78C**+C==
)! : 9"9?:3958!9?(%,9
9!(D,9
< 5.3
9('!9 !;999!9'
,1!

9! ,1 '
?7BB9!&! 2'91

9G , "
9
9'!;999 #59G
91
9!9
!99 92&$8-=$8%9
?'!
1 &9%9 69! ,- " 7*+*7
E9'! 99778=C+7887 9"9?8!9?:395(1' ,
0 87H$99
1' I%9'
&1%9 99
#<;! >'99%99!9 1

1!&9989!9 J8&%%& **+CB ;99'!9(%9
997C87+8*=
'2 --. 9
-79!&9"
'
?5*(
! >'99>"! 9
1 2&1!99% 9!
,(%9
99 1!' >"!929"9!,9?'!
1) 99=
7*+7 B+7
% 119!9 <E?)! 49!?D$%99!9 !&9!6=: %9 !&)'!%D O<)!99
1%991K(
E(
@' 9<(
!9
9 E9;N!3EN
8/89 " &(%9
998LCM*B*+* ', -7BB89" 9;$99!9 ,
$8

(( ,19@'
%99! %9!&(9
'
!-
9'
!6 19 998*
&#$99%9
6.&6-:!%'3?: 7BBB+7B=

;L9M9
$9 <9
%. 99"9 0
& ?5%!
,
!%9
'
E9;N!3EN*C=+*8 1%9!
91 119!99!"929!!99!9 ,
929!KD1191!9
&%9
9! 9 1 ?'!
18%&
,&L8! M*
29"9!,9%092
0
& ?'!
1 7=+7=
5,! '
'!
)(%9 !&C*=*+*87 $ (Q%'(Q%'E7BB$85#$%99!

"9!O6985'3!
96 OP ,:
9'
91!%9
9 , %;&R$%9?'!
7BB$8%9
%99 (%999B 1E9'! 997L=M#B+B7
*+8 $%(?(6;
9<$=5! ,9'2@'

09!)<!! E=$%9 >'9911 9 (%9
997B=*+8BB
"999!'!9%!9%
9!9 1 $'39!.E
1!& !991! %9
(%9
99B7+78 " &#?2S/!,%9! ,9
&9#
01 2 34567898
79 3654898
798 78 68 685 78
!"#$%&''(')*%'+,-+''.+/0/! @+'?("OE$%$7?-$;1"UC2855H&=+=' ;
*1%."%"112324252 $%'=%'%.%=?$+=;%,'?%. ')=-
67'$"#2859*%'+1': ');( <=->$+%.)$= D$='%$H?$;+')*?%? [*H"52"B5884B8PN
?'%=%=&%;==@%?-$)=7%6?-%''."28" W$;$.7?-"*289Q6-+%.=?=$=))?=');'%''7;
ABB4AB3 .7=$;$= $% '7; 3\%'%$=E'7%$')F''
>$%C$$7<"DE$%F$%("GH2889*?$%.?';1'I *?%?"AN"BQA4BQ3
%%=%=%=')?';1J =;7K6-%L7%?')1'% W$;$.7?-"*$%C'&'"/288B#7;$% $%$11?$='%
$=%$=>M%>'%;%=/%=%$='%$"NN"N24A2 ')7;$;=$=S)$?='%$%6$=]//%KCC7-$$
0%($=$?-$$;"C$%'%="ONPPQ6G,?-$%%H%%7$ !",'?%. ')=-22=-/%=%$='%$*+;1'7;'%
G>< %'?-;=+"Q9"A5Q4B2B S)$?='%$%6$=$%NQ=-=%."E$1$%H '?$='%
0-$.%"E0$% M%.%"NPP96-%7'?'.%I )'*;$%6$=*?%?*1%."6'(+'"11A3A4A39
=>&$ ')-7;$%)''1?1='%K*%'+%=.$='% W$7"DED$%?V$%"G288P6-1'<)7%?='%')
D7'?%?$%'&-$>'$G>< "AP"92A493P ?$&'%$='%E'7%$')*%'+*=7"3"22Q42N5
>'%*+'<"M"'('<=R"#"E$?'&"#$%;$%"# W$7"DED$%?V$%"G28826-))?=')=;I
28QBS'I=$=%=$?='% %)7=T7?&%;== 1$=7'%?$&'%$='%1?1='%O-;?$*%"29"
@%?-$)=7%6?-%''."Q"254NP AAQ4AB5
@))%&$?-"E2882O-;?$% %$.%./%K60 W'(';7($"W"O'<$="E$%$7?-$;1"UC288A
U=?-"GV'=+"$='-7($%E*%'<!" /%>7$))%? % %=>=+='&==I=$=%. 7&I
*;$%6$=%#$=-$%V$G$>%"D<W'(" =$%?O-;?$*%"25"9984952
11A984AQ5 X$R'"$%*=$%='%"V6NPP9/%=:?$='%')$=%=>$I
@-=-$"O";$%"O*$%C%$"H2853 $& %$ ;$%=?''1':$=$&$7%.1%?1$
V=&7='%')=$=$%.%$ %'+%>%%. % ?';1'%%=$%$+O-;?$*%"A2"Q2A4QNB
)7%.)'; 1$1$')=--$; =H;?$%E'7%$') X%"VH$%C$7=R"H288PO''$))?=1?>
H%$=';+"2QA"2534NP2 ''%=%=+E'7%$')MJ1;%=$,+?-''.+K#7;$%
@;$%"EE$%?V$%"G2858':?$='%') ,?1='%$%,)';$%?"29"A824A8Q
)7=L$>' &+$1$=$;$% 7?'/%==7=')F'' X-$'"#"/>?""S=$("E"#$-;'='""('-&$"
6?-%''.="H%%7$=%.H&=$?="O-?$.'"/ C$%F=%"*2885F7%?='%$J1 '%')$;$;I
0?("X2882*'7%1?1='%$%)''Y7$=+E'7%$') ;$$%''$%=?1='*?%?"NQ8"NAQ4NBN
F''Z7$=+"2B"5Q489
Chapter 17
Quality Control and Shelf-Life (Stability) Testing
Abstract Two routine functions of a sensory department may be quality control

testing and the measurement of product stability or shelf life. These activities may
involve any of the three main kinds of sensory testing or modifications of them.
However, there are unique constraints for these tests, different types of analyses,
and specific models for these data. This chapter discusses different procedures for
sensory quality control, presents a recommended procedure, and outlines the pro-
grammatic requirements for establishing and maintaining a sensory QC function. The
second section of the chapter presents an introduction to shelf-life testing, its special
considerations, and some of the models used for stability testing data.
Consumer researchers are well aware of the quality of products. The food industry constantly faces
the demand to maintain both quality and profitability simultaneously. Quality, however, is an
elusive concept and as such must be operationalized and measured in order for it to be maintained.
—H. R. Moskowitz (1995)
Contents 17.7.1 Desired Features of a Sensory

QC System . . . . . . . . . . . . . 422
17.7.2 Program Development and Management
17.1 Introduction: Objectives and Challenges . . . 408 Issues . . . . . . . . . . . . . . . 423
17.2 A Quick Look at Traditional Quality 17.7.3 The Problem of Low Incidence . . . . 424
Control . . . . . . . . . . . . . . . . . . 409 17.8 Shelf-Life Testing . . . . . . . . . . . . . . 424
17.3 Methods for Sensory QC . . . . . . . . . . 409 17.8.1 Basic Considerations . . . . . . . . 424
17.3.1 Cuttings: A Bad Example . . . . . . 409 17.8.2 Cutoff Point . . . . . . . . . . . . 426
17.3.2 In–Out (Pass/Fail) System . . . . . . 410 17.8.3 Test Designs . . . . . . . . . . . . 426
17.3.3 Difference from Control Ratings . . . 411 17.8.4 Survival Analysis and Hazard
17.3.4 Quality Ratings with Diagnostics . . . 412 Functions . . . . . . . . . . . . . . 427
17.3.5 Descriptive Analysis . . . . . . . . . 413 17.8.5 Accelerated Storage . . . . . . . . . 428
17.3.6 A Hybrid Approach: Quality Ratings 17.9 Summary and Conclusions . . . . . . . . . 428
with Diagnostics . . . . . . . . . . 414 Appendix 1: Sample Screening Tests for Sensory
17.3.7 The Multiple Standards Quality Judges . . . . . . . . . . . . . . . 429
Difference Test . . . . . . . . . . . 414 Appendix 2: Survival/Failure Estimates from a Series
17.4 Recommended Procedure: Difference Scoring of Batches with Known Failure Times . . . . 429
with Key Attribute Scales . . . . . . . . . . 415 Appendix 3: Arrhenius Equation and Q10
17.5 The Importance of Good Practice . . . . . . 417 Modeling . . . . . . . . . . . . . . . . . . 430
17.6 Historical Footnote: Expert Judges References . . . . . . . . . . . . . . . . . . . . 431
and Quality Scoring . . . . . . . . . . . . 419
17.6.1 Standardized Commodities . . . . . . 419
17.6.2 Example 1: Dairy Product Judging . . 419
17.6.3 Example 2: Wine Scoring . . . . . . 420
17.7 Program Requirements and Program
Development . . . . . . . . . . . . . . . . 422
H.T. Lawless, H. Heymann, Sensory Evaluation of Food, Food Science Text Series, 407
DOI 10.1007/978-1-4419-6488-5_17, © Springer Science+Business Media, LLC 2010
408 17 Quality Control and Shelf-Life (Stability) Testing
17.1 Introduction: Objectives tests (Reece, 1979). Such constraints and demands
and Challenges often entail compromises in sensory practices.
A basic requirement of any sensory QC system is
the definition of standards or tolerance limits on a sen-
Product quality has been defined in a variety of dif- sory basis for the product. This requires calibration
ferent ways (Lawless, 1995). Most sensory researchers studies. If the sensory QC program is new, manage-
focus on issues of consumer satisfaction as a measure ment may be surprised to learn that some research
of quality (Cardello, 1995; Moskowitz, 1995) although needs to be done before the QC panel can be trained
there is an historic tradition of using expert judges, and begin to operate. Sometimes the identification of
commodity graders, or government inspectors to be the standard products and tolerance limits may incur more
arbiters of product quality (Bodyfelt et al., 1988; York, expense than the sensory panel operation itself, espe-
1995). This tradition is tied to use of the senses for cially if consumers are used to define the limits of
detection of well-known defects or expected problem what is acceptable quality. Maintaining reference stan-
areas. The approach was well suited to standard com- dards for a standard quality product may also present
modities where minimum levels of quality could be difficulties. Foods and consumer products may have
insured, but excellence was rarely the issue. Another short shelf lives, and even with optimal storage con-
strong tradition has been the emphasis on conformance ditions the standards will need to be replaced sooner or
to specifications (Muñoz et al., 1992). This approach is later. It is difficult to prevent some drift in the product
useful in the manufacturing of durable goods whose over time. Multiple references including both opti-
attributes and performance could be measured using mally stored and fresh products may be needed (Wolfe,
instrumental or objective means. Another popular def- 1979). Some products simply change with age and this
inition of quality has been fitness for use (Lawless, is a desirable feature like the proteolysis in ham or
1995). This definition recognizes that quality does not in cheese ripening (Dethmers, 1979). Furthermore, the
exist in a vacuum, but only in a context or frame of frame of reference of the panel can drift or change sea-
reference for the consumer. Finally, the reliability or sonally. This makes it difficult to insure that a sensory
consistency in sensory and performance experiences specification of a standard product is in fact the same
with a product has been recognized as an impor- as the last standard.
tant feature of product quality. Consumer expectations Other barriers to acceptance involve the different
arise out of experience, and maintaining the con- ways that sensory evaluation is performed as opposed
stancy of that experience does a lot to build consumer to traditional quality control. Most sensory tests are
confidence. designed to look at a few or limited number of prod-
There are a number of challenges and problems ucts. Sometimes the products are even considered to
that face a sensory evaluation program when trying to be identical, as in a homogeneous product evaluated
provide sensory information for quality control (QC). from the same batch, like a well-mixed tank-produced
Difficult situations occur in the manufacturing envi- beverage. The major source of variability in this sen-
ronment where sensory assessment is needed during sory test is in the measuring instruments, the panelists.
the processing itself. Such online sensory quality test- Statistical tests are designed to look at mean scores
ing is likely to be done under tight time constraints, against the background of variation among people.
for example, while the product is cooling and before This is quite different from the usual operation of qual-
a decision is made to bottle or pack a production ity control, where many samples of the product are
run. Only a few qualified judges may be available on taken and measured only once or a very few times on
third shift in the middle of the night when these deci- an instrument. The variability measured by traditional
sions have to be made. There is little luxury involved QC and pictured in control charts and other plots is
in terms of time, and a detailed descriptive evalua- across products, not instruments. Sensory QC has to
tion and statistical analysis may not be possible due deal with both sources of variation. In the instrumen-
to time and resource constraints. At the same time, tal measures, one can sample hundreds of products
a flexible and comprehensive system may be desired, and take one measurement on each. In sensory QC,
one that is also applicable to raw materials testing, there may be one sample of each product but multiple
finished products, packaging materials, and shelf-life measurements across panelists.
17.3 Methods for Sensory QC 409
17.2 A Quick Look at Traditional Quality is necessary. If action levels are surpassed, then there
Control is good evidence for an out-of-control situation or
assignable cause, and the process must be changed.
The X-bar chart plots the mean scores for different
Traditional quality control involves three major test batches over some time period. Typically three to
requirements: the establishment of specifications, the five products are pulled for evaluation (Muñoz et al.,
establishment of tolerance limits, and a sampling plan 1992). Upper and lower confidence limits (UCL and
appropriate to the product being manufactured or the LCL) are generally set by ± 3 standard errors (some-
system being monitored. By specification, we mean times referred to as 3-sigma) as shown in Fig. 17.1.
the characteristics of the ideal or average product. Out-of-control conditions are spotted by a number of
To set tolerance limits, the liabilities of Type I error, criteria, such as a point beyond the 3-sigma UCL or
rejecting product that is acceptable and therefore incur- LCL, or some pattern of points such as “nine points in
ring unnecessary cost, must be weighed against the a row all on one side of the historical mean” or “six
potential for Type II error, letting bad product into points in a row all increasing or decreasing” (Nelson,
the marketplace and offending loyal consumers. This 1984). The R-chart measures the range of observations
is a management decision that will impact the nature in any batch of product. Upper and lower limits are set
of the tolerance limits that are set. These tolerance as in the X-bar charts with warning and action levels at
limits are the levels of variability and/or the ranges 2-sigma and 3-sigma (or 95 and 99% confidence lev-
that are deemed acceptable (in specification) versus els). Sometimes the X-bar chart and the R chart may
unacceptable (out of specification). be combined to give a fuller picture. If only one unit
Historically, such analysis was born in the advent is evaluated per batch, then the mean and range cannot
of statistical quality control. W.A. Shewhart, work- be used, only the observed value itself. This is plotted
ing with Bell Labs in the early 1900s, noticed that on an I chart (Muñoz et al., 1992). As in the X and R
there was variability in the functioning of some sig- charts, a mean and confidence limits can be set, as well
nal transmission components and that since these were as warning levels and action levels.
often buried, they were a problem to dig up and repair.
Failures or severe problems needed to be differentiated
from the normal expected variability in these systems.
To address this, he coined the ideas of assignable cause 17.3 Methods for Sensory QC
versus chance cause variation. The idea was that some
variation was to be expected, but when the observation 17.3.1 Cuttings: A Bad Example
was outside of some common range, it was likely that
some other cause was at work, and the item would need Muñoz and coauthors (1992) give both good and bad
to be replaced, repaired, or otherwise dealt with. Thus examples of applications of sensory QC procedures.
the approach was statistical and involved a number of Here is an example of a poor implementation of the
charts or graphs depicting this variation and the limits in/out procedure:
at which an assignable cause might be suspected. His
notions of statistical quality control were later adopted The panel consists of a small group of company employ-
ees (4 to 5) mainly from the management team. The panel
by W.E. Deming in support of the war effort and later
evaluates a large amount of production samples (up to
in the reconstruction of Japanese industrial practices. 20 to 40) per session without standardized and controlled
There are several common types of charts used protocols. Each product is discussed to determine if it is
in statistical quality control and the sensory evalua- to be considered “in” or “out” of specifications. In this
program, no defined specifications or guidelines for prod-
tion specialist should be familiar with them as they
uct evaluation exist, and no training or product orientation
are part of the common language used by traditional was held. As a result, each panelist makes decisions based
quality control departments. Three kinds of control on his or her individual experience and familiarity with
charts are common: X-bar charts, R charts, and I charts. production, or based on the highest ranking person on the
panel. (p. 141)
Various rules exist for warning levels and action lev-
els using these charts. Warning levels generally mean This scenario highlights some of the pitfalls of
that the process needs to be investigated but no change a pass/fail procedure. It resembles a common daily
Fig. 17.1 X-bar and R charts X-bar Chart for Thickness

for a hypothetical analysis of
product thickness ratings over
several batches. The X-bar 14
UCL = 13.72
Mean Thickness Rating

chart shows the historic mean
12
and upper and lower control
limits, usually set at three
10 X = 9.87
standard deviations. The
X-bar chart shows one batch 8
with a mean value below the
lower limits and the batches 6
on either side show a range LCL = 5.95
beyond the range limit as 4
well. This should suggest
action and/or investigation by
process control personnel. 2 4 6 8 10 12 14 16 18
Batch 13 was also above the
R Chart for Thickness
range limit suggesting an
out-of-control situation.
Batches 5–10 also show the 5.0
alarm pattern of six points on UCL = 4.72
one side of the historic mean. 4.0
Range of Ratings
3.0
R = 2.85
2.0
1.0
0.0
2 4 6 8 10 12 14 16 18
Batch Number
check on production that was often done by a convened scenarios such as Canadian fish inspection (York,
committee of technical personnel and managers, called 1995).
“cuttings.” Without the guidance of a sensory evalua- Panelists are trained to recognize the characteris-
tion specialist, such a method can be put in place with tics that defined “out-of-spec” products as well as the
a number of poor practices, such as having an open range of characteristics that are considered “in spec”
discussion to reach consensus and determining a final (Nakayama and Wessman, 1979). This enhances the
score. uniformity of criteria among the panelists. As in any
yes/no procedure, the effects of bias and criterion set-
ting can be as influential as the actual sensory experi-
17.3.2 In–Out (Pass/Fail) System ence (see Section 5.8). Different panelists may be more
or less conservative in the degree of sensory difference
Muñoz et al. (1992) discuss four different approaches they require in order to call something out of spec. In
to sensory quality assessment. Their book, Sensory quality control, the liability of differences in criterion
Evaluation in Quality Control, gives a detailed treat- setting is high, since there are always pressures to pass
ment of each. One of the methods is the in/out or pass– poor products to maintain productivity. Obviously, the
fail procedure. This method differentiates normal pro- presentation of blind control samples is necessary to
duction from products that are considered different or estimate a false alarm rate (false positives), and the
outside specifications. It is a popular procedure at the introduction of purposely defective samples can be
plant level and is used in some binary decision-making useful in estimating the false-negative (miss) rate.
Muñoz and coauthors stressed the need for standard- can be used. Additional points along the scale are
ized protocols for sample handling and evaluation and sometimes labeled with other verbal descriptions of
the need for independent judgments, rather than dis- different degrees of difference.
cussion and consensus. York (1995) described how Training with a range of references and establish-
government fish inspectors are involved in standards ing the nature and conditions for reproducing the
development workshops. Training includes definition control sample are critical in this procedure. The pan-
of sensory characteristics that define wholesomeness, elists must be shown samples in training that represent
taint, and decomposition and how these characteristics points along the scale. These can be cross-referenced
at different levels contribute to the binary decision for to consumer opinion or chosen by management tast-
acceptance or rejection. ings (Muñoz et al., 1992). Preferably there is some
The major advantages of the in/out procedure are its consumer input for calibration at an early stage of
apparent simplicity and use as a decision-making tool. the program development. Muñoz and coauthors also
It is especially suited to simple products or those with presented a more descriptive version where differ-
a few variable attributes. The disadvantages include ences from control on several individual attributes
the criterion-setting problems described above. Also, of a flaked breakfast cereal were evaluated. This
the method does not necessarily provide diagnostic more detailed procedure can provide more action-
reasons for rejection or failure, so there is a lack of able information about the attributes responsible for
direction to be used in fixing problems. It may also any differences. If just a single scale is used, pan-
be difficult to relate these data to other measures such elists may weight attributes differently in determining
as microbial or instrumental analyses of food quality. an overall degree of difference. Specific character-
The data necessarily consist of mere frequency counts istics may be more or less influential for a given
of the number of panelists judging the product out of panelist.
spec. Finally, it may also be difficult for some panelists Management should choose some level of the differ-
to be analytical and look for specific problems and ence as a cutoff for action. The scale is useful in that
defects, while at the same time providing an overall it provides for a range of differences that are accept-
integrated judgment of product quality. able. At some point, regular users of the product will
notice and object to differences, and this should be
the benchmark for action standards. If at all possi-
17.3.3 Difference from Control Ratings ble, the panelists should not be informed of where the
breakpoint in decision making occurs along the scale.
A second major approach to sensory quality control is If they know where management sets the cutoff, they
to use ratings for an overall degree of difference from a may become too cautious and tend to give scores that
standard or control product. This works well if it is fea- may approach but not surpass the cutoff (Rutenbeck,
sible to maintain a constant “gold standard” product for 1985).
comparison (Muñoz et al., 1992). It is also well suited As with the pass/fail method, an important part of
to products where there is a single sensory character- this procedure is the introduction of blind control sam-
istic or just a few sensory characteristics that vary in ples. During every test session, a blind-labeled sample
production. The procedure uses a single scale as illus- of the standard should be inserted into the test set,
trated in the paper by Aust et al. (1985) such as the to be compared against the labeled version of itself.
following: This can help establish the baseline of variation on
the scale, since two products are rarely rated as identi-
_________________________________________________________
cal. Another way to think about this is that it provides
| |
a false alarm rate or an estimate of a placebo effect
extremely different the same as (Muñoz et al., 1992). In the original paper by Aust
from the standard the standard et al. (1985), an additional control sample was a prod-
uct from a different batch of the same production. Thus
Ratings on this scale may be transcribed from zero a test product’s variability could be measured against
(rightmost point) to ten (leftmost point). For purposes the response bias or variation within the ratings of the
of rapid analysis, a simple 10-point category scale standard against itself, as well as the batch-to-batch
variation. This approach could also be useful in com- or mishandling will go through a sequence of flavor
paring products from different manufacturing sites. changes and sensory spoilage characteristics. These
Aust et al. proposed an analysis of variance model for changes in sensory characteristics can be translated
this design. If the control comparison is simply the into a scale for fish quality (Regenstein, 1983).
standard against itself, a paired or dependent t-test can The common characteristics of quality ratings are
compare the mean scores for the test product against these (Muñoz et al., 1992): Scales directly represent
the mean score of the standard product rated against the quality judgment, rather than just sensory differ-
itself. This presumes that there are sufficient judges to ence, and can use words like poor to excellent. This
warrant a statistical test. With small panels, more qual- wording itself can be a motivator, as it gives the
itative criteria for action have to be adopted, e.g., any impression to panelists that they are directly involved
three out of five panelists below some cutoff score. The in decision making. Quality grading works best when
false alarm level should also be considered in making there is management or industry consensus on what
decisions. The ratings for the blind standard against is good. In some cases specific product characteristics
itself must be low relative to the test product scores can be rated in addition to overall quality, for exam-
in order to reject a batch. ple, quality of texture, flavor, appearance. In some
The major disadvantage of this test, like the yes/no schemes like wine judging, quality scores for individ-
procedure, is that it does not necessarily provide any ual attributes are then summed to give the overall score
diagnostic information on the reasons for the dif- (Amerine and Roessler, 1981). Unfortunately, the qual-
ference if only the single scale is used. Of course, ity scoring approach is prone to abuse, where small
open-ended reasons for difference can be given, or numbers of poorly trained judges evaluate dozens of
additional questions, scales, or checklists can be pro- product “cuttings,” use their own personal criteria, and
vided for attributes that are common problems or show use consensus (discussion) methods to make decisions.
common variation. Muñoz and coauthors presented this example of good
practice:
The panel consists of 8 to 12 panelists who are trained
in the procedures to assess the quality of a given prod-
17.3.4 Quality Ratings with Diagnostics uct type. They learned the company’s quality guidelines,
which were established using the input from consumers
A third method, similar to the overall difference from and management. . . . These guidelines are shown to
panelists by actual products representing various quality
control method, is to use quality ratings. This entails levels. The program was designed by a sensory profes-
an even more complex judgment procedure on the part sional using sound methodology and adequate testing
of the panelists, since it is not only the differences that controls for the evaluation process. In routine evalua-
matter but also how they are weighted in determining tions, panelists rate “overall” quality as well as the quality
of selected attributes using a balanced scale (very poor
product quality. This idea of an integrated quality score to excellent). The data are treated like interval data and
is part of the tradition of food commodity judging, as panel means are used to summarize the results of the eval-
discussed in Section 17.6. uations. The results are provided to management, which
There are three main abilities of the trained or expert makes decisions on the disposition of the production
batches evaluated. (p. 109)
judge that are necessary in order to use a quality judg-
ing system. The expert judge must maintain a mental Although there are apparent time and cost advan-
standard of what the ideal product is in terms of sen- tages in this direct approach, there are also disad-
sory characteristics. Second, the judge must learn to vantages. The ability to recognize all the defects and
anticipate and identify common defects that arise as a integrate them into a quality score may require a
function of poor ingredients, poor handling or produc- lengthy training process. There is a liability that indi-
tion practices, microbial problems, storage abuse, and vidual subjectivity in likes and dislikes can creep into
so on. Finally, the judge needs to know the weight or the judge’s evaluations. The specialized vocabulary of
influence of each defect at different levels of severity technical defects may seem arcane to non-technical
and how they detract from overall quality. This usu- managers. Finally, with small panels, statistical differ-
ally takes the form of a point deduction scheme. In the ence tests are rarely applied to such data, so the method
case of seafood, deterioration as a function of aging is primarily qualitative.
17.3.5 Descriptive Analysis Descriptive analysis requires extensive panel train-

ing. Panelists should be shown reference standards to
learn the meaning of the key attributes. Next they must
A fourth approach to sensory quality control is
be shown intensity standards to anchor their quantita-
a descriptive analysis method, as described in
tive ratings on the intensity scale. They do not have
Chapter 10. The goal is to provide intensity ratings for
to be shown examples that are labeled as “in specifi-
individual sensory attributes by a trained panel. The
cation” or “out of specification,” however, since that
focus is on the perceived intensity of single attributes
decision is based on the overall profile of the prod-
and not quality or overall difference. Intensity rating
uct and is done by the sensory panel leader or QC
of single sensory characteristics demands an analytical
management. Defective samples can be used in train-
frame of mind and focused attention on dissecting the
ing intensity ratings, but the actual cutoff points are
sensory experience into its component parts. Muñoz
better kept in confidence by managers making the deci-
and coauthors called this as a “comprehensive descrip-
sions about product disposition (Muñoz et al., 1992).
tive method,” but they do allow for limitation of the
This will avoid the tendency for panelists to gravi-
scorecard to a small set of critical attributes. For QC
tate toward scores that are just within the acceptable
purposes, attention to a few critical attributes may be
range.
appropriate.
Advantages. The detail and quantitative nature of
As in the other techniques, calibration must be
the descriptive specification lends itself well to correla-
done. Specifications for the descriptive profile must
tion with other measures such as instrumental analysis.
be set via consumer testing and/or management input.
The second advantage is that it presents less of a cog-
This will consist of a range of allowable intensity
nitive burden on panelists, once they have adopted
scores for the key attributes. Table 17.1 shows an
the analytical frame of mind. They are not required
example of a descriptive evaluation of potato chip
to integrate their various sensory experiences into an
samples and the range of sensory specifications, as
overall score, but merely report their intensity per-
previously determined in a calibration study with con-
ceptions of the key attributes. Finally, the reasons for
sumers and/or management input (from Muñoz et al.,
defects and corrective actions are easier to infer since
1992). This sample is below the acceptable specifi-
specific characteristics are rated. These can be more
cation limits in evenness of color and is too high in
closely associated with ingredients and process factors
cardboard flavor, a characteristic of lipid oxidation
than an overall quality score.
problems.
Limitations. Because the method depends on good
intensity anchoring, it tends to be more laborious in
panel training than some of the other techniques. Due
to the need for data handling and statistical analysis, as
Table 17.1 Evaluation of potato chip samples using descrip-
tive specification well as having a sufficient number of trained judges on
Mean panel score Acceptable range hand for the method, it is better suited for quality evalu-
Appearance ation of finished products. It may be difficult to arrange
Color intensity 4.7 3.5–6.0 for descriptive evaluation for ongoing production, par-
Even color 4.8 6.0–12.0 ticularly on later work shifts if production is around
Even size 4.1 4.0–8.5 the clock. The training regimen is difficult and time
Flavor
consuming to set up since examples must be found for
Fried potato 3.6 3.0–5.0 the range of intensities for each sensory attribute in the
Cardboard 5.0 0.0–1.5 evaluation. This can require a lot of technician time in
Painty 0.0 0.0–1.0 sample preparation. Another liability is that problems
Salty 12.3 8.0–12.5 may occur in some attributes that were not included on
the scorecard and/or outside of the training set. Thus,
Texture
Hardness 7.5 6.0–9.5 the method lends itself to situations where the problem
Crisp/crunch 13.1 10.0–15.0 areas are well known and the production and ingredi-
Denseness 7.4 7.4–10.0 ent variability can be easily reproduced to make up the
From Muñoz et al. (1992) training set.
17.3.6 A Hybrid Approach: Quality procedures, the boundaries for out-of-spec product and
Ratings with Diagnostics the selection of a gold standard must be undertaken
before training, preferably in a consumer study, but at
least with management input. These defined samples
Gillette and Beckley proposed a reasonable compro- must then be shown to subjects to establish concept
mise between the quality rating method and a com- boundaries. In other words, tolerance ranges must be
prehensive descriptive approach at the 1992 meeting shown to panelists (Nakayama and Wessman, 1979).
of the Institute of Food Technologists (described in
Beckley and Kroll, 1996). The centerpiece of this pro-
cedure is a scale for overall quality. The quality scale 17.3.7 The Multiple Standards
is accompanied by a group of diagnostic scales for
Difference Test
individual attributes. These attributes are key sensory
components that are known to vary in production.
Muñoz et al. (1992, pp. 138–139) describe a similar Amerine et al. (1965) mentioned a variation on sim-
modification of the overall quality ratings method to ple difference tests that would include a non-uniform
include the collection of descriptive information on key or variable standard. This has come to be known as
attributes. In the method of Gillette and Beckley the the multiple standards difference test. Although there
main scale takes this form: is scant literature on the procedure, it has apparently
1 2 3 4 5 6 7 8 9 10
Reject Unacceptable Acceptable Match
On this scale, a product that is so clearly deficient enjoyed some popularity. The idea is to give a forced-
to call for immediate disposal gets a score of 1 or choice test in which participants pick which one of
2. Products that are unacceptable to ship but might several alternative products is the most different from
be reworked or blended get a score in the range of the rest of the set. The simplest approach is to have
3–5. If evaluation is online during processing, these one test product and K alternative versions of the
batches would not be filled into retail containers or standard product. Rather than representing identical
packaged but would be held for rework or blending. versions of a gold standard, the standards are now cho-
If the samples are different from the standard but in an sen to represent the acceptable range of production
acceptable range, they receive scores of 6–8, and sam- variability. The choice of standards to represent the
ples that are a near match or considered identical to range of acceptable variation is critical to the success
the standard receive a 9 or 10, respectively. According of this approach. Historically, this method resembles
to Muñoz et al. (1992), the use of the terms “accept- Torgerson’s “method of triads” of which the triangle
able” and “unacceptable” here is unfortunate, for it test is a special case (Ennis et al., 1988). Pecore et al.
gives the panelists an impression of the action stan- (2006) and Young et al. (2008) used a similar approach,
dards for products passing or failing and the feeling except that overall degree of difference rating was
that they are responsible for decisions about product used (discussed below). The choice of products to be
disposition. This creates a tendency to use the middle included in the set of acceptable standards is critical. If
to the upper end of the scale, to avoid grading products they do not reasonably bracket the range of acceptable
as unacceptable (Rutenbeck, 1985). variation, then the test will be too sensitive (if the range
The advantages of this method are its outward sim- is small) or too insensitive to detecting bad samples (if
plicity in using an overall rating and the addition of their range is too large).
attribute scales to supply reasons for product rejec- If there are a large number of testers (N = 25 or
tion. The method also recognizes that there are situ- more), as in a discrimination procedure, the z-score
ations where products will not match the gold standard approximation to the binomial distribution may be
exactly, but still are acceptable to ship. As in the other used for hypothesis testing. The approximation is
17.4 Recommended Procedure: Difference Scoring with Key Attribute Scales 415
(P − 1/k) − (1/2 N) these three pieces of data, three comparisons can be

z= (17.1)
made: Each mean difference score of the test–control
(1/k)(1 − 1/k) N
pair is compared to the mean difference score of
the control–control pair. Also the average test–control
where k is the total number of alternatives (test product rating is compared to the same baseline. Thus the
plus variable standards), P is the proportion select- within-control variation is taken into account in the
ing the test product as the outlier (the most different comparisons, which must be significantly exceeded if
sample), and N is the number of judges. This is the the test lot is to be found different (and thus action-
same formula as in the triangle and other forced-choice able). Of course, this kind of test requires a sufficient
procedures, except that k may be 4 or larger, depending panel size to get a meaningful and statistically pow-
on the number of references. erful test. Later, Young et al. (2008) extended the
Although this method appears simple at first, there model and procedure to include two test lots as well as
are a few concerns and potential pitfalls in its appli- two control lots and used an incomplete block design
cation. First, in many QC situations, it may not be to limit the six comparisons to three comparisons
feasible to conduct a difference test with sufficient per panelist. The critical comparison in this case is
numbers of judges to have a meaningful and sensitive between the average of the four means comparing tests
application of the statistical significance test. Second, to controls versus the average of the mean control–
the failure to reject the null (failure to get a significant control score and the mean test–test lot score. So the
result) does not necessarily imply sensory equivalence. baseline becomes the average difference within control
From a statistical perspective, it is difficult to have con- lots and within test lots.
fidence in a “no-difference” result, unless the power of
the test is very high. Statistical confidence in the equiv-
alence decision can only be obtained after beta-risk 17.4 Recommended Procedure:
is estimated against a suitable alternative hypothesis
Difference Scoring with Key
(see Appendix E on test power). One approach is to
use the analysis for significant similarity, as outlined in
Attribute Scales
Chapter 5. This will necessarily entail a larger number
of judges (N = 80 or so). Finally, the tests for over- This method is similar to the hybrid procedure of
all difference such as the triangle procedure are known Gillette and Beckley except that it substitutes the over-
to have high inherent variability, so the introduction of all difference scale for the quality scale. This avoids the
even more variability by multiple standards makes it problem that panelists may react to words like “reject”
very difficult to get a significant difference and reject and avoid them. So the method is similar to that of
product. This factor may contribute to a high level of Section 17.3.6. Some category or line version of the
beta-risk, i.e., the chance of missing a true difference. difference scale should be used such as the following:
1 2 3 4 5 6 7 8 9 10
Completely Very Somewhat

Match
different different different
A similar approach to the multiple standards choice The ballot should also include diagnostics on key
test was described by Pecore et al. (2006) and Young attributes, those that will vary in production and are
et al. (2008) but using degree of difference ratings, likely to cause consumer rejection. For attributes that
rather than a choice test. This approach was part of can be too strong or too weak, using just-right scales
the original intent of the degree of difference test pro- is appropriate. Some defects may be a problem at
posed by Aust et al. (1985). In this method, a test higher levels, and intensity scales are useful for those
lot is compared to each of two control lots. The two attributes. Others may warrant product rejection at any
control lots are also compared to each other. From level whatsoever, and a checklist can be provided for
Fig. 17.2 A sample ballot for Apple Juice QC Ballot

apple juice, using the
recommended procedure of Sample____589___________ Judge_____14 (MK)____
degree-of-difference scale Date/session___12/25/09____ Plant site_Dunkirk______
plus diagnostic attribute
ratings. Note that some Overall difference Rating:
attributes use the
just-about-right scale while 1 2 3 4 5 6 7 8√ 9 10
Extremely Moderately Slightly Match
others are better suited to a
Different different different
simple intensity scale. For
off-flavors or defects that are
objectionable at any level, a Attributes:
simple checklist is useful. too low about right too high
Note that this method must be Sweet ___ _X_ ___ ___ ___
used with a well-trained Sour ___ ___ ___ _X_ ___
panel. Color ___ ___ _X_ ___ ___
too sour about right too sweet

Sweet/sour ___ _X__ ___ ___ ___
RATIO
Strength:
none / low very strong
Sweet ___ _X__ ___ ___ ___
Sour ___ ___ _X_ ___ ___
Apple Aroma ___ ___ _X_ ___ ___
Apple Flavor ___ ___ _X_ ___ ___
Off aroma _X_ ___ ___ ___ ___

(list/describe) ____________
Off Flavor _X_ ___ ___ ___ ___

(list/describe) ____________
Checklist: (circle any defects)
Vinegar-like Butyric Lactic acid Painty/solvent Fusel Oil
Sauerkraut-like Other fermented Bitter Astringent Musty
Other (list)_________________________________________
Comments_________________________________________
those more serious faults. A sample ballot for apple be invited for panel training and others who score well
juice is shown in Fig. 17.2. can be kept on file for future replacements as panelist
Screening panelists for sensory acuity and a good attrition occurs (plan from day 1!). Ideally the screen-
training regimen are key here, as in setting up other ing pool of volunteers should be two to three times
quality control panels. The screening procedure should the desired panel size. Supervisory approval is a key
use the types of products that people are going to to good attendance and participation.
eventually judge and insure that they can discrimi- After screening, training may take six to ten ses-
nate among common levels of ingredients like sugar or sions depending on the complexity of the product.
acid content and process variables like heating times Gross differences are illustrated early in training and
or processing temperatures. A sample screening pro- smaller differences shown as training progresses. The
cedure for an apple juice panel is shown in Appendix 1 goal is to solidify the conceptual structure of the pan-
of this chapter. Screening should involve a number elists so they know the category boundaries for the
of attributes and, if possible, different tasks or tests quality ratings and the expected levels of the sensory
(Bressan and Behling, 1977). The top performers can attributes. Panelists must also come to recognize how
17.5 The Importance of Good Practice 417
off-flavors, poor texture, or appearance problems factor using a hot-iron technique to elicit the aroma and
into their overall score. came to a consensus judgment using a 6-point scale
With small panels, there is no statistical analysis, for odor intensity. A second evaluation was done after
but rules of thumb must be established for taking the samples were sent to the laboratory for instru-
actions. It is difficult to apply mean ratings to less than mental analysis. In this case, three panelists were
about eight panelists. Since there are differences in screened for sensitivity to androstenone, and means
individual sensory ability, poor ratings by just a few were calculated from a 9-point scale for odor intensity.
individuals may be indicative of potential problems. Evaluations were performed in a laboratory exhaust
Thus action criteria should take into account negative hood and preparation procedures were standardized.
minority opinions and weight them more heavily than The correlation with instrumental measures was +0.27
high outliers or a few panelists who thought the prod- for the first evaluation (not significantly different from
uct matched the standard (but who may have missed zero correlation) and +0.40 (statistically significant)
some important difference). For example, if two pan- in the second. The increased correlation could be
elists rate the sample on scale point two, but the rest due to a number of methodological factors that were
give it six, seven, or eight, the mean score could be improvements in the second evaluation. These include
in the acceptable range in spite of the two panelists (1) better evaluation location (fume hood versus pack-
who spotted potentially important problems. The panel ing house), (2) screening of judges, (3) constancy of
leader should take note of the two low values and at panel members instead of people dropping in and out,
least call for retesting of this questionable sample. Of (4) averaging scores versus a consensus procedure, and
course, consistent patterns of disagreement between (5) a more standardized sample preparation method.
panelists are a hint that some retraining may be Each of the shortcuts might have been introduced on
needed. some practical grounds, but their combined effect was
to increase the error level in the data. This made
a difference in the statistical significance and con-
clusions of the study regarding instrumental-sensory
17.5 The Importance of Good Practice correlations.
Table 17.2 gives a number of guidelines for good
In all small-panel sensory assessments, the general sensory practice in quality evaluations and Table 17.3
principles of good testing become especially important gives guidelines for judges (adapted from Nelson and
since there are shortcuts often taken in these proce- Trout, 1964). As in any other sensory test, product
dures that are not part of standard sensory evaluation samples should be blind coded and presented in dif-
practices. Most notably, quality assessment may not ferent random orders to each panelist. If production
entail any statistical analysis, due to the small numbers personnel are used in the panel who know the identities
of panelists. Statistical methods provide some insur- of some of the products pulled for evaluation, another
ance against false alarms due to random variation or technical person must blind code them and insert blind
errors of missing important differences. Without the controls into the test set. The person who pulls the
aid of statistical analysis, other safeguards for insuring samples must not evaluate the samples. It is not reason-
the quality of the information take on an even higher able to expect that person to be objective and discount
level of importance. any knowledge of the product identity. Serving tem-
It is worth considering an example concerning pork perature, volume, and any other details concerning
inspection since it illustrates many of the pitfalls product preparation and the tasting method should
involved in small-panel experiments. This was a study be standardized and controlled. Facilities should be
of boar taint or sex aroma from androstenone, a prob- odor free and distraction free. Evaluations should be
lem odor in the fatty tissues of adult male swine. made in a clean sensory testing environment with
The goal was to correlate sensory panel scores for booths or separators, not on the benchtop of an ana-
this taint with instrumental measures of androstenone lytical instrument lab or on the manufacturing floor
content (Thompson and Pearson, 1977). Two sensory (Nakayama and Wessman, 1979). Warm-up samples
analyses were done. In the first, three to five pan- are useful. Blind replicates can be introduced to check
elists sampled boar taint aroma in the packinghouse judge consistency. Judges should taste a representative
Table 17.2 Ten guidelines for sensory quality testing

1. Establish standards for optimum quality (“gold standard”) target plus ranges of acceptable and unacceptable products
2. Standards should be calibrated by consumer testing if possible. Alternatively, experienced personnel may set standards but
these should be checked against consumer opinion (users of product)
3. Judges must be trained, i.e., familiarized with standards and limits of acceptable variation
4. Unacceptable product standards should include all types of defects and deviations likely to occur from materials, processing,
or packaging
5. Judges may be trained to give diagnostic information on defects, if standards are available typifying these problems. Scaled
responses for intensity or checklists may be used
6. Data should always be gathered from at least several panelists. Ideally, statistically meaningful data should be gathered (ten
or more observations per sample
7. Test procedures should follow rules of good sensory practice—blind testing, proper environment, test controls, random orders
8. Blind presentation of standards within each test should be used to check for judge’s accuracy. It is important to include a
(blind) gold standard for reference purposes as well
9. Judge reliability may be tested by blind duplicates
10. Panel agreement is necessary. If unacceptable variation or disagreement occurs, re-training is warranted
Table 17.3 Guidelines for participation in sensory assessments

1. Be in correct physical and mental condition
2. Know the score card
3. Know the defects and the range of probable intensities
4. For some foods and beverages, it is useful to observe aroma immediately after opening the sample container
5. Taste a sufficient volume (Be professional—not timid!)
6. Pay attention to the sequence of flavors
7. Rinse, occasionally, as the situation and product type warrant
8. Concentrate. Think about your sensations and block out all other distractions
9. Do not be too critical. Also, do not gravitate to the middle of the scale
10. Do not change your mind. Often the first impression is valuable, especially for aromas
11. Check your scoring after the evaluation. Get feedback on how you are doing
12. Be honest with yourself. In the face of other opinions, “stick to your guns”
13. Practice. Experience and expertise come slowly. Be patient
14. Be professional. Avoid informal lab banter and ego trips Insist on proper experimental controls—watch out for benchtop
“experiments”
15. Do not smoke, drink, or eat for at least 30 min before participation
16. Do not wear perfume, cologne, aftershave, etc. Avoid fragranced soaps and hand lotions
Modified from Nelson and Trout (1964)
portion (not end of batch or other anomalous parts of trained to recognize the attributes, scoring levels and,
production). of course, know the scorecard. Judgments should be
Other rules of thumb apply to the judges or pan- independent without conferring, jury style. Discussion
elists. Panelists should be screened, qualified, and or feedback can be given later to provide for ongoing
motivated with suitable incentives. They must not calibration. A special liability arises when manufac-
be overtaxed or asked to test too many samples in turing personnel are used who have a lot of pride in
one day. Rotation of the panel at regular intervals the product accompanied by false confidence in the
can improve motivation and relieve boredom. Judges infallibility of the manufacturing process. Such pan-
should be in good physical condition, i.e., free from elists may be unwilling to “rock the boat” and call
ailments like colds or allergies that would detract attention to problem areas. Testing with blind out-of-
from their performance. They should not be men- spec samples, known defects, and other such “catch
tally harried from other problems on the job when trials” accompanied by feedback when they pass defec-
arriving for testing, but should be relaxed and able tive samples can help counteract this overly positive
to concentrate on the task at hand. They must be attitude.
17.6 Historical Footnote: Expert Judges and Quality Scoring 419
The data should consist of interval scale mea- various languages of olive oil producing countries
surements where possible. If large panels are used (International Olive Oil Council, 2007).
(ten or more judges), statistical analysis is appro- Two further examples of commodity judging sys-
priate and data can be summarized by means and tems by trained or expert panels are shown below. The
standard errors. If very small panels are used, the sensory specialist should search for such professional
data should be treated qualitatively. Frequency counts organizations and specifications if they are assigned
of individual scores should be reported and consid- to develop methods for such a food commodity. The
ered in action standards. Deletion of outliers can be methods are poorly suited to processed engineered
considered, but a few low scores (i.e., a minority opin- foods that do not fall into the category of a standard-
ion) may be indicative of an important problem, as ized commodity, but they can provide a useful starting
noted above. Re-tasting may be warranted in situa- place for development of a quality monitoring system
tions where there is strong disagreement or high panel for a closely related product. The sensory specialist
variability. should be careful, however, not to force-fit a standard-
ized grading scheme to a product that is substantially
different. For example, the quality evaluation scheme
for grading vanilla ice cream would be only poorly
17.6 Historical Footnote: Expert Judges suited to sensory testing on a frozen yogurt product
and Quality Scoring made from goat’s milk.
17.6.1 Standardized Commodities

17.6.2 Example 1: Dairy Product Judging
The food industry benefits from standardization of
grades for foods that are minimally processed from A longstanding tradition in the field of dairy products
raw ingredients, from a single source without multiple has been the quality grading schemes for assessing
components, and closely associated with a single agri- product defects and assigning overall quality scores.
cultural commodity. Such “food commodities” include The American Dairy Science Association continues
many dairy products such as milk, cheese, and but- to hold a decades-old student judging competition, in
ter; fruits such as olives; some kinds of meat; and which students and teams of students attempt to dupli-
wine. Various industries and governments have estab- cate the quality scores of established experts. Various
lished quality grades for food commodities or systems defective products are supplied, and students must be
for scoring them based on two main factors: sim- able to recognize the defect, subtract the appropriate
ilarity to a product ideal and lack of defects. The penalty given the type and severity of the problem,
value in such quality grading, scoring, or monitoring and arrive at an overall score (Bodyfelt et al., 1988).
is the assurance to the consumer that the product will The support for quality judging in dairy products is not
have the sensory properties that they have come to universal, however. Some countries like New Zealand
expect. have replaced the overall quality judging method with
Sometimes these systems are defined by interna- ratings on specific key attributes for dairy product
tional organizations in order to provide standards of analysis.
identity for the food commodity. An example is the However, these methods do persist and find some
International Olive Oil Council (COI). The COI pro- utility in small plant quality control and in govern-
vides written standards for sensory evaluation includ- ment inspections (Bodyfelt et al., 1988; York, 1995).
ing definitions for the vocabulary of sensory properties An example of the quality judging scheme for cot-
and defects, a standardized scorecard, a point sys- tage cheese is shown in Table 17.4, listing defects and
tem for assigning grades or classifications, methods their point deduction values for slight, definite, and
for panel training, certification of laboratories eval- pronounced levels of sensory intensity. An extensive
uating olive oil, and even specifications of the tast- discussion of quality judging for dairy products can be
ing glasses that are to be used in the evaluations. found in “Sensory Evaluation of Dairy Products” by
Their website provides all of this information in the Bodyfelt et al. (1988).
Table 17.4 A point deduction scheme for cottage cheese quality grading. Cottage cheese scoring guide
Slight Distinct Pronounced
Appearance (5 points maximum):
Lacks cream 4 3 2
Shattered curd 4 3 2
Free cream 4 2 1
Free whey 4 2 1
Texture (5 points maximum):
Weak/soft 4 3 2
Firm/rubbery 4 2 1
Mealy/grainy 4 2 1
Pasty 3 2 1
Gelatinous 3 2 1
Flavor (10 points maximum):
High acid 9 7 5
High salt 9 8 7
Flat 9 8 7
Bitter 7 4 1
Diacetyl/coarse 9 7 6
Feed 9 7 5
Acetaldehyde/green 9 7 5
Lacks freshness 8 5 1
Malty 6 3 1
Oxidized 5 3 1
Fruity 5 3 1
Musty 5 3 1
Yeasty 4 2 1
Rancid 4 2 1
Rate the presence of each defect as slight, distinct, or pronounced. Give scores for appearance, texture, and
flavor based on the table.
Other problems may include discoloration, matted curd, slimy texture, foreign flavors, unclean flavors
(describe), and fermented flavors
Modified from Bodyfelt et al. (1988)
Such methods are poorly suited to food research in milk were viewed less critically on the average by
where the processed or engineered food is not a stan- consumers than the suggested ADSA scores would dic-
dard commodity and/or when the sensory changes are tate. Of course, having a point deduction scheme that
not likely to be a set of predictable defects. In new is more severe that the average consumer opinion pro-
food product development, it is not necessarily clear vides a kind of safety net and insures that the most
what consumers or segments of consumers may like, sensitive consumers will not be offended by poor prod-
so assignment of quality scores based on some arcane ucts. The liability in a stringent “safety net” is that
or traditional knowledge of experts is not useful. The acceptable product batches will be rejected.
dairy judging methods have been repeatedly criticized
for lack of applicability to research problems, viola-
tions of sensory evaluation principles, and problems in 17.6.3 Example 2: Wine Scoring
scaling and statistical analysis (Hammond et al., 1986;
McBride and Hall, 1979; O’Mahony, 1981; Pangborn Beyond manufacturing control and government
and Dunkley, 1964; Sidel et al., 1981). Furthermore, inspection there are other situations in which the
the opinions of expert judges and standard point consuming public desires information on product
deduction schemes may not correspond to consumer quality. Rather than deducting points from some
opinion as shown in Fig. 17.3. The oxidized defects widely accepted standard, there are also products
17.6 Historical Footnote: Expert Judges and Quality Scoring 421
Consumer Ratings An early method for wine quality assessment was

15 the 20-point rating system developed at the University
Metallic Oxidized
of California at Davis (Amerine and Roessler, 1981;
13 Ough and Baker, 1961). This was an additive scheme
Mean Acceptability Score
Light Oxidized
Rancid for giving overall quality scores. It was based on
11
the analysis of quality for sensory categories such
as appearance, body, flavor, and aftertaste, as well
9
as some specific attributes like sweetness, bitterness,
7 and acidity. As shown in Table 17.5, different points
are given for different categories, i.e., there is uneven
5 weighting, presumably due to the different contribu-
tions of each category to the overall quality. Note that
3 this does not produce a scale value in the psychophys-
ical sense, but a score.
1
0 1 2 3 4 Like other quality grading schemes, this method can
(control) be criticized on a number of grounds. First, the weight-
Nominal Level of Defect
ing system is somewhat arbitrary—different versions
10 Metallic Oxidized
can be found—and it was based on the expert opinion
Light Oxidized of the method’s originators, rather than any consumer
Recommended ADSA Score
Rancid opinion. Second, whether wine quality can actually

8 be captured by an additive scheme is questionable.
Some defects (e.g., bitterness) are simply too serious
to provide any good score at all, even though all other
6
attributes might add up to some positive number. Some
versions of the technique try to allow for this by pro-
4 viding a few overall quality points to add into the total,
a kind of global fudge factor. Also, it is an anecdo-
tal observation that some judges who gain experience
2
with the technique score an overall quality level of
wines first and do not bother assigning individual cat-
0 egory points to start. Rather, they first decide on an
0 1 2 3 4
(control) overall score and then allot points into the individual
Nominal Level of Defect
categories, using the method backward.
Fig. 17.3 Consumer ratings of abused milk samples compared A simplified alternative procedure is based on
to the recommended ADSA scores for those products based on hedonic scoring by experienced fine wine drinkers
the recipes and rating systems shown in Bodyfelt et al. (1988). (Goldwyn and Lawless, 1991). The assumption was
From Lawless and Claassen (1993) with permission.
that a small panel of experienced tasters can provide
recognition of good to superior products based on
where excellence is recognized beyond the merely their personal likes and dislikes. The method works
acceptable. Garvin (1987) remarked that considera- to the extent that fine wine drinkers form a cul-
tions of quality should include the ability to please tural and linguistic community (Solomon, 1990) with
consumers, not just protect them from annoyances. known and consensual standards of taste at least within
This idea is further developed in the Kano model in a given geographical area. The method used a bal-
Chapter 19. Some products show a wide range of anced 14-point hedonic scale (like extremely to dislike
better-than-acceptable variation. Wines are a good extremely) with no neutral center point. Wines were
example. What kind of quality measurement system tasted twice and the second final score was recorded.
can go beyond point deductions for defects to provide The procedure followed principles of good practice
degrees of difference on the positive end of the quality such as randomized orders, blind coding, independent
continuum? judgments (no conferring), standardized presentation,
Table 17.5 Example of 20-point wine scoring scheme

Characteristic Scoring guide Maximum points
Appearance Cloudy 0, clear 1, brilliant 2 2
Color Distinctly off 0, slightly off 1, correct 2 2
Aroma and bouquet Vinous 1, distinct but not varietal 2, varietal 3 subtract 2 for off-odors and 1 for bottle 4
bouquet
Vinegary Obvious 0, slight 1, none 2 2
Total acidity Distinctly high or low 0, slightly high or low 1, normal 2 2
Sweetness Too high or low 0, normal 1 1
Body Too high or low 0, normal 1 1
Flavor Distinctly abnormal 0, slightly abnormal 1, normal 2 2
Bitterness Distinctly high 0, slightly high 1, normal 2 2
General quality Lacking 0, slight 1, impressive 2 2
Total score 20
Modified from Amerine and Roessler (1981)
palate cleansers, and a reasonable pace of tasting. schemes must be developed and agreed upon as well
Flights consisted of seven wines and the pace was as standard procedures for sample handling and stor-
limited to at least 30 min per flight, allowing time age. Data handling, report format, historical archiving,
for palate recovery. Blind duplicates were periodically and tracking and panelist monitoring are all important
introduced to check on judge reliability. tasks. It is extremely important that a sensory evalu-
These methods represent improvements over the ation coordinator with a strong technical background
types of informal consensus tastings done by juries to in sensory methods should be assigned to carry out
award medals at state fair competitions. Such evalu- these tasks (Mastrian, 1985). Aside from these prac-
ations have almost no scientific merit, i.e., they are tical operational concerns, the system should also have
of about the same value as a movie critic’s review. certain features that maintain the quality of the eval-
Analysis of 3 years of data from blind duplicate sam- uation procedures themselves. For example, a method
ples in the California state wine judging has shown that for measuring the overall effectiveness of the system
among non-defective wines, any grade or medal may should be identified (Rutenbeck, 1985). External audit-
be assigned to any wine in different competitions, and ing at periodic intervals may be useful (Bauman and
about 90% of judges were unable to reproduce their Taubert, 1984).
scores (Hodgson, 2008). Gillette and Beckley (1992) listed requirements for
a good in-plant sensory QC program and ten other
desirable features. These concerns are taken from the
17.7 Program Requirements perspective of ingredient suppliers to a major food
and Program Development manufacturer but can be modified to fit other man-
ufacturing situations. A sensory QC program must
involve human evaluation of the products. It must be
17.7.1 Desired Features of a Sensory QC acceptable to both suppliers and customers. Results
System must be easily communicated so that reasons for rejec-
tion and actions to be taken are both made clear. It
Rutenbeck (1985) and Mastrian (1985) outlined the should take into account an acceptable range of devi-
program development of a sensory QC system in terms ation, recognizing that some products will not match
of specific tasks. These included research into avail- the gold standard but will still be acceptable to con-
ability and expertise of panelists, availability or access sumers. Of course, the program must be able to detect
to reference materials, and time constraints. Panelist unacceptable production samples.
selection, screening, and training on objective terms Additional desirable features include the following:
(such as “high saltiness,” as opposed to vague terms Potential transfer over time to an instrumental measure
like “poor quality”) must be undertaken. Sampling is a good goal if the evaluations are very repetitive,
17.7 Program Requirements and Program Development 423
as instruments do not become fatigued or bored with QC program and suggested calculations of measurable
the testing regimen. This presumes that tight sensory– results, such as reductions in consumer complaints,
instrumental correlations can be established. Ideally, a cost savings in avoiding rework or scrapping materials,
sensory QC program should provide rapid detection for and potential impact on sales volume. Manufacturing
online corrections. Information should be quantitative executives unfamiliar with sensory testing can eas-
and interface with other QC methods. As quality con- ily underestimate the complexity of sensory tests, the
trol and shelf-life tests are often similar, the methods need for technician time to setup, the costs of panel
will be more useful if they are transferable to shelf-life startup and panelist screening, and training of techni-
monitoring. Many of the changes over time in a stored cians and panel leaders as well as panelist incentive
product are also quality problems, such as deterioration programs (Stouffer, 1985). If employees are used as
in texture, browning, oxidation, syneresis, oiling-out, panelists, another stumbling block can be the personnel
staling, and off-flavor development (Dethmers, 1979). time away from the person’s main job to come to sen-
Sensory evaluations may include raw ingredient test- sory testing (and any associated costs). However, panel
ing as well as in-process and finished products. A good participation can be a welcome break for workers,
sensory QC program will produce a track record of can enhance their sense of participation in corpo-
actually flagging bad products to prevent further prob- rate quality programs, can expand their job skills and
lems down the line or in consumer opinion in the their view of manufacturing, and does not necessarily
marketplace. result in a loss of productivity. There are considerable
Considering these desired features, there are some advantages in using panelists from the processing oper-
traditional test methods from the mainstream of sen- ation, notably in accessibility and interest (Mastrian,
sory evaluation that simply do not apply well to in- 1985). Arranging for a sensory testing space may also
plant quality control work. Problems arise if the tests involve some startup costs. An important issue con-
cannot handle a sufficient volume of production sam- cerns what will be done to insure continuity in the
ples. Any test procedure that has a slow turnaround program. Management must be made to see that the
in analysis and reporting of results will not be suit- sensory instrument will need maintenance, calibration,
able for online corrective action in the manufacturing and eventual replacement. Concerns include panelist
environment. For example, it is difficult to implement attrition and retraining, refreshment, or replacement of
a descriptive analysis panel for QC work if decisions reference standards (Wolfe, 1979).
have to be made in the middle of the night on third shift An early issue in program development concerns the
and the data cannot be statistically analyzed through definition of standards and cutoffs or specification lim-
an automated system. At first glance, finding defective its (Stevenson et al., 1984). Management or preferably
products would seem to suggest that a simple differ- experienced technical personnel can do the evaluation
ence test from a standard product would be a good and set the limits. This approach is fast and simple, but
approach. However, most sensory difference tests take risky, since there is no consumer input (McNutt, 1988).
the form of forced-choice tasks, like the triangle pro- The safest but slowest and most expensive approach
cedure. The triangle test is useful for detecting any is to give a range of products with representative pro-
difference at all, but is not suitable when there is a duction variation to consumers for evaluation. This
range of acceptable variation. Just because a product is calibration set should include known defects that are
found to be different from the standard does not mean likely to occur and all ranges of processing and ingre-
that it is unacceptable. dient variables. As a small number of consumers will
always be insensitive to any sensory differences, a
conservative estimate of problem areas should be set
based on rejection or failing scores from a minority of
17.7.2 Program Development
participants.
and Management Issues A third issue concerns the level of thoroughness
in sampling that is needed for management comfort
Management may need to be educated as to the cost versus the cost of overtesting. Ideal quality control
and practical issues that are involved in sensory QC. programs would sample materials along all stages of
Rutenbeck (1985) described the “selling” of a sensory production, in every batch and every shift (Stouffer,
1985). This is rarely practical for sensory testing. 100 are objectively defective, while 900 are objectively
Sampling multiple products from a batch or produc- trouble free. If the tests are properly done, and there is
tion run or performing replicate measurements with a statistical protection against Type I and Type II errors,
sensory panel will give insurance against missing out- there will still be some occasions where errors do
of-spec products, but will increase time and costs of occur. For the sake of easy calculation, let the long-
testing. term alpha- and beta-risks for the testing program be
Additional challenges arise from reporting struc- 10%. This means that 10% of the time when a defective
tures, multiple test sites, and the temptation to sub- product is sent for testing, it will go undetected by the
stitute instruments for sensory panels. A built-in con- evaluation, and 10% of the time a product which has
flict of interest occurs when a QC department reports no defects will be flagged, due to random error. This
directly to manufacturing, since manufacturing is usu- will lead to 810 correct “pass” decisions and 90 correct
ally rewarded for productivity. A separate reporting detections of sensory problems. Unfortunately, due to
structure may be desirable for quality control, so that the high incidence of good products being tested, the
executives committed to a corporate quality program 10% false alarm rate leads to 90 products also being
can insulate the QC department from pressures to flagged where there is no true sensory problem. Note
pass bad products. Across multiple plants, there is a that this assumes that there is good sensory testing and
need to standardize sensory QC procedures and coor- proper statistical treatment of the results! The prob-
dinate activities (Carlton, 1985; Stouffer, 1985). This lem arises when the sensory QC leader picks up the
includes maintenance of consistent production sam- phone and calls the manufacturing manager and “rings
ples and reference materials that can be sent to all the alarm bell.” Given this incidence, the probability
other plants for comparison. Setting up similar sen- of being right about the problem is only 50%, in other
sory QC systems in different countries and cultures words, no better than a coin toss! Even if alpha is
may present difficult challenges to the sensory program reduced to the usual 5%, there is still a one-in-three
coordinator. Considerations in panel setup in other cul- chance of false alarms.
tures are given in Carlton (1985). Finally, instruments How can this occur if good sensory testing is done
cannot replace sensory evaluation for many impor- and proper statistics are applied? The answer is that
tant product characteristics (Nakayama and Wessman, our normal inferential statistics are used to view the
1979). Odor analysis is a good example. In other cases, outcome chart in Table 17.6 across rows and not down
the instrumental–sensory relationship may be nonlin- columns. The problem in sensory QC is that given a
ear or indicate changes that are not perceivable at all low incidence of problems, there is simply a high rate
(Rutenbeck, 1985; Trant et al., 1981). of false alarms relative to correct detections. This can
hurt the credibility of the program if manufacturing
managers develop a feeling that the sensory depart-
ment is prone to “cry wolf.” Thus it is wise to build
17.7.3 The Problem of Low Incidence in a system for additional or repeated testing of prod-
uct failures to insure that marginal products are in fact
A special problem with QC testing as well as shelf- defective before action is taken.
life studies is that the majority of the evaluations result
in positive results (favorable decisions for manufactur-
ing). This is in the very nature of the test scenario. 17.8 Shelf-Life Testing
Good products are much more often tested than defec-
tive ones, and there is a much lower incidence of
negative test results than those found in research sup- 17.8.1 Basic Considerations
port testing. This can present a special challenge to the
credibility of the sensory testing program. Shelf-life or stability testing is an important part of
Table 17.6 shows an incidence diagram for a fairly quality maintenance for many foods. It is an inherent
high rate of problem products, in this case 10% (the part of packaging research because one of the primary
credibility problem gets even worse if there is a lower functions of food packaging is to preserve the integrity
rate of defective products). In over 1,000 tests, then, of a food in its structural, chemical, microbiological
17.8 Shelf-Life Testing 425
Table 17.6 Bayesian incidence chart

Outcome of evaluation
Problem reported No problem reported Incidence (=total across row)
Problem exists 90 10 100
(description) (“hit rate”) (Type II error)
No problem exists 90 810 900
(“false alarm”) (correct acceptance)
(Total) 180 820 1000
Let alpha = 0.10 and beta = 0.10
Assume 1,000 tests are conducted, with a 10% rate of faulty products
Numbers in cells show estimated numbers of problems reported or not, based on alpha and beta rates of 10%
Given that a problem was reported, you stand a 50/50 chance of having made the wrong decision (90/180)!
and sensory properties. A good review of shelf-life to establish a suitable degree of stability, i.e., that the
testing can be found in the packaging text by Robertson failure time exceeds the typical distribution and use
(2006) and the reader is referred there for further infor- time by consumers, a combination of two tests may be
mation on modeling and accelerated storage tests. For appropriate. It is cost efficient in this case to perform
many foods, the microbiological integrity of the food discrimination tests against a fresh control or standard
will determine its shelf life, and this can be estimated product and then perform consumer acceptance tests if
using standard laboratory practices; no sensory data any difference is detected.
are required. The sensory aspects of a food are the According to Peryam (1964) and Dethmers (1979)
determining factor for the shelf life of foods that do not a shelf-life program will involve the following steps:
tend to suffer from microbiological changes such as (1) formulating objectives, (2) obtaining representative
baked goods. Sensory tests on foods are almost always samples, (3) determining the physical and chemical
destructive tests, so sufficient samples must be stored composition of the test products, (4) setting up a test
and available, especially during the period in which the design, (5) choosing the appropriate sensory method,
product is expected to deteriorate (Gacula, 1975). (6) choosing the storage conditions, (7) establishing
Shelf-life testing may employ any of the three the control product or products to which the stored
major kinds of sensory tests, discrimination, descrip- product will be compared, (8) conduct the periodic
tive, or affective, depending on the goals of the pro- testing, and (9) determine the shelf life based on the
gram (Kilcast, 2000). Thus one can view shelf-life results.
tests as no special category of sensory testing, but Important strategic choices include the nature of
simply a program of repeated testing using accepted the control product and the storage conditions. Storage
methods. The objectives of the study may dictate conditions should mimic the conditions found in dis-
what method is most suitable to answer the research tribution and those in stores, unless some accelerated
questions (Dethmers, 1979). For a designed study to storage conditions are required. Ideal conditions are
evaluate the effects of a new packaging film, a sim- generally a poor choice. The control product presents
ple discrimination test might be appropriate to test special problems. If a fresh product is available at the
for changes versus the existing packaging. For pur- different time intervals, how does one know that sub-
poses of establishing an open dating system, consumer sequent batches have not drifted or changed since the
acceptability tests would be appropriate to establish initial product was manufactured? If a fresh product
the time that the product is likely to become unac- is stored from the initial batch under ideal condi-
ceptable. If the product is new, a descriptive analysis tions, how does one know it has not changed? There
profile is needed to establish the full sensory specifi- is no perfect solution to this problem. Sometimes a
cation of what a fresh product tastes like. If a product study will involve more than one standard. Reference
has failed a consumer evaluation, it is often appropriate standards should be clearly identified by date, lot
to submit the samples to descriptive testing to try and number, production location, etc. A separate program
understand the reasons for failure and which aspects may be instituted to insure the integrity and con-
have deteriorated (Dethmers, 1979). If the purpose is stancy of reference standards. Descriptive evaluation
may be beneficial for this purpose. Options for refer- above neutral) (Muñoz et al., 1992). Another option is
ences include the following: current plant product that to use consumer rejection (“I would not buy/eat this
has passed QC, current pilot-plant prototype, historical product”) (Hough et al., 2003). These two measures
product, optimally stored product, a written descriptive are not necessarily equivalent. Giminez et al. (2008)
profile, and a mental reference (Wolfe, 1979). found that for certain baked products, consumers might
Two main choices are used for criteria for product not like the product, but they would answer “yes”
failure. These include a cutoff point from a critical when asked if they would consume it at home (having
descriptive attribute (or set of them) and consumer already purchased it). This finding suggests that con-
data when the product is rejected as unacceptable. sumer rejection may not be sufficiently conservative,
Statistical modeling with equations such as a hazard i.e., that a product may become disliked and even gen-
function or survival analysis is discussed below. Note erate consumer complaints before it reaches the point
that product failure is an all-or-none phenomenon, and of rejection. Giminez et al. (2007) found that accept-
decreases in sensory measures such as falling acceptability scores could be related to percent of rejection
ability or increasing percents of consumer rejection are by logistic regression analysis. The logistic equations
more continuous in nature. This opens the opportunity for two different countries (Spain and Uruguay) for
for other kinds of models, such as logistic regression a baked product were different, a warning about cul-
against percent rejecting (Giminez et al., 2007). tural and/or national differences. Logistic regression is
a useful general approach to data in the form of pro-
portions that accumulate in an S-shaped curve. The
general form is
17.8.2 Cutoff Point
p
ln = bo + b1 X (17.2)
1−p
The choice of a cutoff point has two implications. The
first occurs when the cutoff point itself is used as an where p is the proportion rejected and X is the vari-
action standard. When the product gets to this point, able that is the predictor, such as time, or in this case
we consider it to have reached the end of its useful life. acceptability scores (b are constants).
It is no longer salable. The time estimate may be used
for some purpose like open dating or “use-by” dates
printed on the product package. The second implica-
tion is that when a product in a designed study reaches 17.8.3 Test Designs
this cutoff point, it defines “failure” and will be used as
a data point in some kind of statistical modeling such Several options are available for shelf-life tests regard-
as survival analysis. ing how samples are stored and test times. The simplest
Determination of a cutoff point requires careful method is to make one large batch of product, store it
consideration. Several options are available including under normal conditions, and test it at various intervals.
(1) a significant difference in a discrimination test, However, this is not very efficient in terms of test time
(2) some degree of difference from control product and risks the panel drifting its criteria. Another option
on a scaled attribute or overall degree of difference is to stagger the production times, so all the products
scale, and (3) consumer reaction. Consumer data may of different ages are tested on the same day. A vari-
involve a significant difference in acceptability rat- ation on this is to store the product under conditions
ings from control, a cutpoint on an acceptance score, that essentially stop all aging processes, for example,
or some percent of consumer rejection (e.g., 50% or at very low temperature. This is obviously not possi-
25%). Giminez et al. (2007) found the first signifi- ble with all products (you cannot freeze lettuce). Then
cant difference to be too conservative an estimate in products are pulled from the optimal storage condi-
the sense that acceptance scores were still above 6 tions at different times and allowed to age at normal
on the 9-point scale. This makes sense because two temperatures. Another variation of this procedure is
products may differ but still be acceptable (Kilcast, to allow products to age for different times and then
2000). Another option is to use any value less than 6 place them into the optimal storage conditions, pulling
on the 9-point hedonic scale (6 = like slightly, i.e., just everything out of storage at the test date.
17.8 Shelf-Life Testing 427
17.8.4 Survival Analysis and Hazard failure data (or percent of failures) over time. Percent
Functions survival (one minus percent of failures) often takes the
form of a decaying exponential function.
An important choice of model includes the distribu-
The literature on survival analysis is very large, tion used to fit the percent of failures. Many distribu-
because a number of different fields use these kinds tions have been tried (Gacula and Kubala, 1975) but
of statistical models, such as actuarial science for the two useful models are a log–normal distribution (sur-
insurance industry. Some of the models are similar to viving is a positively skewed distribution, few people
those used in chemical kinetics. These functions are live to 100) and a Weibull distribution. Weibull func-
useful when the product has a single process or a group tions are useful distributions that can be used to fit a
of processes that are occurring at about the same time. variety of data sets. They include a shape parameter
However, some products will show a “bathtub” func- and a scale parameter. When the shape parameter takes
tion with two phases of product failure (Robertson, a value greater than 2, the distribution is approximately
2006). In the early stages, some product failures occur bell shaped and symmetric. These equations for failure
due to faulty packaging or improper processing (see take the following forms:
Fig. 17.4). Then the remaining products from that

batch enter a period of product stability. After some ln(t) − μ
time, X, the products begin to fail again, due to dete- F(t) = (log −normal) (17.3)
σ
rioration. Gacula and Kubala (1975) suggest that the
shelf-life modeling should only consider those failures where is the cumulative normal distribution func-
after time X2 . tion, t is time, F(t) is the failure proportion at time
t, μ is the mean failure time, and σ is the standard
deviation.
RATE OF PRODUCT FAILURES, UNITS / WEEK

ln(t) − μ
F(t) = 1 − exp − exp (Weibull)
σ
(17.4)
x
where exp(x) is the notation for e .
Period of
Period of failures deterioration
due to manufacturing If we make two substitutions and determine the
defects or packaging
mean (μ) and standard deviation (σ ) of our failure
times, a simple model can help us find the time for
a given percentage of failures from the fitted Weibull
Period of product stability
equation. Let ρ = exp (–μ). Then the following rela-
tionship holds for any proportion (F(t)):
X0 X1 X2
TIME
− ln (1 − F(t))σ
t= (17.5)
Fig. 17.4 The “bathtub” function showing a common pattern ρ
of failure rates changing over the sampling time in a shelf-
life study. From time X0 to X1 , some products will fail due to Using the log–normal model for F(t), a simple
improper processing or faulty packaging. This is followed by graphic method for finding the interpolated 50% fail-
a period of fairly low failure rates when products are stable or ure level is to do the following: For N samples of foods
within specification limits. At time X2 , failures start to increase
sampled over time that have known failure times Ti ,
markedly. Researchers fitting hazard functions or doing survival
analysis for estimation of shelf life should consider using only rank all the batches, i as to the time of failure (i = 1
those times after X2 in curve fitting as earlier failures are due to to N). Calculate the median ranks, MR values. The
causes other than the time-related deterioration. median rank can be found in some statistical tables or
estimated as
Survival analysis as applied to sensory data has two
main tasks, the fitting of a hazard function to the data MR = (i − 0.3)/(N + 0.4)
and the interpolation of some point used as the criterion
for shelf life (such as 25 or 50% consumer rejection). Plot the median rank on log probability paper ver-
Various functions have been used to fit the function of sus Ti and interpolate at the 50% point. If a straight
line fits the data, the 50% point can be estimated from different kinetic constants change at different rates
the linear equation and standard deviations estimated with different temperatures, the one with the higher
from the probability paper. This is essentially a fit of value may come to predominate. The sensory special-
log MRs to z-scores. Other percentages may be inter- ist should be familiar with the logic of this testing and
polated, of course, because 50% may be too high a the modeling that is commonly done, as well as the
failure percentage for many products in setting a useful pitfalls.
limit. An equivalent mathematical solution is given in
Appendix 2 of this chapter.
Hough et al. (2003) point out that the usual sensory
experiment produces censored data. That is, for any 17.9 Summary and Conclusions
batch that has failed, we only know that the time of
failure was in some interval between the last test and Insuring the quality of products on a sensory basis is an
the current test. Similarly, for a batch that has not failed important corporate goal in a competitive environment.
at the final interval, we only know that its failure time Consumers have fixed expectations and will become
is sometime after that final test. So the data are cen- disloyal to a brand if they experience substandard prod-
sored and the survival function can be estimated using ucts. However, in spite of the need for sensory quality
maximum likelihood techniques. control, setting up and maintaining an in-plant QC
program is difficult and costly. Commitment to the
program is the corporate equivalent of diet and exer-
cise. Everyone admits that it is a good idea. However,
17.8.5 Accelerated Storage maintaining program integrity, avoiding shortcut pro-
cedures, and dealing with dwindling panel size can
Sooner or later, product developers may figure out that be challenging. The success of any program demands
sensory specialists do not own a time machine and can- strong management commitment. Without manage-
not deliver shelf-life estimates without a long study. ment support, especially from manufacturing, a sen-
So they may request some accelerated storage tests to sory QC program is bound to fail. In a typical case
shorten the time. Such tests are based on the idea that the program will amount to nothing more than “rub-
at higher temperatures, many chemical reactions will ber stamping” of supervisory opinion, thus supporting
proceed in a predictable manner, according to simple a management policy that maximizes productivity at
kinetic models, and thus the shelf life at long time the expense of producing unacceptable products. Such
intervals can be simulated by shorter intervals at higher programs will blow hot and cold, usually receiving
temperatures (Mizrahi, 2000). Kinetic models are often some emergency attention when a truly bad product
based on the Arrhenius equation, and rate constants hits the retail shelves and consumer complaints filter
can be found from experiments conducted at differ- back (Rutenbeck, 1985). After a period of improved
ent temperatures. Some of the models are shown in production, some complacency may set in and loss of
Appendix 3 of this chapter. interest in sensory QC efforts (until the next disaster).
Problems in accelerated testing occur when changes Implementation of a sensory QC program will
in the product due to temperature are not the same involve four technical tasks. First, a range of products
as those due to storage time at normal temperatures must be prepared for establishing quality specifications
(Robertson, 2006). Obviously, trying to measure the and limits. Specifications must be set in a research
shelf life of a frozen food at higher temperatures study with consumers or by management or expert
makes little sense. Other foods may not follow simple opinion after sampling the various products. This range
predictive models as multiple processes occur with dif- of products can also be used in panel training. The
ferent rates. For example, at higher temperatures phase second step in the program is recruitment, screening,
change may occur, solid to liquid. Carbohydrates in the and training, in other words, panel setup. Next, stan-
amorphous state may crystallize. The water activity of dard protocols for product sampling, handling, storage,
dry foods may increase with temperature causing an serving, blind coding, and maintenance of reference
increased reaction rate and overprediction of true shelf standards must be established. The fourth step is in
life at the normal temperature. If two reactions with systematizing the paperflow (Mastrian, 1985). This
Appendix 2 429
includes establishing standard reporting formats and Provide palate cleansers (water, crackers).
processes for data handling, recommendations, and Run duplicates of each test.
action criteria. This activity should also include mech-
anisms for archiving results and tracking both products Good performance: 3/4 correct.
and panelist performance across time. Acceptable performance: 2/4 correct.
Part 4. Acidity test

Appendix 1: Sample Screening Tests Adjust pH to about 0.5 versus 1% titratable acidity.
for Sensory Quality Judges Use four paired comparisons, as in sweetness test
above.
Part 1. Paired comparison of sweetness levels
Adjust samples to three levels, e.g., 10, 11, and 12% Good performance: 3/4 correct.
sucrose wt/vol. Acceptable performance: 2/4 correct.
Give four pairs in counterbalanced orders, e.g.,
After scoring, rank order candidates from highest to
10 versus 11%, 11 versus 10% (hard discrimination) lowest.
10 versus 12 %, 12 versus 10% (easier discrimination) Invite the top 50% from each shift for training.
Send thank you notes to all the people who try out.
Use a different order for different panelists. Blind Keep the rest “on file” for possible replacements if
code with random 3-digit labels. scores are acceptable.
Good performance: All four correct.

Acceptable performance: One error if other test sec-
Appendix 2: Survival/Failure Estimates
tions perfect.
from a Series of Batches with Known
Part 2. Multiple choice odor identification. Done with Failure Times
blotters in capped jars
Circle correct answers on sheet. Make up multiple This procedure follows the graphic method given in
forms with different orders. Section 17.8.3 but allows a more exact fit by least
Use random 3-digit codes on bottles. Odors repre- squares regression.
sent common notes in the product.
Use four alternatives, e.g., fruity, smoky, vinegar, 1. For N samples of foods sampled over time that have
onion known failure times Ti , rank all the batches, i as to
the time of failure (i = 1 to N).
(a) Dilute ethyl hexanoate (or similar ester) 2. Calculate the median ranks, MR values. The
(b) Dilute ethyl 2-methyl butyrate median rank can be found in some statistical tables
(c) Dilute vinegar or estimated as
(d) Dilute phenylethanol
(e) Trans-2-hexenol(dilute until green or leafy smell is MR = (i − 0.3)/(N + 0.4)
obtained)
Good performance: 4/5 correct 3. Convert each Ti to ln(Ti ), called Yi . This will permit
Acceptable performance:3/5 correct a fit of MR to the log–normal model.
4. Calculate the z-score for each MR at each Ti . Call
Part 3. Odor discrimination test this Xi .
5. Regress Y against X using least squares to get the
Run triangle tests with base juice and base juice + 1% linear equation Y = a + bX.
vinegar.
Run triangle test with base juice + 0.1% butyric acid. This is equivalent to finding the straight line fit to
Subjects should sniff first, then taste. the log probability plot described in Section 17.8.
6. Then solve for Y = 0 (z-score for 50%) which is This sometimes takes its derivative form:
X = –a/b, to get the 50th percentile.
7. Convert back to the original units by exponentiating d(ln k) EA
= (17.11)
Time at 50% failure = eX = e−a/b . dT RT 2
The activation energy is fictitious in a way, because
there is not a single chemical reaction going on dur-
Appendix 3: Arrhenius Equation ing the aging of a food product, but a large number of
and Q10 Modeling simultaneous processes. Nonetheless, we can think of
this as useful for two reasons. First, it gives an indi-
The reaction time for product failures may be linear cation of the fragility of the food (lower activation
(zero order) or a decaying exponential (first order). energy would mean faster deterioration). Second, the
Both allow determination of a rate constant, K. Let us EA value becomes useful in predicting what happens at
consider a cutoff point rating on some scale, R, as the different temperatures. Of specific interest is predicting
event to be modeled as function of time, t. R could also what will happen at “normal” temperature given that an
be any event that signals product failure. accelerated storage study has been conducted at higher
The zero-order equation is temperatures.
Experiments are often performed at varying temper-
atures, 10◦ C apart, to generate what is known as the
R = Ro − kt (17.6)
Q10 factor.
And the first-order relationship is
kT+10 ST
Q10 = = (17.12)
−kt kT ST+10
R = Ro e (17.7)
where kT+10 and kT are the rate constants at
where Ro is the rating or failure at t=0 temperature T and T + 10, and ST and ST+10 are the cor-
and responding shelf-life estimates. Note that the ratio of
rate constants is the inverse of the ratio of the shelf-life
R
ln = −kt (17.8) times.
Ro This produces some useful relationships, for exam-
Reaction rates are also dependent on temperature, ple, to use in estimating the activation energy, EA
so in accelerated storage studies, the Arrhenius equa-
10EA
tion provides a starting point or a generally useful ln Q10 = (17.13)
approximation: RT 2
Again, EA can give some idea of the susceptibility

−EA
of the product to deterioration. Once the Q10 factor has
k = ko e RT
(17.9)
been determined for a product, the time–temperature
and relationship can be predicted for the accelerated tests.
A useful factor to determine is an acceleration factor,
AF. This will help us convert from the accelerated tem-
EA 1
ln k = ln ko − (17.10) perature back to a usage temperature or normal storage
R T
condition temperature like 20◦ C. Hough (2010) gives
where k is the rate constant to be estimated, ko is a con- the following example, based on an off-flavor rating
stant independent of temperature (also known as the (OF) as a function of time and temperature:
Arrhenius, pre-exponential, collision, or frequency fac-
tor), EA is the activation energy (J/mol), R is the ideal OFT,temp = OFo + (AF)ku (Ttemp ) (17.14)
gas constant, T is temperature (absolute, K).
So a plot of ln(k) versus 1/T can be used to find the where OFo is the off-flavor at time zero, ku is the
activation energy, EA . rate constant at usage temperature, and Ttemp is the
References 431
accelerated test temperature. Knowing EA , we can also Carlton, D. K. 1985. Plant sensory evaluation within a multi-
estimate the AF from plant international organization. Food Technology, 39(11),
130–133, 142.
Dethmers, A. E. 1979. Utilizing sensory evaluation to determine
EA 1 1
AF = exp − (17.15) product shelf life. Food Technology, 33(9), 40–43.
R Tu Ttest Ennis, D. M., Mullen, K. and Frijters, J. E. R. 1988. Variations
of the method of triads: Unidimensional Thurstonian models.
where exp(X) is eX , Tu is the usage temperature, and British Journal of Mathematical and Statistical Psychology,
Ttest is the accelerated test temperature. 41, 25–36.
For example, if we determine that the EA is Gacula, M. C. 1975. The design of experiments for shelf life
study. Journal of Food Science, 40, 399–403.
6,500 cal/mol, then we can calculate an acceleration Gacula, M. C. and Kubala, J. J. 1975. Statistical models for shelf
factor based on a test at 40◦ C and a usage temperature life failures. Journal of Food Science, 40, 404–409.
at 20◦ C: Garvin, D. A. 1987. Competing on the eight dimensions of
quality. Harvard Business Review, 65(6), 101–109.

1 1 Gillette, M. H. and Beckley, J. H. 1992. In-Plant Sensory
AF = exp 6500 − = 4.13 Quality Assurance. Paper presented at the Annual Meeting,
293 313 Institute of Food Technologists, New Orleans, LA, June,
1992.
Suppose we determine that at accelerated tempera- Giminez, A., Ares, G. and Gambaro, A. 2008. Survival analy-
ture of 40◦ C, we have a failure time of 35 days. Then sis to estimate sensory shelf life using acceptability scores.
we can find the failure time, FTu , at usage temperature Journal of Sensory Studies, 23, 571–582.
Tu , we merely multiply by the acceleration factor Giminez, A., Varela, P., Salvador, A., Ares, G., Fiszman, S. and
Garitta, L. 2007. Shelf life estimation of brown pan bread:
A consumer approach. Food Quality and Preference, 18,
FTu = FTtest (AF) = 35(4.13) = 145 196–204.
Goldwyn, C. and Lawless, H. 1991. How to taste wine. ASTM
Thus our accelerated test predicts a (mean) fail- Standardization News, 19(3), 32–27.
Hammond, E., Dunkley, W., Bodyfelt, F., Larmond, E, and
ure time at 145 days for the product stored at room
Lindsay, R. 1986. Report of the committee on sensory data
temperature or about 20◦ C. to the journal management committee. Journal of Dairy
Science, 69, 298.
Hodgson, R. T. 2008. An examination of judge reliability at a
major U.S. wine competition. Journal of Wine Economics, 3,
References 105–113.
Hough, G. 2010. Sensory Shelf Life Estimation of Food
Products. CRC Press, Boca Raton, FL.
Amerine, M. R. and Roessler, E. B. 1981. Wines, Their Hough, G., Langohr, K., Gomez, G. and Curia, A. 2003. Survival
Sensory Evaluation, Second Edition. W.H. Freeman, San analysis applied to sensory shelf life of foods. Journal of
Francisco, CA. Food Science, 68, 359–362.
Amerine, M. R., Pangborn, R. M. and Roessler, E. B. 1965. International Olive Oil Council. 2007. Sensory analysis of olive
Principles of Sensory Evaluation of Foods. Academic, oil. Method for the organoleptic assessment of virgin olive
New York, NY. oil. http://www.internationaloliveoil.org/.
Aust, L. B., Gacula, M. C., Beard, S. A. and Washam, R. W. Kilcast, D. 2000. Sensory evaluation methods for shelf-life
1985. Degree of difference test method in sensory evaluation assessment. In: D. Kilcast and P. Subramaniam (eds.), The
of heterogeneous product types. Journal of Food Science, 50, Stability and Shelf-Life of Food. CRC/Woodhead, Boca
511–513. Raton, FL, pp. 79–105.
Bauman, H. E. and Taubert, C. 1984. Why quality assur- Lawless, H. T. 1995. Dimensions of quality: A critique. Food
ance is necessary and important to plant management. Food Quality and Preference, 6, 191–196.
Technology, 38(4), 101–102. Lawless, H. T. and Claassen, M. R. 1993. Validity of descrip-
Beckley, J. P. and Kroll, D. R. 1996. Searching for sensory tive and defect-oriented terminology systems for sensory
research excellence. Food Technology, 50(2), 61–63. analysis of fluid milk. Journal of Food Science, 58,
Bodyfelt, F. W., Tobias, J. and Trout, G. M. 1988. 108–112, 119.
Sensory Evaluation of Dairy Products. Van Nostrand/AVI, Mastrian, L. K. 1985. The sensory evaluation program within
New York, NY. a small processing operation. Food Technology, 39(11),
Bressan, L. P. and Behling, R. W. 1977. The selection and train- 127–129.
ing of judges for discrimination testing. Food Technology, McBride, R. L. and Hall, C. 1979. Cheese grading versus con-
31, 62–67. sumer acceptability: An inevitable discrepancy. Australian
Cardello, A. V. 1995. Food quality: Relativity, context and Journal of Dairy Technology, June, 66–68.
consumer expectations. Food Quality and Preference, 6, McNutt, K. 1988. Consumer attitudes and the quality control
163–170. function. Food Technology, 42(12), 97, 98, 108.
Mizrahi, S. 2000. Accelerated shelf-life tests. In: D. Kilcast and Rutenbeck, S. K. 1985. Initiating an in-plant quality con-
P. Subramaniam (eds.), The Stability and Shelf-life of Foods. trol/sensory evaluation program. Food Technology, 39(11),
CRC /Woodhead, Boca Raton, FL, pp. 107–142. 124–126.
Moskowitz, H. R. 1995.Food Quality: conceptual and sensory Regenstein, J. M. 1983. What is fish quality? Infofish, June,
aspects. Food Quality and Preference, 6, 157–162. 23–28.
Muñoz, A. M., Civille, G. V. and Carr, B. T. 1992. Sensory Sidel, J. L., Stone, H. and Bloomquist, J. 1981. Use and misuse
Evaluation in Quality Control. Van Nostrand Reinhold, of sensory evaluation in research and quality control. Journal
New York, NY. of Dairy Science, 64, 2292–2302.
Nakayama, M. and Wessman, C. 1979. Application of sensory Solomon, G. E. A. 1990. The psychology of novice and
evaluation to the routine maintenance of product quality. expert wine talk. American Journal of Psychology, 103,
Food Technology, 33(9), 38, 39 ,44. 495–517.
Nelson, L. 1984. The Shewart control chart-tests for special Stevenson, S. G.,Vaisey-Genser, M. and Eskin, N. A. M. 1984.
causes. Journal of Quality Technology, 16, 237–239. Quality control in the use of deep frying oils. Journal of the
Nelson, J. and Trout, G. M. 1964. Judging Dairy Products. AVI, American Oil Chemist’s Society, 61, 1102–1108.
Westport, CT. Stouffer, J. C. 1985. Coordinating sensory evaluation in a multi-
O’Mahony, M. 1981. Our-industry today—psychophysical plant operation. Food Technology, 39(11), 134–135.
aspects of sensory analysis of dairy products: A critique. Thompson, R. H. and Pearson, A. M. 1977. Quantitative deter-
Journal of Dairy Science, 62, 1954–1962. mination of 5 Androst-16-en-3-one by gas chromatography-
Ough, C. S. and Baker, G. A. 1961. Small panel sensory mass spectrometry and its relationship to sex odor intensity
evaluations of wines by scoring. Hilgardia, 30, 587–619. of pork. Journal of Agricultural and Food Chemistry, 25,
Pangborn, R. M. and Dunkley, W. L. 1964. Laboratory proce- 1241–1245.
dures for evaluating the sensory properties of milk. Dairy Trant, A. S., Pangborn, R. M. and Little, A. C. 1981. Potential
Science Abstracts, 26, 55–62. fallacy of correlating hedonic responses with physical
Pecore, S., Stoer, N., Hooge, S., Holschuh, N., Hulting, F. and chemical measurements. Journal of Food Science, 46,
and Case, F. 2006. Degree of difference testing: A new 583–588.
approach incorporating control lot variability. Food Quality Wolfe, K. A. 1979. Use of reference standards for sensory
and Preference, 17, 552–555. evaluation of product quality. Food Technology, 33(9),
Peryam, D. R. 1964. Consumer preference evaluation of the 43–44.
storage stability of foods. Food Technology, 18, 214. York, R. K. 1995. Quality assessment in a regulatory environ-
Reece, R. N. 1979. A quality assurance perspective on sensory ment. Food Quality and Preference, 6, 137–141.
evaluation. Food Technology, 33(9), 37. Young, T. A., Pecore, S., Stoer, N., Hulting, F., Holschuh, N.
Robertson, G. L. 2006. Food Packaging, Principles and Practice, and Case, F. 2008. Incorporating test and control product
Second Edition. CRC/Taylor and Francis, Boca Raton, variability in degree of difference tests. Food Quality and
FL. Preference, 19, 734–736.
01234567
869
935 862495
!"#$%&! !&'"'#()##'"!&!"(#"*&)

!&#"+!,%$#&$'!-./"!0+#,!" !!&#"+ !&)!&,"#&*!& &'
"!1$"!*!&2!",&)*%! #&!%2&'"!&)!",&)"#!'$"!2
%&&&)-!"!#"!#&#&'!)&&)!3!"*!&&%$'&)'!)&&'
"!*!& "$$"!-4$/!1$!&%+2 !&#,!")!&!"%&!%"!!&&)2!%!5
&)2&'"&&)6!%%&#,!",!6#(&!%&!&,!-!%!)%"*7#&
&'"!1$"!*!&#($&)$*&$/8!#(!&#"+ !"!%#'!"/!'-
9%+!!"'$!'#%%!#&&'/$%#&-
:;<=;>?@=A=BC;DCC;<==EF=AG<=D?=ACBHIBDJ=GD?AKGDL?;A=CGC?BMCJG=D?GNJBFFA;BJ@O?;?@=GA
P;AQBA=?@B?RMBOS?@=A=GCD;A=BC;D?;CTFF;C=?@B??@=A=PGUUV=BVGBCWMVOSG?<=BDC<TJ@<;A=
P;AQWMJOS?@GDLC<GL@?L=?<GE=HTFS
X@=A=GCD;A=BC;D?;CTFF;C=?@B??@=A=PGUUD;?V=BVGBCSYCA=LBAHCMVOZ;D=<BKBCQZ[<;A=
?@BDP@B?\]>;A?@B?BIBUGH=EF=AG<=D??BQ=C<;A=P;AQ?@BDBDGDIBUGH=EF=AG<=D?GCGAA=U=IBD??;
B<BDP@;GCPBD?GDL?;<BQ=IBUGHGD>=A=DJ=CSYCA=LBAHCMJOZ;D=>==UCCK<FB?@KZVT?G>BD
=EF=AG<=D?=AGCD^?PGUUGDL?;H;BH=J=D?_;VP@KH;=CD^?@=J@;;C=C;<=;?@=A=BCG=APBK;>
=BADGDLBUGIGDLS
à"#6&%!!bcdef2-cg
0
45
4 ~-e-~ &!%!"$*!& ----------
~-e- &!%4!%!#&&'4"!!&&) ----
r
r
hij klmnopml----------------- ~-e-e "&&)#(&!%---------- rq
his tuvwvlmxtvplyzl{pml|vl ------
qr ~-e- &!%!"(#"*&!!*!& ---- rq
~--c ,%$#&"!------------
q} hi topmllnlxp ----------- rq
~-- %*!#&"#%------------
q ~--c &"+4+!* ---------- rq
hih tvmmmmlpnvpml---------
s hir mlopml ----------------- r
~-~-c 4*%!4!",&)"#!'$"!-------
h
h vvvlv -------------------- r
~-~- 4*%!4!-------------- h
~-~-~ 4*%!4!",&)!*!"$"!------
~-~- 4!",&)#&&!" ---------- 7
4649
~-~-e ""!"---------------- q
~-~- %!%!&&)------------ q
~-~-f 46%%#6&)&'3!#"#& ----- .&%!"!"#(!3/##6!6%%#(!&!
~-~- .&"$#&#&!% -------- "$%"*!#'#$%'/!!"(#"*!'$&)&5
~-~-d &'#*#&&'a%&'9/!%&) ---
hi zvp|vlvpyl------------- '"'!&#"+"!-!"6%%'!"/!6
~--c !)&&)4$'+----------- 6!*!&/+&'"'!&#"+"!-/%!~-c
~-- !)&&'"!*!&4"$$"! ---- "#,'!!% #(*&+#( ! )##'"!
hiq lvpmlpnvpml ----------- rs )$'!%&!'$!'& !"/%!&/!
~-e-c .&!&,!--------------- rs
~-e- !#($*&4$/8! -------- rh $!'/+!&#"+!%#!&$"! !$'+
--96%!2-!+*&&2:=DC;AK¡IBUTB?G;D;>¢;;H2£##'4!&!!34!"!2
¤.c¥-c¥¥f¦df5c5cd55e§~2¨4"&)!"4!&!©a$&!ª!'299¥c¥
01 2 34567589
34757
!"!#$%&' <%$$)&'%!%&&H'!&%')$2'&'GD<+H'

(&)*!&%+ 9,%!2 !%&%&<%$$!!'%'$$ )'#
(&&,
-'$%&
&"&'2'2,'!&%!4/%2$%GI")'#
.!/%&0& &"/$'$<'"'&)3/$$'<'3&"!:
1&"23!&'!& F/!H&"%#&'%$2H'2<"&"&%$$!'
/,+% ',,+'$ 4&+'$%22'&'30&"&/2G
!%4
5302!&
5!&%+ JKLMNOPQNRSMTUMVSPOMWMR
('%%4
'0,$
%6'2"', 1/!"3&"%30'&%%&"%!&%!030
7$/0 /9,%!%+%%&%4H2%4%4H'2,'&%4
8'% 3'!%$%&%)&"%%2/&%'$'2%/%+%&
+%4&0,'&/
1'9%0/0"$2%4&%0 &&%4G("!&%<''$<%&&<%&"3!
(&&/, &;0%H-'4)H'2.$AXYZ[BH\$$%#
-'$%&!"!#:% AXY][BH^44&'2_#Aàa]BH&'2%2$
-'$'&!$'
5&/!&%
AàabBH'21%$4''2&'$GAàaZBGI3$&"'&':
(&!"%!%' ,$'%4!&/!&%4+'&%4'
(,'$%& 3'!%$%& "/$2'2^44&'2_#Aàa]B'2+%<
!"& &"'!!0,'%48@H&"%%'9&0$ +'$/')$
5&/!&% /!G(" 3'!%$%& "/$2)$!'&2!$
(,3!'$ &,&&%'$*/24)/&&%&"0%22$3''
;&&%)/&<2 <%&"9&'/2'2=%G("%0'&"'&
;!"<2
82%4
%'0'&:,!%4,$'&&" ''"/$2&
.'20%6'&%=!/&)'$'!%4 )'&"0#"/'2%'<% &"3'!%$%&
>&"%&0 "/$2)/&3'"&30&"%3&")&&$%4
-!%$ $%G(" )&"''0/&)'%$ '!!%:
?',#% )$&&",'$%&'2%3&"3'!%$%& <%$$)/2)
,%&!/, !/0,'$%&,'$%&&"'&<%$$)&'+$%4
8$'/, 02%&'!&"&""/$2)'0,$H' ,'#:
%4'+'%$')$G("%3F/&$ 0'&"'&&"
@%,'$''40&A%0,&'&%3!/%&%#B
.!%,&%3%!&%+%0&'
-'$%&2)%C4 3'!%$%&"/$2)&"4/2c3')/%$2%4'2
(&'' &"'&&"''"/$2)'&"&'!&&"!0:
,''&%3,'$%& ,$9G5!0,'%<%&"!/%&!!H&"
(0,'&/ ,,''&%3'!%$%& "/$2)<%&"%&"!/''
D/0%2%& )/&&",'$%&<'%&%40'2,%)$&"
E%4"&!2%&% )&"''"/$2)%'''&"'&%'%$'!!%)$
?%A'/2%&B '2,%)$&!/G
I"2%4%4&" &&%4''H&"&'3:
>'!#4/22=!$''%"'2$%4=,%&%+,/
;!!%)%$%&
!/%& C!,'&&3&",'$%&"/$2)#,&%0%2G
-'$%&"/$2&'29%&&"3'!%$%& <%&"/&
'("%!"!#$%&%'F/%!#<' 30'#%4/&"'&&"
,'%4&"/4"&",,''&%''&"3C!''
,!%'$%&"'&"/4"&30' 3&"42,'!&%!4/%2$%
2%!/2%&"%!"',& 3&"3'!%$%&G("%%&,+&,'$%&30"'+%4
,"%!'$+%/'$'!!&%30'&%&"'&0' )%'
&"%,Gd9'0,$H%3,'$%&"',,&
)&"/4"&&"/4"G5&"/$2)00)2&"'&' 00,& *'3',!%C!)'2%'&'"!'%&
42,!%'$%&<%$$'$<'3$$<&"&'2'2 0' )%'&"%,%3&" 9,!&&+'$/'&&"'&
,'!&%!)!'/&"'&</$2"$,/&"'&"=" )'2'3&"%!22'0,$G;22%&%'$$H3
012 34565789
358 77
9757
!"#$ K&&&#$&
%!!!%!!& $%!!& &$$)$
#'#(&)#*!#$+ %!#(&# !&#$+
>!"(!!#)$
,-.-/01234256789:;2 C!"$)#!$@(
&@!!)&@+G !%!$#
"")$&$$+<(
<&#$(&!($%$ "$= $##&#&'!!
+>!&)&#$$&! %$$$%!)$ $!=#A=
$ ) %!)$&#)!#$ )$)@!%"@L!#DM((!
)$+<$%&#&&)!( (&$"$+
!"$?@# & <%!"$$
!$'$! (+<#$$ #(!#"$#& ?$
&#!!#$A!!+<( $(!#&#&"$
&#)!"$)$!#$= !$ )+<&($@!
!$ !"$!)$& "$!$&# @
!!## (!!+B&&$ )!@(?$@%&(!
$+CDEEFG &%!)$! #$CH+I+MG+
%!!#$C!!$G$ >!%!$ !"&()$@
A!!+<($$#)$@#! )%$$$@ $+>!(!#$N
#$%!#)$#$% )!CH+I+D O!($ !$&'!&
(&(G+>!)& ($!!##($ %$$=
J#"$ %$$$$%#$)# P+>!!$ )&$ !%
+ &(&$#+>!#$
QRSTUTV L
(&(#)$
)!+
01 2 34567589
34757
!"
!""#"#$%&&&"#&!"
'(")*+&,"#$
-.&*&/!"0&+"#&$
1&&2+&#13.#4
567&#"&"!
&%.*"#/"*.
+"#859""&
"#:&!.%;&/
%&&&"#8-6&##"#$
(!":&!9<&*&+&*.
(!#8-9":#""
#"#$(!#"#7"%!
:&!#&=(!#81&#!"
#"#$:&&/":&!
%!&#7("#7*##8
2&*&%"#!":;#%"#
&*&+&*."3$""#*
#.*"#8>&#%.#3
:&!:?:$3&4@0#"*
:&!"3&##&3AB
5%!&"%#CCD7D*7
E"/F4
#!.*:$#$3&&3&G"!":&&/&3"? J3$#"#$%&&&"# !"(!"&# !"

"&##(.#3"&3"#!&#&#.+&*("4H"&"+"!" !"!""&4H!&#"#!.*("&#"*
"*&.3:&&/ !""#!.*(""I.&"*:&! 3 !""&"7("%3("(.
#3"&/! "*&/4J#3"%&&&"#7%!&*%" %#.&"%"4H!""#!.*:$#("%"
""&##K%!&*"3$#("&%.*"*4J *"##&;&/4E%"/&7".?
!&#%#"7%"3.#(";" "+" !"&#"* *&#%&/%#"*+&#("4H!"3#!.*("
*&#%&3 !&#"3 &""&/:&! !" ;"I.&" %&&""&#%%"&4H!"""
"&##K%%"&*.&/*.%"+.&4 (($#3$+"#&#(!"##!"""
J#3"%#.3""#&/%&&&"#7(&"L/" #"#$%&&&"#4,3"!"+&&#"%#3"&%
3$("*M%" !":&&/3&"&# *!"#"% !".%&&$!"#%"4J!&#
3$("*"& !":&&/3 &#"4H!"&"? #"%&:":&*"#%&("#3"+&&#7!&/!&/!&/
&"&#+"$.#".&%!&#& !"3 " *+/"#**&#*+/"#"%!4P":&%?
/"*&:##"3&?%&%"4H!"&#.%&#7 %""(!#.#"**"+.&#8!:"+"7
# %"*."#7%("/&+" :!"/. #"%&&G"*(!#"""I.&"*!""+.?
%"("" !"$""!" "# (!#*&#%.##& &"#%"*.%##.%!##!+&/%"3#7
34N."#&#%("L"*"**+.""? ##7*"*#*!3"%"*.%#&;"&#"%?
"&##!+&/*&L%.&"#%(".!"&#.%"* &%&*"#7Q:="#7**""/"#45"=3"
:""*"*.4 .#"?(.&#"#$(!&#'".3%3?
H!"*&#%.##&3 :.*.#.$("/"* $K#$?#"#&(!*"#%&("*($>"*##&
#&3&$ %""%"3(. !"*"%* !" "4@9RRSF4
.&#!&/##!.*("#&3"*&%#!:.* H!".3("(!#&%&&&"#:"!+"#""
"% !""&##K%%"&4H!""#!.* /"#3#:#T #!&/!#9U4H!".3("
(""#&$%%"##&("!""&##*!""? &#.#.$%#&"*($!"#%"+&("4O:"+"7
&"4O:"+"7!""&###!.*!+"+&#. !"#"#$#%&"&##!.*"3 !+"!"3=&?
!$#&%%%"##!""&"4H!"#"%&# 3.3##&(".3("(!#%#.%"*7#&%"!"
%&3"%7&/!&/7"%47!"(!"" (!+&(&&$&#"I."$(""%;&"#+?
"I.$&%("!"*&#%.##&"4 .3"%.#&/.*."*"$"&##*"%"#&/
012 34565789
358 77
9757
!""###$$% !!*$)!"#!!!'$
&"!'"()*)!$!#!!!" !"")"#!!""$"!
!$ !"+ )+ !"!'%, "#3"(!*""%4$2$"!*
")-!" .#$/ "#!"!
$!"#$!)###!%0 "(!*1"!")%
$12#""!() *"3""#%4 4$(*!"$2(!")""
""$"$#)5 !""(!*L!
$!(!$"2$67#)$$* $!"$(*!"#-"#"#
#$+(#%4!" !$"(!*#%4"(9
!"(!""!$8#"#9 !*#"$$N"3!#
!*#:"##!%4#!$!$ 13!*"")"#!!""!)"!$
""$3!$*!"" ")"!$%
#)(!$% 4"$5!$3!#!# 9
;$$!!)1!<!$,"1!" ""3!#!'$$)")"""
"$)$!"$""*!$!"#!$ 3 "#!#!#")$$!
3!$"1"!*#=*!"1$#!*" (!"#!=#$#%>)")""
;!#"3!>!"!!!";#+??7@AB 3! $!"#""$ !!"#%4!"!
<,C+BDE+B+7+F+BB+GH"$3$% "!-"!"$$!"#*$%4"
4#!*!"")!$"- "!-"5)8"#$#!9
!*@ID#H-!##!*@?B#H% !!$()$!=#!!
4!*#!"#"!$) !#)"!)"!%K!""$!"9
!*"(!*#"!$ ""!"$3*""""!-"%
""9*#%J(""3 J"(!*#!"#"$1!""$
"(!*#3"-!##!*$ (""#!#!*3!"
#3"!*%49 "(!*""%4!$#!#! !-!"
! ""3""3""$!* !*$393)"!*")"%K"!"#"1
-!##3#"$$!"9 #!#! !-#"!#3
#*)" #"#!*!") "$!%K#""1
%K*1!#!*!""$%4 #$"!!##$"!#!""$1
!*#"3!"""!!#9 !""!*"!**" "$
#!"$$""")"#!!"" "(!*#*!!"
$""""*"(!*#"%4 "(!*""%
$!*!"!!!'$3#!" O!"3! $!"$$#
-!##!*!"""$ !"($!!##!!)%P21
"!$""$8"!*% #!!)$"!*$"$2#"!()'
,(!*#""$**##9 $"""3$($("13$
$")"1"#""1$)" $'"#%0$1#!9
*!!!'"(!9 !)$"!*$(!"3$$'
!L"(!*%4#"A6# $!*"#"3 "("1"("1$
3!$$A7# !*M3(12#"!'!" !#""$#-%P"""
$$"!'""(!*)" !!""3$!=#*!()""
3$"$!#!!)%4""(!* !#"$$!!1
#"!"!$!*$")$ "!#"$"3$5!$!
2")%4"!$!*$")"$! ) !"%
"!$""!$%4"$"($(* 4$"*$ "*"#%
)$##)"#!! Q!*$"*!"#"")""19
"(!*#%48$!"$(*" #"$$"1$$"@!#!("H9
$"!"!""#"* !""%P'"*"#!"$$""
"#!!%4(!" 5!'!*%;$$!!)1#!"*"#
01 2 34567589
34757
! "# !""&"(,!""!!"

$$"$%"&'$$""" 7&"(2!$"$
"&"""#$$!&$" !""&"''."'!$$"-
&()"#$$""""!"""&*"" """(,!#&!'"*!"
"$"&(+#"!#$&"!" "#&" $$!%!"&-
"#$!"#"&# """7"(,!""
"'$"$"&( !%"""&"$%(
,!!""!"""!!- E &!!"/"#"!
."$"& $#(,! "" '"!$(2"#!
&$"&!%"!"/!$- "%#"%#'&!"&"#'((.-
&"$/0/! "&$& $'!%
!"!"''(,!$"&&"%- ''.(
&"#"&"&&" ,!'$"""!'I#!%!
"""! $"&$ "'$( "&""!%FGHFFJ"KGHKKL
,!""!%&&/!'"" "!'#(,!&/'"*!
!"""#&""'""(1!/"! '&'"%!$""/$-
*/! "%"$"""!&"!% !''% "&%#!'$"
"!$$""""(,!!"% !!'#(
"""&"/"$$#&" $$ +'"!""!%""
"/""$$#""/" MGGHKGG7"!"%"&(+"#!%
%%#!$" %/"'- &"%/!"''/&!"'"7''
$(1%/"%/"'" NGGHOGG7!"'"#3&(
$"%"" "#$(2"# +&"&!"'3"%%#&!" !
$ !#$$&%! %%""%#"/ &%!
"$$"&&!"&&" "&*$" !!#"!!&(P/!"
'&/"$-"#'"#"% %$&"%"$%'"!%"&&
(+"&*$""!"" /! !%!""(,!! !%
!!/!&!"&"3 "!"/!&"&(,!"
&' !&* ""45( $&"! '&"""
,!$%"$$"&"'(+ !/%&J!"$QF(
'"&67%#!"%! ,!"%&/""""#
$$""""/!'"%&" 67% "&#(P/'$-
&&""/"!*$"!$" &"'$&"."&!%&-
/"$$"&"'"$$"&!"" &(R7"'$!"&#%
"$&3&(2 "$"&" "'%!!'"-
$&".$'&!"&&*" ""&"/!!!&"&(
$'%&/! ""&#( !!"'"#"'&"&#
+"&"""% &!% ""!%&"(
&'"*!"$*43 2 #"'&"&#""/&""
"#5/""#/""!" "&'(+""&'" ""'$
""!$$""""( 6/&#!!%"!"'$
%#'7 /!-"(S"/
""!"'$""7$ ""&'"*
89:9:;<=>?@A;BC@DB< $&"#3 $4,"*"QTOT5(
2' &"%%#'!"$"-
'"#(,!""""&-
,!%!"&""!%&'"&- ""'""'(,!"'
"(,!$"&"%"& !!&!"'%' !7"-
&"%3!"#'&$$# '(,!""!%&&
010 23456785898
84 375 84 0
!"# '!+!"^!V!
$%!" '!,''+'
&'!'()' !' ))V'(("&'!(
!!)*)))( '!,'!'(V''(
)+,$()" ''()'!(')++%(
''_"
Z+!'%!'
-.-/012345256578591:;04<2:591 '()+('+V$
V'V$')V'+!
=>=>?@ABCDE@EFGHIJKFLMENOFEP $'()`' (+'!
!*)')!$!
("Z))''()))
&''() +( V''$'()()V'''
))*$)()! !'())V''V!
'Q(%' RS()("T !'())V'''
%!()('+++ '()'!('V!
)$'))'UV))) V''!%)VV)
)$"&V'V$) )'V!!'())"
'WY)')))($''! ^'+V$'+'(V
X! (''($'' )()!+'($$!)'
Z'["\'V Y) )(V% +
'('(V%X! "&''' $"
V)'( )$'
*"\'(V)$
!(''())+ =>=>a@ABCDE@HbE
'))! ("&''
!V'VV)'!'
V$))'($"]V ^+'!())!)
'()('')!!)" '+'!' ()
&'('))$'+$ (''!'())"^+
"&'')))'(+! ))*'%
'Z+$)'+!'[ Q(!)$)!()R#
'"&'V+$ )$'' )#)cddeS"]V+''
!($"&'%!$'! '!!!''!V
V)'V'!(' )')!
$'!())+'))!"&' "
!''))$('! f))#$RXghgS+()'!
+$Q('!()++" *++)'$)%(
^)++()$''))! (V''V
)!$'(+'! (V+'!*)++"ijW,"
$$'V)! RcddYS_!)'(V')++)("
()!! &'(!, !''
)"#!(),('! _'!*)'()
'V)++)'!!( ()'%')++!
)++" *!'))++("&'+
^+!+)(Q() V'))$'!*'
'!$+'!())*)"T '(),+VQ(!)!
%!+!,(),+' j''(+'()k]V$'
!(+!V($)$!( !*+')(k]V$
01 2 34567589
34757
!"#$%& U&&&&'&%&&&&
&&&'&&'&!( &&$&&&&&
%&&$&&)&&* %&&&&-&&.V*
!+&#&&&,&&& &&'&*%&&&'-&&&
&$&&&&&&* %&&0&#&&&*
-&)&&&%&&'&./&& -&&&1#&&%1
%&&&-$&&& &NL1#&%1#-%&1&-&*
-&&)&-&&.(& 1&&)&1&.+&&&&&&
&%&&&$%&&0&. &'&*%&&&&%& &L&
+&$ &&#&& &&&'-&&&&0&&*
&$%&%-'& &&#&&&'&.T$1&&
."#&'&1&%&%&%&#&& &$&&-&&&0&
&#&1-&1& &&&.
&&&)&%&&. (&&%&&&E&&&1
&&&&1
#0&#&F&&J&PK&&&
2323245678949:;<=>?9679:5@A:9B &-&& -&&&$
%&&.( -& &
&& %&&&E&&&.
+&&'-&&&&%&&* "#&'&1&%&&&&'&&*
0&&&.C&'-&&& &&$&&&$&
-& &&0&#&* &%-#
&.D&&'& &&&$&&.
&& &&&&'&#&.
(%&&1% &&
'&0E&&&&F$GHIJK&&.
C-%&L&& %&&* 2323W49:;<=>XY=@5<=9:B
-&&%&&&%&&
&&%&&&.M'&#
-&&)&&&* (0-'&-&& &
&&&&'&$&& &&&&&E&&
&. &$&.(&&1&E&'&
($NL1&$ * & -#$&1&*
&$%& &&&#& && &#&#-%&
&&&%'&&-&.(&& &.(&&1&&$%&0
#&&&'$&&$ &'& &&&&
&1&& %&'-&&* %&&.+&%&'&&
&#-%&&I#&.+ &&#&&-# &&.+&&*
N L %&&'&OGJ&&& $& && &
PJ&&&&&'&N'. '&&1$&& &&&-*
I
(& &%&&&&QOGIJQORIJ '&$&&&$%&&.V
&OS%&&&'-&-0 &E&1C$ &'&$ '&&&
&& & &.T-&&&&& &&$&%& &$%&%&&
& &#&.($%& -&&&&LL&L*%&1
& &&$&&&)&&&$%&& $&#&'&&& & '&&$
&'-&&&&& &&N' &%&'&-&.(
&&)&.(& &&&&* &&*- &&&'L-&1 &
&&%&&&&& %&L&&&L&
&$&&. .
010 23456785898
84 375 84
+'!&.%.13(%",.'!&!"$#$
$',&$'(&,)',(%"-$+$())%!('1<&!*
'($*(-+(./(+,$!.!$$$+!$#$(&.
!"#$%!%&!$$! '(%%%)!!* #$!$$'(%%%"$*(&.%!%')%*$1
)%!+,*"!%$&!%-)'(,$(&..%/'(%0 '(&+/%-&"!%*($/(+!$$?!"$)%'/0
",'!&.%($!&123(%%%4 (,,-%"%$!*($%(, (+,.""%&'(#,,($&($%!"$&$%('$!&
$($"!%*(+(!%/',"!%$"!!.+&5$$.6 +$#&$'(%%%(&.$"!!.$!+$$.1
+$*(-*!%+%!(.,-+'!&.%.((&-!$%"!!. F$!&(&.F.,7IJJC85/$"!,,!#&5&$%$0
$($(''!*)(&$!&+&5$$.!$($$-(% &5:(*),!"('!*)%!*&$ !"('(%%%6
&5$.7(&.$($.8$!5$%19:(*),(%'%(*;,,0 #%$"!!.)%!.'$7()??(('8&A&'.
&5&)($%6+$$%!&+%(.6)'&(('6(&. +-$'(%%%7'%$8&'*)!%$(&$#(-$($$
(,(..%&5!&,$$',(/1 )%)(%($!&6+$&!$$$$&56(.$!&/!,/$+0
<&.'%*&($!&$$&56$5!(,!"$&$!*(=( $%($G2KKK$#((5%.$($A(/!%&$%('$!&#$
$$$($#,,+/%-&$/$!)%!.'$.""%&'1 '%$'!*)!&&$%,$.&(('A(/!%$($'!,.
>'(%%%'(&*(=!%.5.""%&'!%*&0 &!$+('/.+-($&5$('(,!&1L!#/%6
*?)(&,$@(+,$$!)%'/$.""%&'. !#&5$!/(%(+,$-&)??('%$#$&(+%(&.6$
$!$(..$!&!"!$%A(/!%(&.*!.;'($!&$!$ #(.$%*&.$($$)??(#!,.+'!!=.(&.
$:$%(&.*!$",'(%('$%$'1<&!*'( $(''%()."%!*$'%$(&.$$.+-$,"1
$'(%%%*(-*),-&'%($!/%(,,'!*),:0 $'*'(,%('$!&#%(,,!#.$!!''%(&.
$-!"$&!%-*)%!&$!$)!&$#%$ $+E'$@%)!&#%&!$&A&'.+-'%$
)(&,$(%!/%#,*.1<&$'($ !" $'=&!%!$%&!&$$/(%(+,14
('(%%%*5$&!$+.%(+,&'$#,,.'%(
$""'$/&$/$-!"$$$"!%.$'$&5&!%-
.""%&'1<"$%(%%!'!&B&'"%!*
*&5(.""%&'7-)<<%%!%63()$%C8 MNOPOQQR
$&$ !"('(%%%$($'!,.)!$&$(,,-*(=
.""%&'&!$%'!**&..1
<"!&$!$%(&.6("(,(,(%*!%.$'$!&!"( 5!(,!")(,($',(&%!,.+$!(.&$
"(,)!$/.""%&'7-)<%%!%63()$%C8 %>&( *!/(,!"%.(,*($%(,"%!*)%/!(*),1
)!%!)%!+,*6$&$!+'%&5!"(.""%0 $($&%'/.!&$5%"%B&$,-$!,.($#&$($&5/&$(-
(%%!($+",'#,,,) &.!$
&'+-$'(%%%,.$%*&$(,1.5%!" ""'$!"5$
%(,*(...+-$'(%%%*(-'!*),'($$$(0 !*&'(* &&&&%.#&(*),1*(=
'(,,-1)%!$&!"$*($(&.
$!&6+$$'!,.)%/&$$.$'$!&!"(.""%&' $E''!,."!%*(' !*),:%*!/&5$(&&&"%!*
$($*5$+*(&&5,$!'!&*%1&0 !,$!&S%.'&5$2
!%-)'(,$!,..'#$$',&$#$% (/+&&*%!$).'! =%4!"$#&1%
$.5%!"%(,*&$$$('!&'%&1D!%( %*!/%.#&($%&5&'-7&) (,($',(&%$!
"!!.)%!.'$$($%(%,-'!&*.(,!&(&.(,*!$ (%,($/,-%'&$:(*),81L!#/% T! $(,16IJJU"!%
(,#(-&/!,/('(%%%6$2(%$;'(,$-4!"$$0 $($&!$%'!&&(+&%('6.! $#!,.*
($!&#%$'(%%%!*$$.*(-+(*(E!% )(,($',(&%$! &$'!&.$!&1 &$.(,
)-'!,!5'(,)%!+,*$!$)(&,$6)'(,,-& V '(=(&.W,#'7IJJX8/(,($.$""'$
'!&*%$$&51>&:(*),#!,.+'%%-) !"(%(&5!")(,($',(&%7'!'!,($6)'$&!,0
;,,&5#'(%%(%,-($&#$!$)'%$1
F!6$%(%$#!5.,&"!%'!&.%($!&& $# !&6$(+,#($%'%('=%6#(%* #($%6#($%6(&.
.$%*&&5#$%('(%%%!,.+ .G$%,0 *!!,$* ,=8!&"!!.%)%&$&5/(%!$($(&.
($/'!&B&'!"*&5(.""%&'/%( 7+$$%86*!"=".
", '$'(E,,-+(&7#$86'!""
"(,)!$/$$%,$(&.$.5%!"%(,*$($ $!%$,,(')7)&5(&$(8657 "($$-86$(7($%&5&$86)'-
*&$7'!!,&586(&.()),('
.*.&'(%-1H"$&$'!*),'($!&'%($. 7&!&0,&5%&581-" !&.$($$(+,#($%'%('=0
+- !"$'(%%%#,,,!#%$.5%!"(*), %#%$!&,-)(,($'
'!&$%!,(&. &"!%*$-$($)!+,6!$* $ %)%&$($/"!!.1 ,(&%""'$/('%!(,,
00 1 234564789
236469
! "#$" !" ""!"*!"+("$!"%"
%&"'"&!"(!"%&")!"'&" !""34!""%!&*"'"""*!
$"!&*"&+%$'%"(!"!+ )$""$&(!"")!"
$+$$,-".!+/00123 '`"%"$'&&(!"$3a"!""6"&
4!"))""!""" !"&!" *""$N"!("&'"&$"$
%&"""*"("""$'&"34!(&$5" "(!""$"*"$'&"+%&"
""!!" !"!""%"*"!" %34!&&%%*"%"(""
%$'6"37(""+!""! $&(!!"")$"!&)!"
"!"%%"!!(&&"$" $"!"%!"&&!""34!""
"#'"")& !$'&"'&" '"%&!&&(*"("!'"&
%&""3 '*&"$3
4!"%"%!%!&&
8989:;<=>>?<@AB=ACDEFGHI?J=I@?A *" "%&"'""*&!&*"("3K
"&!"!""%!%"'"!""#'&
!"'%""!""'"%&34!(&&"
K$&%&""+(&&()" !!""(""%$$%)'*"(""!"
$'&"""#'"%"34!$"' "'"%&!"'"'&"'"$)!"3
"&"%"(""%""' &&+$""&""&'
%!""#3&(&&()!)!'%% '")'%""5""'!&*&"
"%"%&"'"&L"3M%"+ &*"*53
%$"")(!""%%"'*&*")$"
"+(&&()')"""%%"% 8989bc=AC?d@e=I@?A=ACf>@ACg=hG>@AB
%$"L'!"'%3&%
$""))""&N)!"&%$' i$'&"!&*"*&&*"&"(!$!""
%%"!""!)"'"(&&() )%"*$'&""!&*"
!"'%%%"'*&"3O"&&7"$,/0P02
"!"""%(&&()""#'"% $ N"%""'""
!"!&)"""%'"%$' '%$3 4 *&"j3U)""'*"'%"
'"""%""4*&"j3j
)&"")"&5"*" "!"$"
"$&55$$&534!")% X)"j3j+j3W+j 3
)+5)+!"%"3
k+j31+j3l!($"!""
""%37(""+!&*""!!"'(" '"'"%%) !"%4*&"j3U
!""(&(!!"&5"&! ) j3j3
""%"(&)!3$"Q""&
R"*$"*S&T"&3,UVVW2
!(&&()""#'"%"%!" mnopqrstuvswxyz{s|u}w
'"%""*"""!"*!
')%""!"""%!""3 4!%!'"")"*"%$'"!"
M")"(&&()&%&"
")!"$&""%"'!" *" %"#'"$"&")3~#%"&&"
!34!%*"$'$"'% &") 5 %!'"!"*""(""#'"$"
"$3X"#$'&"+!*$' S#,/0k("( &""!"""S%!
'"''"$'&"Y!%%!Z,!"'" .!,/l02+ PW
2+
%&i)!,/0PW2+&&5"
%X",/0P12+""",/0Pk2+
%!"$%&2%!%"%%!%&"3 7",/0012+ %&,/00l23
8989[\A]IJ^HI@?A]I?_=AG>@]I] 899G]@BA@AB=;I^C
4!""!&*""%&"%%"3K"6"& K!"%("(!)!&)!$"$`"
"*&")"!"%!('"$!" !!&*"5"'$*!""'"%&
""&*!"*&&+*""!"'"& (!""))"#'"$"3!"*"))
012 34567896
68

!"#$"%&'#($")(%"#$*&+#"'%#,#$-#($-$+%)%$'""
./0%-%,-"%"12"1'(,3+,-"%"1"#$4#*"/5/565/75/86-$+5/9:/
-/4#33#$1"-,3#+$#;'-#($-(/4(%-#%++#))%$'(%-#%+%)%$'""<('(3&,$""1(&3+ ;33+#$
2=6>:/4(%'($"-$%)%$'+&(%#(""1%'(3&,$""1(&3+ #$+#'-+2?)%$'=6=6-$+>:/4(% -3-$'+
%)%$'+&(%#("")(&%'(3&,$"-%$++2?)%$'=6?)%$'>6=6-$+>:/@%#-$*3""-3"($+)(&%
'(3&,$";33+#$2=6=6>6-$+>:/A$3!1%"-%'1%""1(&3+B$(<1#+$#!()1=-$+>/
/4#33#$C&+*$&, %"2#//6.6D65EEE:/=""#*$-'1C&+*-$&, %-$+,-B"&%1--B!(1"$&, %"#"#$'3&++
#$1$(((B-""('#-+<#1"&+!/F#""#,3%#)-"'#;'C&+*%-#$"1-$&, %1%(&*1(&1"&+!/
'/G%-%,&-#($"()"-,3%"$-#($/4(%-#%++#))%$'(%-#%+%)%$'""1%-%<((""#3
%,&-#($"2=>6>=:/4(%1 -3-$'+%)%$'+&(%#(""1%-%)(&%(""#3%,&-#($"2?= =>6?= >=6?>
=>6?>>=:/-$+1'($"-$%)%$'+&(%#(""1-"<((""#3%,&-#($"2?= =>6?= >=:/4(%%#-$*3""
1%-%"#H(""#3%,&-#($"2==>6=>=6>==6>>=6>=>6=>>:/I-'1"%J#$*(%+%"1(&3+ -""#*$+-
$&, %/
+/K%,#$(%+%()"-,3%"$-#($/L"#$*--3()%-$+(,%,&-#($"6$&, %"-%%-+)%(,(( ((,
<#1#$-'(3&,$/L"($3!$&, %"'(%%"($+#$*(1$&, %()"%J#$*(%+%"#$1"/M%#1$&, %2<#1-
%+$6#$+#'-+#$NOPQ($4#*"/5/565/765/865/9-$+5/R:#$- 3-$B'(3&,$-$+1$<%#1(%+%1-1"-,3"
<#33--%($1"%J#$*%-!#$1&%%#*11-$+'(%$%()-'1"S&-%($1,-"%"1/@1#"#$+#'-"1(%+%#$
<1#'1-'1"-,3#"%"$+(-'1C&+*/
/=""#*$1%+#*#%-$+(,'(+$&, %"(-'1"-,3)(%-'1C&+*/-%)%(,-$!(#$($1-3()%-$+(,
$&, %"-$+&"1%+#*#")(%-'1$&, %/K($(&"$&, %"1-,-!1-J,-$#$*(1C&+*"2#//6.569996
T..:/M%#1%-$+(,$&, %"($1,-"%"16($)(%-'1"-,3)(%-'1C&+*2&" 3&(% 3-'B$6#$+#'-+
#$UVWXUYZ#$4#*"/5/565/765/865/9-$+5/R:/=$(''-"#($-3+&3#'-()-$&, %,-! )(&$+($-%-$+(,$&, %-36
#)"(6"B#1+&3#'-$&, %/
D/M%#%-$+(,$&, %'(+"($1"-,3'($-#$%"/L"1%-$+(,'(+$&, %"<1#'1<%<%#$($1,-"%"1/
G(+$&, %"($"-,3'($-#$%""1(&3+,-'11-%(%#-'(+$&, %"($1,-"%"1/@1"-,3'($-#$%"
( ;33+<#11%)%$'"-,3""1(&3+$('(++?= (%?> &"1(&3+ '(++($3!<#1-$?/
5/0%-%"'(%"1/4#33#$1+-61C&+*$&, %6-$+1%-$+(,'(+$&, %"#$1"S&$'#$<1#'11"-,3"-%
( J-3&-+2-"#$+#'-+ !1%-$+(,%,&-#($":/
7/0%-%"-,3"/
-/0%-%-$(%*-$#[+-%%-$*,$)(%(%#($#$*"-,3"/="#,3,1(+#"(,-B-,-"%"1,3-<#1
"&);'#$"-')(%1"-,3'($-#$%"( 3-'+#$1"S&-%"/@1#",3-,-! ,-+(&()-$!3-%*-%(%
-J-#3-3"&"#&/=33(<#$*-5#$/"S&-%)(%-'1"-,3#""&**"+61(<J%61#"<#33J-%!+$+#$*($1"-,3
'($-#$%#"3)/
/="", 3"-,3'($-#$%"($,3-/A$'-331'($-#$%"-%3-'+($1,3-#"1(&3+ #+$#'-3#$
--%-$'(1,-"%"1/
'/0(%#($"-,3"#$('($-#$%"/
8/="", 3"-,3")(%-'1C&+*($-%-!#$1"S&$'1-1!-%( J-3&-+/=3"(63-'1"'(%"1($1%-!
-$+<-%)(%%#$"#$*1-3-/K(&3'1'B"%J#$*(%+%/
9/%J"-,3"(C&+*")(%J-3&-#($"/
R/K'(+"'(%"1($1,-"%"1/G#%'31'(+1-1C&+*'#%'3+/2L"-$6\]^]_&"$'#3($,-"%"1"(%
"'(%"1":/F$1#"<-!6+'(+#$*#""#,3-$+(%+%3!/F$(%+%(-$-3![1+--6#,&"%%"$+$&,%#'-33!/
@1#",-! -''(%+#$*(1$&, %()'(%%'C&+*,$"2-#%++#))%$'"6%#-$*3"6-$++&(%#(":(%$&, %()
C&+*"%)%%#$*"-,3=(%>2-#%+%)%$':/`-B"&%1--'(3&,$#"3))(%1#"&%("/
a/=$-3![1+--/
()-$!%(C'1"$"(%!"'#-3#"-$+-331-% H%'#"#"#$"%&'#J '-&"1%('""-33(<"1

#"1--%%S&"#$*1"&+!"1(&3++;$1 "'#-3#"('3-%3!#$+#'-1b#+-3/c@1$<1$
(C'#J()1"&+!/@($"&%1--33-%#"-% #,-$+'("'($"%-#$"-%-+++-$+1"'#-3
'3-%3!'(,,&$#'-#$*61"$"(%!"'#-3#""1(&3+ #"1-"(%+"#*$1"&+!(-"'-3++(<$J%"#($
%1%-"-331(C'#J"-"S&"#($"/@1""1(&3+ 1%#"-'3-%#'&%()<1-#"b*#J$&c#$1#"
'#%'&3-+-,($*-33-%#"6<1("1(&3+%(J#+ %('""/F$"(,"#&-#($"1"'-3++(<$J%"#($
)+-'B(1"$"(%!"'#-3#"/@1"$"(%!"'#-3 ,-!$( '--3()-$"<%#$*1"(C'#J"/
#"6#$'($"&3-#($<#11'3#$2":6"1(&3+#+$#)! M1$1#"1-$"61"'#-3#"-$+1'3#$2":
1""%S&#%+(-$"<%1S&"#($"/=1#"(#$ ,&"%$*(#-1'("-$+#,'($"%-#$"-$+d(%
##"&"&-33!#$"%&'#J)(%1"'#-3#"(+"#*$1 1"(C'#J"/F#"&"&-33! %(+'%-"1
%)'H%#,$<#1(&-$!'("'($"%-#$"/@1#" $&, %()"(C'#J"(1("<#111#*1"
01 2 34567589
34757
!"#$%!&" '&#!"($#)"*!"(+#)!"()",-,&"!%

./0#)#1)#-23!(/4/5'&#%&16,1)#-7/
)/3!66!" )16!,"!8%)!&")&/9"-:)16+!"3!(/4/5'&#) $,&'8-+-!1)%#&,+;&,+<$")+=)*/
>"6-#)#%-# -&$6,*"&?-!,"!&'-#&,$%&#)16/
/3!66!"@$,("$1#2!//+.+A+4BBB7/C !(")%-@$,()"$1#)",1)* $#-))*&-"$1#!6)%,!"
- $,"&&&*/9! !16#!') %!8%@$,(#)!" -)"$1#-#&$(-&$- $,/
%/C !("-#,!(!#)",&1%&,"$1#&)%- )16'&#)%-@$,(/)#'#&1)"&!"&"-)6&'#)",&1
"$1#)",$-#,!(!'&#)%-"$1#/DE#$"$1#-)1)-)E1)"!"(&-@$,( 2!//+.4+FFF+G..7/
H#!-#)",&1"$1#&"-1)#-+&"'&#)%- )16'&#)%-@$,(2$6$&#6)%*"+!",!%),!"
IJKLIMN!"3!(/4/O7/C"&%%)!&")6,$6!%)&')"$1#1)'&$",&")#)",&1"$1#)6/9'&+*!-,$6!%)
"$1#/
,/P#1!"&#,#&')16#")!&"/Q !"())6&'#)",&1#1$)!&"+"$1#)##),'#&1&&&&1
?!-!")%&6$1"/Q &"6"$1#%&##&",!"(&-"$1#&')16 !"(,2!//+'&#'&$#)16R$&"6
"$1#.+A+4+)",ST#),-"$1#!"-&#,#-))#7/H#!-"$1#2?!-)#,"+!",!%),!"&6,!"
3!(/4/57!"-$##!(--)",%&#"#&')%- U$)#&"-1)#-/<-!!",!%) -&#,#!"?-!%-)%- )16!
#",&)%-@$,(/9"-:)16+-8#)16! #E,'&$#-+- %&", )16! #E,8#+%/'&#@$,(./
A/H#!-#)",&1%&, &"- )16%&")!"#/Q -#)",&1%&,"$1#?-!%-)#?#!"&"-1)#-/;&,
"$1#&" )16%&")!"# -&$6,1)%--)#&#!)%&,"$1#&"-1)#-/9'-#)#"&$(-&6
?&#*!"(&(-#+-!%)",&") #)",&1"$1#)##%&#,,&"-1)#-/
4/0#)# %&# -/3!66!"-,)+-@$,("$1#+)",-#)",&1%&,"$1#!"- U$"%!"?-!%-- )16 )#
&E)6$),2) !",!%),#)",&1#1$)!&"7/
S/0#)# )16/
)/0#)#)"&#()"!V,)##)"(1"'&#&#!&"!"( )16/C !161-&,!&1)*)1)#-16)?!-
$'8%!")%'&#- )16%&")!"#&6)%,!"- U$)#/<-!16)1)1),&$&')"6)#()#&#
)E)!6)6 $!$/C66&?!"()4!"/U$)#'&#)%- )16! $((,+-&?E#+-!?!66E)#,",!"(&"- )16
%&")!"#!6'/
/C 16 )16%&")!"#&"16)/>"%)66-%&")!"#)#6)%,&"-16)!-&$6,!,"!%)6&-
1)#-/
%/0&#!&" )16 !"&%&")!"#/
W/C 16 )16 '&#)%-@$,(&")#)!"- U$"%-)-)#&E)6$),/C6&+6)%- %&# -&"-#)
)",?)#'&##!"!"(-)6)/P&$6%-%* #E!"(&#,#/
F/#E )16 &@$,( '&#E)6$)!&"/
5/P%&, %&# -&"-1)#-/H-"@$,( )#)*,&#)&"6&")#!$)6)"*%&6$1"!6'?"
%&6$1" &'#)",&1%&,"$1#/H-")*,&#)1&#-)"&"#11&#6)"*%&6$1" 2&"%&6$1"'&#)%-#1
#),7-&$6,6'/<-%&6$1" #&E!, )%'&##%&#,!"(@$,( %&# )'#%&16!&"&'-/2Q )"+XYZY[
$"%!6&"1)#-&#%&# -7/9"-!?)+,%&,!"(! !16)",&#,#6/
O/C")6V-,))/
#!&#!! #)-#-)"%$!"(-&?#&'-2 1!-)",6,/0)"6! 1),#&&$'&#%&16

C",!:\'&#&?#!$7/<-#!"&&!"!"# !"(-U$"%/0)#!%!)"1)"&%&##%6
'&#1!"() $,-)! !"),U$)!")"?#!"(- '&66&?-#&&%&6&#-1)1!$",#)",
1)@&#&@%!E/9'!! "&& !6&,!(" !"#$%!&"/<%-"!%)6#&""6%)"1)*1!)*
)"),U$) $,+- %!)6!1$)*'&#1&# !" #E!"(&#,# U$"%/Q"?)",]$%$)!&" !"
#&$#%/ )161#)$#&#&-#%&",!!&" 1)"#-
D:+- "&# %!)6! -&$6,1!%$6&$6 !%$#/^&&'-#&61 %)"6!1!"),&#
%#$!"!V- $, /<-!,)! &)* 1!"!1!V,!")?66,!(",/
U$!&" ))%-&!")&$-?&#& !6 % >"%-$,-)"#,!(",!!)(&&,!,)
")#!&)",-&?- $,%&$6,!1#&E,&1!" &?#!,&?")_*6&"` )!!%)6)")6!/<-!
!1!V-%&"!"("%!/"&# $,! )#1&# ?!66(!E- %!)6!)(&&,!,))&$-,(#
%&16:-)"-))#)8#(6)"%)",-&" &''#,&1) &%!),?!- !("!8%)"%/9!)6&
!)6'&#%&16!%)!&" )",1!)* !)6?) #"/ -6'$6& *%-&$&"!)68($# )",)6 -)
)16 1)6&+%&")1!"),+&#&-#?! ?!66$,!"-8")6#&#/
012 34567896
68

ÿÿÿ!"ÿÿ #$ÿ%ÿ&ÿ ÿÿÿ!"ÿÿÿÿÿÿÿÿÿÿÿÿÿÿO#ÿ#$ÿ%ÿ&ÿ
ÿÿ ÿ ÿ $#ÿ%ÿ'ÿ ÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿO#ÿ$#ÿ%ÿ'ÿ
ÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿO$ÿ#$ÿ%ÿ4ÿ
ÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿO$ÿ$#ÿ%ÿ3ÿ
()* +ÿ # $
!
()* +ÿ O# O$ # $
& ÿ'ÿ &,-. /012 !
' ÿ&ÿ '&'2 &-3. & 3 O2 //1z 3Pÿ/ÿÿÿÿ.
4 ÿ&ÿ -,ÿ'ÿÿÿ2 ,33. ' & O2 ,4,. ,P3z
3 ÿ'ÿ '14. ,-42 4 ÿ'ÿ O2 4'/z '4ÿÿPÿÿÿ.
67+ 3 ÿ4ÿ O2 30&. &/-z
7++
89:;<;< =>?@ABCDE?@?FGCHFICCGEDH?A?JHCKAHCECHCLMCGCFGN
89:;<;{ =>?@ABCDE?@?FGCHFICCGEDH?j?B?LMCKHCECHCLMC
ÿÿÿ!"ÿÿÿÿO#ÿ#$ÿ%ÿ&ÿ KRDSGHJDGCFGN
ÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿO#ÿ$#ÿ%ÿ'ÿ
ÿÿÿ!"ÿÿÿÿ$##ÿ%ÿ&ÿÿÿÿÿÿ$$#ÿ%ÿ3ÿ
()* +ÿ O# # $ ÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿ#$#ÿ%ÿ'ÿÿÿÿÿÿ$#$ÿ%ÿ/ÿ
! ÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿÿ##$ÿ%ÿ4ÿÿÿÿÿÿ#$$ÿ%ÿ,ÿ
& & O //12 3Pÿ/ÿÿÿÿ.
' ' O ,4,. ,P32 ()* +ÿ # # $ $
!
4 ÿ&ÿ O 4'/2 '4ÿÿPÿÿÿ. & / 3-/. -',2 &Pÿ4ÿÿÿÿz
3 ÿ'ÿ O 30&. &/-2 ' 3 '-'z P--2 P/3.
7++ 4 ÿ'ÿ 1-12 ,40ÿÿÿÿÿz 4&/.
3 ÿ4ÿ PPP2 /,,. -P&z
89:;<;Q =>?@ABCDE?@?FGCHFICCGEDH?MDLFG?LGHCECHCLMC / & ',1. /4&z 3,-2
KRDSGHJDGCFGN
, , '0&2 '4-. P'ÿ1ÿÿÿÿz
TUVUWXYZ[\]^]_àY^bcY]bdbaefbeaYZ 7++
gCBJhCGDRFCKCFMHJAGJDLDEGICC>ACHJ@CLG?BKCFJiL
CBRMJK?GCKjklJBBJhCL?LKmDILFDLnopqrsNtICFC
?RGIDHFKJuJKCKC>ACHJ@CLG?BKCFJiLJLGDGvDj?FJM 89:;<;| =>?@ABCDE?@?FGCHFICCGEDH?GHJ?LiBCGCFGN
FGHRMGRHCFwL?@CBkGHC?G@CLGFGHRMGRHC?LKKCFJiL
FGHRMGRHCNtICkKCFMHJjCGICGHC?G@CLGFGHRMGRHC?FFCG JFKCxLCKjkFCLFDHkFACMJ?BJFGFvICLGICkiHDRA
DEF?@ABCFDHGHC?G@CLGFGI?GGICMBJCLGnFsFCBCMGCK C>ACHJ@CLG?BRLJGFJLGDjBDMhFNtICFCGvDFGHRMGRHCF
GDFGRKkJLGICFACMJxMAHDyCMGNtICKCFJiLFGHRMGRHC ?HCBJLhCKjkGICH?LKD@J}?GJDLACHEDH@CKjkGIC
01 2 34567589
34757
JE@@>C>FBGC?N<A?@<>?<=>>Q:=N:BEFG>:KD<C?JNKBL
OD><C?JNKBGC?N<@?C;A:l=?KILrF>9:;<=>j?N=J
l>:ABNJMjEBDBDC>>Q>CAE?FA?@:<C?JNKBLOD>B?B:=
! "#"$ %&ÿ'ÿÿÿÿ) *#'+ "#', K?FAN;>CGC?N<EAJEQEJ>JEFB?BDC>>ANlGC?N<AjEBD
<>?<=>C:FJ?;=M:AAEGF>JB?:GC?N<L8:KDGC?NABA
& %"# --. -*.
, + ) '.'$ ?F><C?JNKBHBD>Fq==A?NB:RN>ABE?FF:EC>LsNABEqK:BE?F
@?C;?F:JEKJ>AEGFA:CEA>AjD>FtuvBD>B>ABj?N=J
' %"/) '*ÿ'ÿÿÿÿ, %"'$ &"#+ l>B??BE;>K?FAN;EFG?C=>FGBDMB?D:Q>:==<>?P
<=>>Q:=N:B>:==<C?JNKBAwtxvBD>NA>?@?F><C?JNKB
/ ."#+ *%'$ *.*, %%!) j?N=Jl>=EI>=MB?EFyN>FK>?<EFE?F?@:F?BD>Cw?C
011 tzvBD>NA>?@BD><C?JNKBKD:FG>ABD>>FQEC?F;>FBH
<>CA?FH?CANlABC:B>LOD>=:AB>@@>KBEAK?;;?FjEBD
K?FAN;>C<C?JNKBAt>LGLHy??Cj:9HEFA>KBEKEJ>v:FJ
<>CA?F:=K:C><C?JNKBAt>LGLHAIEFKC>:;HD:ECK?FJEP
2345657 89:;<=>?@:;:AB>CAD>>B@?C:C:BEFGHC:FIEFG?@ BE?F>CvLOE;><C>AANC>B?K?;<=>B>:B>AB;EGDB:=A?
D>J?FEKB>ABL JEKB:B>:;?F:JEKJ>AEGFEFK?FAN;>Cq>=Jj?CIL
{EBDBC:EF>J:F:=MBEK:=<:F>=AHBD>A:;<=>AAD?N=J
A>FA?CMA<>KE:=EAB<CE?CB?BD>ABNJM:FJB?G>BD>CBD>M |NJGQ>B
l >> :=N:B>JEFC><=EK:B>t?@B>FBCE<=EK:B>vB?:AK>CB:EF
;:I>N<BD>>9<>CE;>FB:=J>AEGF?@BD>ABNJMLOD> EAAN@qK?| NJG>Q:CE:BE?FLh@BD>FN;l>C?@A:;<=>A
A>FA?CMA<>KE:=EABAD?N=J=>BBC>:B;>FBABCNKBNC>JEKP <:F>=EAB>Q:B==MA
E
>F
N :B
;:==EB;:Ml><?AAEl=>B?D:Q>>:KD
>:==A:;<=>AEFC><=EK:B>EF:AEFG=>
B:B>F>EBD>C:<??CJ>AEGFABCNKBNC>F?C:@:Q?CEB>?C A>AAE?FNAEFGmnoL }?j>Q>CHBDEAEA?@B>FF?B<?AAEl=>
@C>RN>FB=MNA>JJ>AEGFABCNKBNC>:@@>KBBD>A>=>KBE?F?@ :FJBD>FBD>A>FA?CMA KE>FBEABj?N=JNA>:C:FJ?;Ek>J
BC>:B;>FBAL K?;<=>B>l=?KIJ>AEGFtnm~ovL
65S5T5UVWX34YZ[\]^[]\WX b6`5c
_`89

97d
5e6fb_dg
_`89

9ab6`5c
d
5e6f_bdg hF:C:FJ?;Ek>JK?;<=>B>l=?KIJ>AEGF>:KDBC>:B;>FB
hFBDEAJ>AEGF:==BD>A:;<=>A:C>C:FJ?;=M:AAEGF>J ttNNAANN::====MA:;<=>AvEAC:FJ?;=M:AAEGF>JB?>:KDNFEB
B?:==BD><:F>=EABALi?AB?@BD>>9<>CE;>FB:=J>AEGFA ODEAJ>AEG:FEM< F>=EABAvjEBDEF>:KDl=?KIt?@B>FA>AAE?FAvL
:AA?KE:B>JjEBDA>FA?CMABNJE>A:C><>C@?C;>JB?:Q?EJ <:F>=EABAK:FFA@ C>RN>FB=MNA>JjD>FBC:EF>J:F:=MBEK:=
?C;EFE;Ek>:CBE@:KBAJN>B??CJ>C?@A:;<=><C>A>FB:P AEFG=>A>AAE?FL?hB> Q:=N:B>:==A:;<=>AEFC><=EK:B>EF:
BE?FLOD>AE;<=>ABA?=NBE?FB?BDEA<C?l=>;EAB?;:I> >:KD<:F>=EAB>Q:=ND:EBAK
FB :A>BD>l>ABA?=NBE?FEAB?D:Q>
ANC>BD:BBD>A:;<=><C>A>FB:BE?F?CJ>CEAK?;<=>B>=M :FJBD>FD:Q>BD>;C>BN=C=A >: :;<=>AEF:AEFG=>A>AAE?F
C:FJ?;Ek>J:KC?AA:==<:F>=EABALODEAB>KDFERN>j?CIA C>P>Q:=N:B>:==BD>A:;<=>AF@ L
?C:ANlA>RN>FBA>AAE?FB?
rF> 9:;<=>EA:J>AKCE<BEQ>
RNEB>j>==EFAEBN:BE?FAjD>C>BD>FN;l>C?@A:;<=>AEA :F:=MAEAABNJM?@AE9EK>KC>:;A;: J>jEBD@:BC><=:KP
A;:==:FJ:==A:;<=>AK:Fl>>Q:=N:B>JlM:==<:F>=EABA >CALhF:AEFG=>A>AAE?FBD><:F>=EABAK
EF:AEFG=>A>AAE?FLmnoEABD>EJ>:=J>AEGF@?C:K>FBC:= AE9A:;<=>AL}?j>Q>CHBD>A:;<=>AAD?:NF? F=M>Q:=N:B>
=?K:BE?FK?FAN;>CB>ABjD>C>>:KD<:F>=EAB>Q:=N:B>A EFBCE<=EK:B>LOD><:F>=EABA;NAB:BB>FJBDC>>> =Jl Q:=N:B>J
>A>AAE?FA
>:KDA:;<=>Lp?C>9:;<=>HEF:;:==EFB>CK><BHB>AB<:FP B?>Q:=N:B>:==BD>A:;<=>AEFBCE<=EK:B>LhFBDEAA BNJM
>=EABA:C>:AI>JB?>9<C>AABD>ECJ>GC>>?@=EIEFG@?C BD>l=?KIA:C>BD>A>AAE?FA:FJBD>AE9A:;<=>A:
>:KD?@@?NCK?=:<C?JNKBAL8:KD<:F>=EABC>K>EQ>ABD> C:FJ?;Ek>J:KC?AABD?A><:F>=EABAjEBDEF>:KDl=?KILC>
@?NCK?=:AEF:C:FJ?;=M:AAEGF>JA>RN>FK>L
mnoJ>AEGFA:=A?EFK=NJ>C:FJ?; :AAEGF;>FB?@
<C?JNKBAB?<>?<=>jD>C>>:KDEFJEQEJN:=?F=MA>>A 67`89

97d
5e6
?F><C?JNKBLOD>A>A?PK:==>JK?FAN;>C;?F:JEKB>ABA
:C>K?;;?FEFK?FAN;>Cq>=JABNJE>ALOD>A>:C>:=A? hFK?;<=>B>l=?KIJ>AEGFA:C>NA>JjD>FBD>C>:C>B??
K:==>Jl>Bj>>FPGC?N<AK?;<:CEA?FAAEFK>BD>C>:C> ;:FMBC>:B;>FBAEFBD>>9<>CE;>FB@?CBD><:F>=EABAB?
012 34567896
68

!"#$%&'( '.''%&'(.-''-"(
"'(')'!'*+ ."''*.'(!.'(.#'
)%,(!"')-!*."(' '("(*'(."'..'9*+
'''-)'*'."' *.'9-..'.-)'%?.05
.'-!'(''-!'*+ '.'.''('0!)5'(.-
%/0*'(1.''-*"'!"# "'"')''('('!)+
'(".')-*23) ."'%,('.'''."'.*.
.*.!-*'('(. 4-52667$% '('-'('99-9'(''*(''.'+
,('"')'23 '*.'''9'.'.$%,9+9-'.'+
%8"("'!"# ''."'..#9*"'.'.''
9'(*..!"# $23!"# '."'.%
$ 22%::5;"(.;0526<=$%/'
'(*'('-'('()'
"(*'(23)*.'%,("'- R\67V7696
W7XYX76
*"'!"#*'".
'9(.'(.'".>)..*.+ ]-'(.!'.'''."'.0'.
.""'*..*''>). .-'%^0"*."'
.*.."%?.05'(.-!'9'-(' *"'.'."'.5+9-'."'.*"'.
!*.."*.9>.905!'' "!9'('9+9-*"'...''.'+
'(*''.*'(*..""."+ ''."'.%_'+'.'.0.+
'.''(.#*..'(*.*.." '.".'*. "'!"#
%@-"('(..'"' '."'.*"'...''.''
!"# 22%3A5;"(.;0526<=$ '."'.9'('9..'-*'.''%&
B3.*".9)'*.*.+ '+'0.''(..'9
."".'''"#'('# `*0.'''('.''"!
*.."% '**.'-`0.''!-.+
`'%/09!'(*9a"(
CDEDFDFGHIJKLIMKNKHOPKOHIQ *0).'*''.9'(..95.+
-51b'(''.(.)'
R
6STUV7696
W7XYX76 '(''*."(.9.#'.'
"'.%,(''.'("#5'(.
&'("'*'.'' $."( "!'"#(
'"(c%^"("'.'''(.
9'('.'('('.''% ,("#. -"#.".%
&.-''("".9(."'' -
''.)'!-".')
*.'0'.%&'("'(0.'
"(*.'(!.'(.#'%&'( '')')
"'(.''(."'!- ''')'' .'-'(.9'(0.+
"-.'''('*.'(."-%?. ]#d((":ee#7$%.".'(!'"(
05'-*'(.-"(."'.'"* /.'..'
!"#'5'(.-"'-"(*.!"# '+''.*'('9`*
'5!-"(**.'-#9"+ 0.''!''()*'''.'+
%,(+9-'.'''."'.*.'('- '-'$"'!.-%?.
'((*.'('.'.''"('(. 05!.""(.)'.-
-9-0"''(''(-.'!.*'% '!"#*.**.'"#'.+
%,("#.'..-".')
VZ[STUV7696
W7XYX76 )''(-'."').
:+9#.%&'("'(0.+
?.'9+9-'.'''."'.'*. '''("#'-'('(.'
".'!-"!)*'9**.''-* -$%
01 2 34567589
34757
!"# $MK&$%7

$$!%&%! $&' ""$%M%&7
$()#*+,-./0123&*3!
,-.45126,-./417
89XY=>Z[B\D]@>\B?Z^=DB@>\
89:9;98<=>?@ABC=DB@> _àb̀cdefdghifj
$%$&"E7F*E7EG
*&%*G"*H*7I R&K %&%G$*G
$$G"$%*&$' $$G&G*K%%K$'
"&*2%"7J%* 7RG%&$* !%& $&
&$"&*&KL$&&&$ % G#* "
L$$"7*%&M&' %*$*"&"$ "M
&$"&*2$* %7TJ kU"
$"&KL$&%%7I L$G%*$*"*G
$G&$"&**%K 7I&*&*G**'
L$&&&$L$$"7I $$* %&%7l#2
%" $%&G**KL$& *$G2G%%&%$*"
G& GK&' K$&KG7IL$*&*
%G&$"&**"$ KK' %G , &**2!&%
L$& &%G"&*7I &*"12$%&%*
*2KG%%&**"GKG' $ "%*$ *$
%L$$",)&N 72-./.O K*%&2K$G
J#*)&N2-..417*%' $%&%K*7
&G%$%"G&G'K& S$*K%&%&'
"%!,)$*P$2-..-Q-..FO &% #&K7mGT#U
R&&-..EOJ*S*2-..-Q-..F17G &Kn$$ %%&'
%$T$&U &G'K %K$2"$$&
&2&$*&G' %&%7o"K$G2%%
K&%&7 %**G*n$"$ G2"$
V*H%*L$* K%**K*$GKK
&K*G"$%*$%' &&*#*$K$7
*&2%&M&G2M'%* I*2Gn$TG7U
&76*&&&$ %&%' %&#&K2%G *
%&*"G % * * $7I&$
%L$& %*7 #2GG2&
I*2M%%&K**G '%2%G
$"L$ %2G*$% %G&&7N%G2$*2&'
'$%7'&*M'%&L$ *&%&%G2
,%&G&$$*1* ' "&KK*2$&%K'
G& $*% &7 %$ %7I
V*H&% M %' &*&$$2&%'
L$$"*&&"G "G$%%$%G"
%*&%7S"$ $ #G"L$7
K*$G%M *"G $ 3 # &K &$* &#
%W & % * G T7U&K&&'"#
$*"K*GMK%2"$ $"# %G* %%$G&
M%*$G*% &G"&K&7
012 3456789
598 64
859 0
!(9%%!+**%
!"#$" %% !%%!%!!(9#
"&" /%(%!*%%!
!%!%% ' %!!(9%P!(QR%
%(!!#)! ST%0U"#VW#&
%*+*!, XYZY2%[+0S&XY\\2#
- #. /%!%*]
%#/%
%(%%*'#/% 0X25%!(9""!
*+*%* #
0!*1%*2% 012/%!(9%!%"% "
%"(# %#
)%", 0^2/%!(9%!((4*
#3% %(!%!#
+* 0Z2/%! %! !!!%
(!%%&! #
(#!, 0_2/%%%!%%*,
"!!4%!%", %!(9#
#5 ! 0\2/%%!!%%+%
&%!6!+ (7 !(9*%%! !*%
"#8!" % %%(;%
* %9(#/%!& %4"!%+*!%
%!% #
:!"*%%% 0`2.("&%%!!%%
!#.%*'% %! %%%
%9(+&, 6!*%!!!
"("9!& !#
"*(+!"%
%"9(#5%9( 5!& !+
% !%(, 7("% + #8/%!
;!!+ %+*%"!:%,
!%%!!+* !0%+(!
# %%2#5,
5"( &%U&
! ,#5% 4%!!(9"%!
%: 9(( !%$T0Q$T_\]XXà.`1bXc12#
(#5 /%!(9 (+%%
!%"&%!(!%*+ " %
+%(! (#<! (%%#d%+ (
*% *+,4, #$4&"
% (0* + !%!"
%"2# (%%"%"",
"!%%7 '
80eT.2%U#R*%
=>?>@ABCDEFGHIJKGLMCNOB +!%%!(
(+*&%!(
"!#.& %!(
%!(" *%%% "%+!%(
%#/%%! *%+#5!!(%&
01 2 34567589
34757
!" &(#)M'%(
##$!%$& (#(%#"
"%'#!'"!(' $'#)
$) M%'"(%$
*%%#+,! % ('"""":
$#$$-%$ $% $#&;$'
"" ("%$$./$ ")*'"(## #
,$-%.*$01210"! %#%#$
345467#0!34489)*$: (%&)X!
#;$)/!%( '"##%#&
"%"%"% %#( %
"%#$$-%2,! %#&)M&(%#!!
344<9) :" '"'#$"$%:
$)M(%#%$%
".$%
=>?>=@ABCDEFGHCIJKEGLCBG YZ"$%)[$%$%
$'%#'"%Q%$(
")
M("%$&$""
%$&N"%'$:
#$()* "%" \]^]_]àbcdefghijklmggnonpqghrh
$(%%$")O$%':
"$% M(%%%'%:
"$%%#$ " ''#'Q%$()*!
" '(%"-%)P (%%'"$%'$:
$ %(((N"%# 'Q%$")
'$(!$%(!()*: [&234s89%$%"$#
#!("%$$" N#%$),;$"
""' $""%") #'$;$!
O(!%%#+,! '""%":
"#"%Q%$./$,$-% ($$)M#(""%
*$01!("%( #''#)*'("#
&$%"$( %"!'('%$
#'%'""%" "%Q%$ %$'
$() "$% )
O(!"'$$#"%#
"%%
=>?>R@ABCDEFGSCDCIGETBABUSIJCCBEBV t& (#t"2'':
9"$%)*"&":
&('&(#
W%"$%%#( %( t"$%&%#
"$#%%#" '"$%)P%$
"%"#($()O(! &$$"%%%
("%(%(%$: !t!$'%%%'
('"")/!"% "$%)M("%%$
$((Q%% N'"#),$%:
'""),""( ( #($#&
%##! $": '(";$#"%%
)O!(#"" #)
012 34567489
4 479
!"#$!# "7 %)$6!#"$I#"
%&"&" "$$ "$!!"'P#,1+",
"'(")"*&"! #"$+"$!
+"!,&"#"$+- "$!""!"$$#,", "+,
&"!""!#""'. "%)"""&'.&$"
!"&#+ "!#%""- """""$) &"&
"$"+!%$$' &+"$$#,#$ "
/""&$&"! %)$ &+"#$ &"",&'
!,!"#$#,- ."$#+""""!"-
"""%!#$+!- ,'( !"&#"&
"'01"&,#!#$$+$$ 6$" !#&'&&
,$&+ ,$&"$1! %"%)%1!$"#!
&""$!,,# #,"$2!- "%"$"$"$#!
+)3"$43'4!" + !",'
$&"!"$+ !"##'.
""&#&"$"!"$
+ %+$"!""&
! &" !#$"$#$ "" 9:;:QD>@EF?GHDE=CB=R>@SETGGEGGRE@H
"&'
5!$"!,#$) &"!"$"#$+,
!&"##&"!"$%"- (
'+ "&"!& &
$&&" U#!#"""!&
&+"$"!""$!!"
"%&++&"$"
%,%'(!"#$"& ")""$$#,+"#"'.
,,%"$"#""!"# +"!"
$ "!""%$#"
%!!#'($!&#&"$%"- &" $ "+--"!!'&"!
%"$"$"$$$+&% "+"""

$I#-",'8" !&"-
"&"$,,"&
6""7%&""," "!'(!" "&",!
"$'
(&&""!&&",$ """""#
,+ "! " ","&"!
""# !"!"$#, ,"&)$!$ ,&#"&-
"$%)""$ !"- X""$V&JLWWLN8","VJ "'JLWWMN
LWWYN"$V!!
'8$+$#""+", "!+!" JLWWZN'(["K!),"&J "+""
"$"++ ,""$"!!#" &\#"V X%)J \VXN%N
+"#"&"!' !"&"+"&,"&
",&--#%""!)",'/%
+#K"LM'
9:;:;<=>?@?@ABCD>@EF?GHG
("&#",I#$$$ ]^_àbcdaefghahijhadklfl
")"$+ "!# $$'0&
$!+1+"$-$",!- 9:Q:mn>H>o@H=pqpGHERG
"JK"LMN'0&" $!&"
&&"",++$'!"
""$")"$"1 /$!"!"!"$%"
",JK"ON' !&#"#&$"" &"+
P#,","!"$!+ !&+ "$ "+""'!%"
" !"&#&")" ,,!&"&!# "+""
01 2 34567589
34757
!"#$ ABCDEFGHIJKEF

"%&"%''()")
*)("!)+)" 8(!!')"4
! ""!!(" "!'"((#L!
!# ''))"(&"!(/
"(0"'%'"
,#-( ()"!( )!%(%*(/
"(#.")"'!/ )"%&!
("!)" ()!())#
")"'"(( M"!)!*(&
)0"!(# ")0"&'"%"&')"/
1#2"'()"3!("4" "%)N""((('
"& /&'"( / #
("!(5!6 O!% ')"4
)!/*(&"&" !( ()"&4! )""(
((( " "0(!
'!(# "('(")""
7#8('&()"/ "!#
("!"( "(
"""!#8"( PQRQSQFGQJ
&( (!"!) "
)""(%()"3!( (!
# O("%L#O#%2'"%T#L#T"%U#T#,V>;#
9#-"!)" 2X"Y ) M"!U& W%O(%
+)&"&"!+: .(T"*)%Z #>#
"#,V[V#U2")\ "
'3'"" ) 2" ](M^T"#-X
'# Z(( "2" ](M^
;#-&' ()"3 "! .
(.&"T"#X8
]#_0 "2" "( T*T"#
$'%?Z#
.%`#a#b%2#.#,VV>#Z' &'
<=) "&'" "!")" /@&"#b"
<="!""!!( ."M "!M%,,%,c,[#
%`#O#,V;[#-)") +)"('#
'")')"'"( 8"d!Z"%,7%,cV#
<=""""&&"! Z&e%() O#%\3f3W"'%M#\""%?#1gg9#U
'()")") "
<5*!)'6= Y"(<hijklmnmopmnqjrsqs=#b"
M"!M%,V%,V7c1,g#
>#?&' Z"%O#t#M'"%T#O#,VuV#U ()3
)""(+"^'(#b"
<=("(!("()" ZM"%$#\#"!M%9%,c,u#
//'!(#-"/ $!%XY"Z *#
+%\#L#,V;[#U+)"(?'#
"(!")!( $ ^*%T#%U'%a#2"3%b#W#$%]#1gg7#
"")" -@ 3()"+"
<=""" (! ' U'"%]# #b" M"!M%,u%917c97;#
"!3 5(!%6#'#% " T(%?#b#,VV;#-"&
!:W"' (
) +)"(#-L*d""!%[7%9uVc;g>
<=()")"'"((!@+' U''""%Mb#v *`#1ggu#2!T4"(\
&"+)"('" M)-2
"!U& a""%MU#OM-L
V,7#O("M! "
"4"(")0!) -'L"%$Z*%2O#
0121314516 77
89

!""
# :: >9
8.77J7
)
;'9&

$" %&$!'$
9"('
9"
9 -
-

A
&99! )#;

&&
98

)""
9&
*'
"9+',

- M'9)
&

/707370D
.""
9/012340 <
= '-
$77:
9
9 9
"
99&
5'+ !-6720999'$
9- & ;9%9- *
9%

)"('!
$

-.
$' 89 6'&
)9
17D3744
5'+74:
"9;

)')"9

9

,572D:
! .')&8?"
$
9
8 <99=$*''+=
':

<
%I
>
)$>70.$"&
"9;
')& ,+8>
+ L
''
E48&()&
;''*)9%
"

9')9
"
'6'& "'9
'

&
9&9!
));
!

)9
73/
"
9
!
9&
%
6'&
)9

>9
8.718?"
$
9'
!@A<
= /3/7
,;8B
C'9;9
.')&:9
) -'-7/
&-!
#;

!
"
9
$
#

8'
;

.$9
$=-
<
9-
'
$
9
) $
9
8 M'9)
""/431
&

0/3/D
6
''
572D
)8;'9&8 A=-
) -'-71:
(! ;' 9!
$"#
9
8''>%

8

$
*9" 9 9
99*9

'9
=
-'!)+--
9
8!' @A<
= ,;8B
C'9;9
.')@&
E
'')8F >
)$>72+99!9-
&&
9 :9
)

8'
;
F.$9
$=-
&!
9
?"
99
) &&

9
9 <
9-
'
6'&
)9
/03DD
*
687'!89-''),
"*'
,

-A
E$$
'.!$#59 5 6#G70 .5
&9
9 @9
',
;
%F .""'

)9
9!%9-)!-'
8 =
-'!)02 ',

-
9-

'/7!
*'9
3 @<
%*)+.
9+9 :
'%-
688&()&;
8,;8 ,N *99
78;'9
"'9
'

%9-
"

99;
& +-
$
) &"
'9"
&$

"9;
"('!&

"9/3/ !
A.$'9;9
9 )6'&
)9

8
>6>721."
9&
?"
$
9'
!@A D//3D
!!996,B
C999'

8 ,

- 9
>
'690
)8;'99

8'
;
.""'

9 ""7372 =- 8 98'
;
.
$:
!+.
8
>6>5
-&&EF9-
''< ''9 =! 8 72 9 O9 '&9$
9

72:
!9*'
9-
&&
9&
&"

9# 6"P
;
% ;
9
)
+-
$'

70
9 (9#
)#;

&&
9-''9
96'& 0301

)9
073702 9
99
;E'F 9
9'70=-

'
!.7
&$

$
9&
) +
A='&L+$'*
&
9-
<
$*
!
"
'96'&
)9
43D '9)=*'
+9'+'9% <7

'! +;''
5 +F=1
) ' 5;
$
99!&(
L-!9:+
8;'9=
-
89-8 9+,+=)'H ""727372/
85"F,989 L
'!@< 8
>6>7D:
!!#
''
5. 6-:8720.')&
) $
9'&(9 -!-

&)#;

&&
9
:9A'7<9 ,
-' <
%I %-
')*
9&$"'
&$"$)*
9
9
8
''
5. 6-:80.')&
) M'9)
&

13/2
:9A'7
8 9+-"$H >''J+,+ L''$.. .' 577J7=-
Q

&
<
%I "

99&99-

)
$
9&*
;
!

+711>$=
9! 9-
+9,$
& 8 M'9)
&

/77374
>$*K
9&
9)8;'9&8 +-
$'
*'9707<9'.
$)&

<9'
,

-+'L-!9:+""7003701
01234567
89
699 249 5
49
! "! #!$%%!%&%%!$ '""!&

("% !)"""' $!(!)#(""(*%&)++"+!!"!'"!!,
-''("$")"*"'"!"+%)!"#'(")!.
(!/)#!0!/1%"/)#%))/)2.!.2/$"%%%"#")"#'
234)!1 ",%&"'5#"+!'"" "")"*")
)"%6*!%/'.5#"/7./)*".*!%)*#!8,2))!%%&/$"
*"1'")#!!+'""++"!+%(!$""!&"9 '+#'"
)#"):'(";) '"2((")<!+'"*!!=,>'""")+!"(%!
"!&)! ") '"%&!+'"")")#"),?%&/$"
)#'""")+!$.#((%""#!)$"%%!"!!
#"1+! '""("!!+'""#%!+"!&)!",
@ABCDEFBGHIJHCKLMAHJENAHOCHNJAHNMHDHPIMAEIQ
9:'%"R!%%"
0 45 4
f,;,f :!!n="n)"'"
-"("!!+!>" ,, _^
VWWXWYWZY[\WY],,,,,,,,,,,, ^_ bb \YvWw WXb{\bbZxZYv{WY]x\
STU
ST̀ Z
Y vwX\VWcc\\Y[\ ,,,,,,, _d
[ab\ZcVWWXWYWZY[\ ,,,,,,,, de t\c\\Y\ ,,,,,,,,,,,,,,,,,,,,
f,g,h i"):!(!>",,,,,,,, de __
f,g,g >1%">",,,,,,,,,,,,, dj
f,g,k #!0>!>" ,,,,,,,,,,,, dS 7 89
699 249 5
49
f,g,f .2%"l"m!"):'!"6.2m:8
n"'!) ,,,,,,,,,,,,,,, do
f,g,; 2.R!.2 ",,,,,,,,,,,,, do ! "'!#%)*"#")$'" '""!&
f,g,p q!1n"'!),,,,,,,,,,,, d^ ("%$!)"""$'"'"$!(%""
f,g,r >'"234!>= ,,,,, dd
f,g,s #%.q))>" ,,,,,,,,,, dd (""(*%&)++"" 62"""%,/hp;n"%1)
STj t\bu\vw\Y]xYvy\zY\\
ZcVWWXWYWZY[\,,,,,,,,,,,, dd "%,/gpi"&h;sq!")q)"%/gf8,-
STS VY{a\,,,,,,,,,,,,,,,, d_ "( !*%"+!$!(%" !*"'"%%&)++".
f,f,h 3!%*#!)>g*%",,,, d_ +!#%!*# +!'#! !(""l"'

f,f,g >'"2)|#"):'.q5#"6} 8>" ,,, _e )++""",i!)# )"l"%!(""<(%! '(!*%&
!i!(!! ,,,,,,,,,,,,, _e $' " '"&"+!#%"(!)# *&#1)++""
f,f,k >'"R!%*#!) '"~.>"
STo u\,,,,,,,,,,,,,,,,,,,, _` #"
1 ")"$'%"#%"!#%&!$1'"!.
f,;,h >'"i!$"!+'"q%>" ,,,, _` !)"")++""",m!"<(%"/""
f,;,g "(%! ,,,,,,,,,,,,, _S #+#"&$ !#*#" '""<("l"
f,;,k .(++" ,,,,,,,,,,, _^ l%%l!#") '"("#l%%""
$''"("l%%l!,!$"l"/'"&%!&
,>,?$%"/,"&/ECJHILGBKPBMHCHFHH/m!!)q"">"<q""/
-h,hrrs.h.ffh.pfss.;f/q(1"q""3#"n")/??:gh
01 2 345674849
49 549
FGHIJKLMNOPQMRSQTQUVWQNUILMWM
!"#$%
&' ( X+)#B!5 $ $ #)#
#)#$### +)#B!Y
$*)#*! ' !
+# $(
'!,'$#)
''
# $ Z[\[]^_àbcdefg_ah̀eijbhkh
!-) ). (
.##$/
#*#$'! +#$#$ /
0/$# #$#3#*
*1/2( (# 9(
#) 3#*#
#)!- l9!+
). .## ).(
$!- #( $!-$*
)#$$ ##$&$)
#12( (#3Y("?49
#))$ !- /##48/
'34567589! #$&#(
-)#$#' #$'$
)/#!- )#3*9
#$)'( *#$#!m
###))#$### /$*
#)!- $)39#
$)##':( :/(
%' ) )#$/):#
#34;9!-/ !
'#) ?))(
#)#!0/)# )#$ # &/)$
** +$--#*#$3# #3''9/!0/
"%<9! %#48## nn
,##$ l))#$
#$#!-)### )$##o!
) -)##$#'
&##$)#!=##$ 3##)###
*'#':##) #9#$'#(
34;9! )#$##*
+) :#) !+))#(
#)##''#/(/ $ ##$pqD!8!r)
/>(# / p/*'3
3?./5@AB9/#$'##$# #9)/#
3C*/5@AD9!-45/)E$( #$'#3
#$'#/7/ #9!+
!- ##' $#$' #
# !s##$(
)#! ##))#n##t-
012 3456789
7 83677
!"#$#%&#'&(%)*%+%,#-%!,--
/#+&0 .1#,)
.&# !"-- - %,)-%!,1# /#+&0--1# #23%,-*4)-%!, *!'#'%&%-
5((%- *%#,6& 78!,9 :;:<;=7!*:;=;=<9 .1!!-1+!-(%""*,- >3?
#+&
@#-)1%,6 .!,-#,-*"*,) B"C: :;= @#-)1 #+&-!*"*,) >3D
(4!A-*%!
=#&#,)(*"*,) B"C:;B"C= :;= @#-)1 #+&-!*"*,) >3D
(4!A-*%!
:=E B"C:;B"C= :7!*=9 @#-)1 #+&-!*"*,) >3D
F4#&-#,(#*( B"C:;B"C= :;= @#-)1'!-1#%* >3D
G!*)()1!%) H#%*()!+#*%!, 78!,9 :;= .1!! #+&I%-1+!- >3D
!")%J(#--*%'4-
?C:G. 78!,9 :;:<;= 7/#+9 >3?
KC:G. 78!,9 :>L:KL>;= < 7/#+9 >3K
F4#&#%* 78!,9 :;=#,(:;: .1!!:;=7(%""*,-#%*9>3D
/!*-%,6 I!!4-!"J$ 78!,9 :;:<;=;=<;=<< /!*-%,-!-I!6*!4 >3>M
N3OPQ#**%AR#&+4S 78!,9 :>L:N;=><L=N /!*-%,-!-I!6*!4 >3TM
U3,! /#+A(%""*,- 78!,9 H#%*0:;: !*:;= .1!!*!,0P/#+S 83:#
!*P(%""*,-S
7B!, :;,!-C: B"C: :!*= .1!!*!,0P:S!* 83:#
)1!%)9 P,!-C:S
#G!*-13,!--;#)*%-*%!,+#' -'#)1%,(%$%(4#&#,(-1*"!*-1*!'#'%&%-+#,!-'V4#&-!>3DW/.1#-*X
"!*"4*-1*(%)4 %!,!")*%-*%!,%,3,!-#2
Y /- %,)!,(4)-%,6#(%""*,)--
>W5'-#%, #+& #,()!,J*+--4*!;(-#%&;-%+-#'&;#,(#,&%-Z-*#%,%,67W6W;-*#%,%,6I%-1-1
*!) 9I%-1)&%,-W
DWF)%(--%,6)!,(%-%!, 7#+& %[;$!&4+;-+*#-4*;-)W9#,()&#*I%-1)&%,-W
?W\*%-%,-*4)-%!, -!-1#,&%-#,()!,-*4)-'#&&!-W
NWB)*4%-!-,-%#&#,&%-W
XW/)*,#,&%-"!*#)4%-W
]W*#%,-!(! )%J)(%""*,)--7)#,4)!&!*!*1# !*%2( #+&9W
TW/-4)!4,-*'#&#,)(!*(*W
OW: %6,*#,(!+-1*C(%6%-)!( #,(&#'&#+&)43&#-W
^W.!,(4)---W
>MW:,#&[*4&-W
>>W.!++4,%)#-*4&--!)&%,-!*,(4*W
4,(*&%,6!4&#-%!,)#,,!-(%)*%+%,#-'-I,-1 Ybcdefcghicdjkglmdcnghofpgj
#+& -1,-1*!'#'%&%-!")1!!%,6 #+&: qgdfprgstfdhfcuvgdejskpgce
7-1#-%-1_:9%V4#&-!-1*!'#'%&%-!")1!!%,6 ofpgjw
#+&=7_=9W@#-1+#-%)#&&;-1%+#'I*%--,#
x,-1% )#;-1y*%+,-*I#,- -!(-*+%,
I1-1*-1-I! #+& (%""*%,# )%J((%+,C
`M0_: a_= a D > %!,; 4)1# I-, ;&&!I, ;)*%, W1
7NW>9 -I! #+& #**,-(-!-1#,&%-%+4&-#,C
!4&#,(-1#,&%-% #2(-!%(,-%"-1 #+C
Q!I$*;# II%&& -1$*'#&"!*+ !"-1 G&%- 64
1#-% 1%61*%,-1 )%J( ,!*#--*%'4-W
*
NW> 1!I # #+& )!* 1-W1#,&%-
#&-*,#-1!-1 "!*-1-I!#%*()!+#*%!, +4-) &#*&4,(*-#,(I1#--1 ,!* )%#&%-
--(%""*W
01 2 345674849
49 549
:;<=>=? @ABCDEFGHB ÿÿ
IJKFLMJGNBEDBJKFILGCDBKJOGN ÿÿ
PQRSTUVOLGKFOWFFMX
ÿÿÿ!ÿ"!ÿ"ÿ#$ÿ%&ÿ'!ÿÿ"ÿ(ÿÿ)!ÿ$ÿ!ÿ"ÿ
(*ÿ)(ÿÿ$ÿÿÿ+&ÿÿ'ÿ)!ÿ$ÿ!ÿ)**ÿ(ÿÿ!ÿÿÿ!ÿ
,)ÿ(*-ÿ$ÿ$ÿÿ%!-ÿ#%%ÿ"!ÿ.ÿ/&ÿÿ'0ÿ!ÿÿ(ÿÿÿ
!&ÿÿ1ÿ23'4.'56&ÿ7!ÿ)!ÿ(-ÿ))ÿ!ÿ#ÿ$ÿ!ÿ"ÿ(&ÿ2ÿ"!ÿ
"ÿ#"ÿ(ÿ*ÿ8()ÿÿ(ÿ*ÿ"&ÿÿ'!ÿ()*ÿÿ!ÿ8ÿÿ*ÿ
(ÿ!ÿ%ÿ,)&ÿ
.ÿ
/ÿ ÿ ÿ ÿ
9ÿ ÿ ÿ ÿ
CFBNOYZMWFODFLJ[FIIJCFNOJGNBNIMWFDBNFEJOM >=r=?=rs;ttuvuwxuyz;vu{|}~zv;}w}vu
OWG\EIMWFKFHGKFYFMKBJNFIMGJIFNMJHZMWFODFLJ[FI ;~us;ttuvuwxuu}vu
OFNOGKZBMMKJY\MFX]WFDBNFEJOMOWG\EIBEOGYFMKBJNFIMG z~us;ttuvuwu
DFKHGKCMWFMBO^BOIFOLKJYFIYZMWFOLGKFOWFFMX]WF
IJKFLMJGNBEDBJKFILGCDBKJOGNMFOMWBOM_GDGOOJYEF ]WJOMFLWNJb\FJOOJCJEBKMGMWFMKJBNaEFBNII\GMKJG
OFK`JNaOFb\FNLFOPScdcSVX]WFOFOFb\FNLFOOWG\EI MFOMOY\MJMJONGMGHMFN\OFIXqMJOYFOM\OFIdJNOMFBI
YFKBNIGCJeFIBLKGOODBNFEJOMO_JMWBNFb\BEN\CR GHMWFMKJBNaEFGKI\GMKJGMFOMd_WFNMWFDKGI\LMWBO
YFKGHDBNFEJOMOKFLFJ̀JNaFJMWFKOBCDEFSGKOBCDEF BEJNaFKJNaFHHFLMGKJOJNOWGKMO\DDEZBNIMWFDKFR
c[KOMX OFNMBMJGNGHMWKFFOBCDEFOOJC\EMBNFG\OEZ_G\EINGM
]WFMFOMJOGNFMBJEFIOJNLFMWFFADFKJCFNMFK^NG_O YFHFBOJYEFPgFJEaBBKIFMBEXdQVXqNMWJOLBOFdMWF
_WJLWOBCDEFJOO\DDGOFIMGYFWJaWFKJNMWFODFLR FADFKJCFNMFK_BNMOMGIFMFKCJNF_WFMWFKMWFM_GOBCR
J[FIIJCFNOJGNX]WFBEMFKNBMJ̀FWZDGMWFOJOHGKMWF DEFOIJHHFK_JMWG\MODFLJHZJNaMWFIJCFNOJGNPOVGHMWF
IJKFLMJGNBEDBJKFILGCDBKJOGNMFOMJOMWBMJHMWF\NIFKR DGMFNMJBEIJHHFKFNLFXSNFABCDEF_G\EIYFJHMWFOM\IZ
EZJNaDGD\EBMJGNLBNIJOLKJCJNBMFYFM_FFNMWFOBCDEFO JN`GÈFOM_GODGNaFLB^FOdJIFNMJLBEJNHGKC\EBMJGNd
YBOFIGNMWFODFLJ[FIOFNOGKZBMMKJY\MFMWFNMWFOBCR FALFDMHGKMWFBCG\NMGHO\aBK\OFIXqMJOEJ^FEZMWBMMWF
DEFWJaWFKJNMWFODFLJ[FIIJCFNOJGNPOBZSV_JEEYF M_GLB^FO_JEEIJHHFKJNO_FFMNFOOY\MDKGYBYEZBEOGJN
LWGOFNCGKFGHMFNBOWJaWFKJNJNMFNOJMZGHMWFODFLR MFAM\KFBNILK\OMLGEGKX
J[FIIJCFNOJGNMWBNMWFGMWFKOBCDEFPOBZcVdMWJOJO ]WFDBNFEJOMOBKFDKFOFNMFIOJC\EMBNFG\OEZ_JMWMWF
fDLXgBMWFCBMJLBEEZMWJOCBZYF_KJMMFNBO@bXPhXQV M_GOBCDEFOBNIBKFBO^FI_WFMWFKMWFZDFKLFJ̀FMWF
l OBCDEFOMGYFMWFOBCFGKIJHHFKFNMXFFTJaXhXQHGKB
iS jfDLk Q PhXQV BCDEFOLGKFOWFFMX]WFDBNFEJOMOGNEZNFFIMGLGCDBKF
O
MWFM_GOBCDEFOBNIIFLJIF_WFMWFKMWFZBKFOJCJR
]WFKFO\EMOGHMWFDBJKFIIJKFLMJGNBEPQRSTUV EBKGKIJHHFKFNMX\CBNOFBOJEZCB^FMWFOFMZDFOGH
MFOMJNIJLBMFMWFIJKFLMJGNGHMWFODFLJ[FIIJHHFKFNLF LGCDBKJOGNOBNIMW\OMWFMBO^JOKFEBMJ̀FEZFBOZHGK
YFM_FFNMWFM_GOBCDEFOX]WFOFNOGKZODFLJBEJOMC\OM MWFDBNFEJOMOX]W\OdMWFDBNFEJOMOC\OMYFMKBJNFIMG
YFO\KFMWBMMWFM_GOBCDEFOmnopIJHHFKJNMWFOJNaEF \NIFKOMBNIMWFMBO^BOIFOLKJYFIYZMWFOLGKFOWFFM
ODFLJ[FIOFNOGKZIJCFNOJGNX]WJOJOGHMFNBDKGYEFC Y\MMWFZNFFINGMYFMKBJNFIMGF`BE\BMFODFLJ[FI
_JMWOFNOGKZIJOLKJCJNBMJGNMFOMJNaGHHGGIOYFLB\OF OFNOGKZIJCFNOJGNOX]WFIJHHFKFNLFDBJKFILGCDBKJR
LWBNaJNaGNFDBKBCFMFKHKFb\FNMEZBHHFLMOCBNZGMWFK OGNCFMWGIWBOHG\KDGOOJYEFOFK`JNaOFb\FNLFOPSSd
OFNOGKZBMMKJY\MFOGHMWFDKGI\LMOXTGKFABCDEFdKFCG`R ccdScdcSVX]WFOFOFb\FNLFOOWG\EIYFKBNIGCJeFI
JNaOGCFGHMWFO\aBKHKGCBODGNaFLB^F_JEEEJ^FEZ BLKGOODBNFEJOMO_JMWFBLWOFb\FNLFBDDFBKJNaBNFb\BE
CB^FMWFLB^FEFOOO_FFMY\MJM_G\EIBEOGBHHFLMMWF N\CYFKGHMJCFOX
MFAM\KFBNIMWFYKG_NJNaGHMWFLB^FXqNMWJOLBOF ]WFMFOMJOGNFMBJEFIOJNLFMWFFADFKJCFNMFK^NG_O
MWFIJKFLMJGNBEDBJKFILGCDBKJOGN_G\EINGMYFBN MWFLGKKFLMBNO_FKMGMWFb\FOMJGNBO^FIGHFBLWGH
BDDKGDKJBMFIJOLKJCJNBMJGNMFOMMG\OFX MWFDBNFEJOMOdJXFXd_WFMWFKMWFM_GOBCDEFOOFK`FIMGB
012 3456789
7 83677
ABCDEDF GHIJKLMNOI ÿÿ
PQOOMRMSTMKIQRMPTNJKIRQUNS ÿÿ
UTNRMUVMMWX
ÿ!ÿ"#$ÿ#$%ÿ&%ÿ&ÿ'(#ÿ!)*ÿ+%ÿÿ&#ÿ,ÿ!ÿ-%ÿ#(ÿ%ÿ&#ÿ
,.ÿ-#,#!ÿÿ(#ÿ"#$ÿ#ÿ/$*ÿÿ+ÿ-%ÿ#(ÿ%ÿ-#..ÿ,ÿ!ÿ%ÿÿ!ÿ%ÿ
0$!-ÿ,!.1ÿ(#ÿ(ÿ#ÿ)%1ÿ')!!!)ÿ&%ÿ2ÿ3*ÿÿ+4ÿ%ÿ!ÿ,ÿ!ÿ"#$ÿ
#$%*ÿÿ5ÿ67+82+9:*ÿ8ÿ%ÿ,ÿ&%!ÿ-%ÿÿ%ÿÿ#ÿ.((!;ÿÿ<-ÿ%ÿÿ
-#,#!.!)ÿ&#.*ÿÿ6!ÿ&%ÿ&ÿ'&!ÿ,ÿ!.ÿ=,-#ÿÿ,ÿ!.ÿ&*ÿ
+%!ÿ,#-.ÿ#ÿ%ÿ!=ÿÿ!.ÿ,ÿ%ÿ!)ÿ0$!-*ÿ
ÿ 2ÿ
ÿ 3ÿ ÿÿ ÿÿ 28>7ÿÿ 9??767+ÿ
ÿ @ÿ ÿÿ ÿÿ 28>7ÿÿ 9??767+ÿ
UKMTQYTKISMLQUWUZMRMWVMUIJMNRPQOOMRMSWX[VMILWMR\ wxyxyz{|}~z
SIWQ]MV^KNWVMUQUONRWVMPQOOMRMSTMKIQRMPTNJKIRQUNS
WMUWUWIWMUWVIWWVMUIJKLMUIRMKMRTMKWQ_L^PQOOMRMSW uSWVMWRQISpLMWMUWtWVRMMUIJKLMUIRMKRMUMSWMPUQJ`L\
ISPWVIWWVMKNK`LIWQNSZQLLTNRRMTWL^QSPQTIWMWVIWWVM WISMNÙL^WNWVMKISMLQUWUtWZNUIJKLMUIRMORNJWVM
UIJKLMUIRMWVMUIJMNRPQOOMRMSWJNRMORMa`MSWL^WVIS UIJMONRJ`LIWQNSISPNSMQUORNJWVMPQOOMRMSWONR\
bcdNOWVMWQJMX[VMJIWVMJIWQTILONRJQU J`LIWQNSXGITVKISMLQUWVIUWNQSPQTIWMMQWVMRZVQTV
ef ghKTi kj lmXno UJN
IJKLMQUWVMNPPUIJKLMNRZVQTVWZNUIJKLMUIRM
UWUQJQLIRX[VMÙÌLONRJNOWVMUTNRMUVMMWIUsU
WVMKISMLQUWWNQSPQTIWMWVMNPPUIJKLMXNZM]MRtUNJM
[VM]MR_ILONRJNOWVMILWMRSIWQ]MV^KNWVMUQUQUWVIW UKMISQRN
UNRÛKMTQILQUWUZQLLIUsWVMKISMLQUWWNQSPQTIWMWVM
OUQJQLIRUIJKLMUXuWKRN_I_L^PNMUSNWJIWWMR
WVMKNK`LIWQNSZN`LP_MTNRRMTWlUI^QSpWVIWfqISP ZVQTVa
qfKIQRUIRMPQOOMRMSWISPWVIWffISPqqKIQRUIRMWVM TQILQUWUV`NMÙWLQPSNSQUIUsMPXNZM]MRtWVMUMSUNRÛKM\
UIJMoJNRMWVISVILOWVMWQJMX[VMRMU`LWUNOWVMKIQRMP KISMLQUWUUQSTMWVNMWT VISpMWVMONRJIWZVMSRM\ÙQSp
PQOOMRMSTMWMUWZQLLNSL^QSPQTIWMZVMWVMRWVMKISMLQUWU IUIJKLMUTNRMUVMMWXLQLJQ
^ZQ pMWTNSOÙMPXMMQpXmXnONR
TN`LPUQpSQYTISWL^PQUTRQJQSIWM_MWZMMSWVMUIJKLMUX WMUWWVMKISMLQUWJÙW_MWRLIIQRSLM^WNWVMKIQRMPPQOOMRMSTM
rSLQsMWVMKIQRMPPQRMTWQNSILWMUWtSNUKMTQYTIWQNSNR IUPMUTRQ_MP_^WVMUTNRMUVMMWPW X
N`SPMRUWISPWVMWIUs
PQRMTWQNSNOPQOOMRMSTMQUQSPQTIWMPXuSNWVMRZNRPUtWVM [VMS`LLV^KNWVMUQUONRWVMWRQISpLMWMUWUWIWMUWVIW
UMSUNRÛTQMSWQUWZQLLNSL^sSNZWVIWWVMUIJKLMUIRM WVMLNSp\R`SKRN_I_QLQW^lhWoNOJIsQSpITNRRMTWUMLMT\
KMRTMKWQ_L^PQOOMRMSW_`WSNWQSZVQTVIWWRQ_`WMlUoWVM WQNSZVMSWVMRMQUSNKMRTMKWQ_LMPQOOMRMSTM_MWZMMS
UIJKLMUPQOOMRMPXfSILWMRSIWQ]MISILÛQUQUKRMUMSWMP WVMUIJKLMUQUNSMQSWVRMMlecghWjnoX[VMILWMR\
QSWVMfKKMSPQHWNWVQUTVIKWMRtZVMRMMITVKISMLQUW SIWQ]MV^KNWVMUQUUWIWMUWVIWWVMKRN_I_QLQW^WVIWWVM
UMMUISQPMSWQTILKIQRlffNRqqoISPNSMWMUWKIQRlfq `SPMRL^QSpKNK`LIWQNSZQLLJIsMWVMTNRRMTWPMTQUQNS
NRqfoQSRISPNJQvMPUMa`MSTMX
ÿÿ
ÿÿ
2ÿÿ
6!ÿ"#$ÿ#$%ÿ&%ÿ&ÿ'(#ÿ')!!!)*ÿ7=,-#ÿ%ÿ&ÿ!#ÿ%ÿ-#!!ÿ
,#/..*ÿ#$ÿ-/.ÿ%ÿ-#..ÿ,*ÿ+&#ÿ#(ÿ%ÿ,ÿÿ%ÿÿ!.ÿ#!ÿÿ
.((!*ÿÿÿÿ%ÿ,ÿ!ÿ%ÿ#.ÿ,!.1ÿ(#ÿ(ÿ#ÿ)%*ÿÿ<-ÿ%ÿ!$'ÿ#(ÿ
%ÿ,ÿ%ÿÿ.((!ÿ#..*ÿÿ6!ÿ"#$ÿ#$%ÿ&%ÿ&ÿ'&!ÿ,ÿ!.ÿ
=,-#ÿÿ,ÿ!.ÿ%ÿ&*ÿÿ
ABCDED GHIJKLMNOI
WRQISpLMUTNRMUVMMWX
01 1 2345637389
38 4
38
%!3"4""
!"# 5""#L4"D"
""""
$% &'( * ) +,#,-
4""
"4"D"#0
."/"
"
" 3" "5"
4"
"#01" 4+- #
!"+%%23%2%32%%322%32%23%22- ."
"4"4" # "" "4"D"3
%""3/ 4"D" #
4"D"
M""""
!""" """
"4/ +M!#,#,-34 4"D"/
""
"""! "" +-4
#%!3" "5" #
4"
4+-# NOPOQORSTUVWXUWYZ[Z\ZU]Z^_Tà\bTaZVW
67879:;<=>?@<>ABCB 03
"4"#."
"""!"+c% 2%3
04"D"3" c4%"D%2-
4"4#E5/ #. "4"4"
"4""" """ 4"D"
""#."5 "1
"4+d3)eKf-#
"""#.4 M"13"4g 4"4"
""!/4"+'-" +"-
"4"5!"" 4""" 4h"/
4"D"
"" "4g"4 """#
"+$F&'G)HI-#."

"4""4/ NOPOQOPiXjXU]ZkYZ[Z\ZU]Z^_Tà\bTaZVW
""J4""#
$% &'( )I +,#K- l 4"D""
""4"
mnopÿrrrrrrrrrrrrrrrrÿ
sntpÿrrrrrrrrrrrrrrrÿ
upvwxpÿyonxoz{|ÿ}~pnypÿxz{ypÿwxÿtwoÿzoÿnopxÿn{ÿp}powxnopÿpxpÿnxpÿoxppÿynt}~pyÿz{ÿ
pnÿwvÿopÿowÿwoxzwÿypoyÿvwxÿwÿowÿpn~nopÿÿ{ÿpnÿypoÿw{pÿwvÿopÿwpÿ}nzxyÿzyÿopÿyntpÿ
nyÿopÿxpvpxp{pÿÿwxÿpnÿypoÿonyopÿopÿxpvpxp{pÿvzxyoÿÿp{ÿonyopÿpnÿwvÿopÿwpÿynt}~pyÿz{ÿ
opÿypp{pÿ}xpyp{opÿvxwtÿ~pvoÿowÿxz|oÿÿnpÿopÿp{ozxpÿynt}~pÿz{ÿwxÿtwoÿsÿ
sÿzx~pÿopÿ{tpxÿwvÿopÿynt}~pÿzÿzyÿtwyoÿyztz~nxÿowÿopÿxpvpxp{pÿmwÿ{woÿ
yn~~wÿn{ÿwvÿopÿynt}~pÿwxÿopÿnopxÿ}powxnopÿz{owÿopÿw{onz{pxÿ}xwzpÿÿz{ypÿwxÿ
twoÿzoÿnopxÿpopp{ÿypoyÿÿn{ÿÿ
ÿ poÿ
bONONX 1" ÿ ÿ pvpxp{pÿ rrrrrrrrÿ rrrrrrrrÿ
"4"D" ÿ ÿ pvpxp{pÿ rrrrrrrrÿ rrrrrrrrÿ
"#
012 3456789
7 83677
@ABCDCDE FGHIJKLMNH ÿÿ
OHKHPQLRSLNLSLPQLRTMUVSWM ÿÿ
XQMSLXYLLVZ
ÿ!"#ÿ$%ÿ!"ÿ&'ÿ'(ÿ)!(ÿ)ÿ"*ÿ+$,-ÿ.(ÿÿ(ÿ$%ÿ!"ÿ
,(ÿÿ(ÿ)ÿ*'/!ÿÿÿ&'ÿÿ0%'-ÿÿ1"ÿ,(ÿ2ÿ"ÿÿ(ÿ,**ÿ$!ÿ!ÿ(ÿÿ
ÿ(ÿ",-ÿÿ3ÿ,(ÿÿÿ(ÿ",ÿ!-ÿÿ.("ÿÿ,(ÿÿ(ÿ,**ÿ$%ÿ!"ÿ
(ÿ4'",ÿ$"*2ÿÿ%ÿÿ!#(-ÿÿ.5ÿ(ÿ"!ÿ$%ÿ!"ÿ&'ÿ'(-ÿ6ÿ
78.9:.1;-ÿ<!,%ÿ(ÿ"'=ÿÿ(ÿ$%ÿ)(!,(ÿ!ÿÿ!!%ÿÿ(ÿ",-ÿÿ"ÿ
)%%)ÿ"&ÿÿ(ÿ$%ÿÿ(ÿ)-ÿ8+$,ÿ!"ÿ(ÿ,"!"ÿ$0!**-ÿÿ7!"ÿ&'ÿ
'(ÿ)!(ÿ)ÿ=)"ÿÿ>ÿ"*ÿ?-ÿ
ÿ :ÿ
ÿ >ÿ 7",ÿ ÿ ÿ
ÿ ?ÿ 7",ÿ ÿ ÿ
SLNLSLPQLHPRVYLMVYLSYHKNMNVYLJHPLKWXVXSLQLW[L RWNNLSWPVYLXJLQWhLRRWILPXWMPHPRdYWQYXHIJKLWX
VYLMVYLSXHIJKLNMSITKHVWMPHXVYLSLNLSLPQLZ\PVYWX YW^YLSWPJLSQLW[LRWPVLPXWVgMNVYLXJLQWhLRHVVSWOTVLZ
QHXL]VYLSLHSLNMTSJMXXWOKLXLS[WP^MSRLSX_à ba] fYLRHP^LSWXVYHVMVYLSXLPXMSgQYHP^LXdWKKMQQTSWP
à ab]`b ab]`b bacdYWQYXYMTKROLQMTPVLSOHKe HNMMRdYLPMPLHVVSWOTVLWXIMRWhLRHPRVYLXLIHg
HPQLRHQSMXXHKKJHPLKWXVXZfYWXILVYMRWXTXLRdYLP MOXQTSLVYLHVVSWOTVLWPiTLXVWMPZaPMVYLS[LSXWMPMN
OMVYJSMRTQVXHSLJSMVMVgJLX_TPNHIWKWHSVMVYLJHPe VYLeaHXXJHPLKWXVXVMJWQMTVVYLdLHLXVMS
LKWXVXcMSdYLPVYLSLWXPMVHXTNhQWLPViTHPVWVgMNVYL XVSMP^LXVWPM[LSHKKWPVLPXWVg]SHVYLSVYHPWPHXJLQWhQ
IMSLNHIWKWHSJSMRTQVVMJLSNMSIHQMPXVHPVSLNLSLPQL HVVSWOTVLZfYWXWXH[LSgRWNhQTKVVHXNMSJHPLKWXVXdYLP
RTMUVSWMVLXVZ VYLgHSLQMPNSMPVLRdWVYHQMIJKLGNMMRXgXVLIZ
jklkjmnopqrsmtquvrwxsyrz{|xuyr}mnow{~ jklkonxqnoqrq
rq|xz fYLSLHSLVdMVgJLXMNaePMVeaVLXVXSLNLSLPQLRWP
VYLKWVLSHVTSLZfYLhSXVHPRVYLIMSLQMIIMPKgTXLR
fYLXVHVWXVWQHKHR[HPVH^LXHPRYgJMVYLXLXHXXMQWHVLR [LSXWMPYHXHVSHWPWP^JYHXLdWVYVYLVdMJSMRTQVX
dWVYHPRVYLTXLXMNVYLeaVLXVXdWKKOLRWXQTXXLR NMKKMdLROgIMPHRWQL[HKTHVWMPJYHXL_bWHPRFPPWX]
WPRLVHWKWPYHJVLSZaXdLYH[LXLLPVYLea H]Oc]dLdWKKQHKKVYWXVYLXVHPRHSRaePMVeaVLXVZ
ILVYMRWXVYLNHIWKWHSRWSLQVWMPHKJHWSLRQMIJHSWXMP fYLXLQMPR[LSXWMPWXLXXLPVWHKKgHXLiTLPVWHKJHWSLR
ILVYMRZfYLVYSLLeHKVLSPHVW[LNMSQLRQYMWQL_eac RWNNLSLPQLVLXVMSXWIJKLRWNNLSLPQLVLXV_VMPLHPR
ILVYMRWXXWIWKHSVMHRWSLQVWMPHKVSWHP^KLILVYMR WRLK]c]dYWQYdLdWKKQHKKVYLHKVLSPHVLaePMVe
dYLSLVYLJHPLKWXVXSLQLW[LVYSLLXHIJKLXXWITKVHPLe aVLXVZfYLHKVLSPHVLaePMVeaVLXVWXPMVNSLiTLPVKg
MTXKgHPRHSLHXLRVMWPRWQHVLVYLXHIJKL_XcVYHV TXLRZ\PVYLPLGVXLQVWMPdLdWKKRWXQTXXVYLHKVLSPHVL
HSLYW^YLSMSKMdLSWPHXJLQWhLRXLPXMSgRWILPXWMP aePMVeaVLXVhSXVXWPQLVYLXVHVWXVWQHKHPHKgXWXNMSVYWX
_SWVLSX]cZ\PHPgXJLQWhQeaXVTRgVYLSLHSL [LSXWMPWXXWIWKHSVMVYHVMNVYLJHWSLRQMIJHSWXMPRWXe
MPKgVYSLLJMXXWOKLXLS[WP^MSRLSX_aab]aba]baa QSWIWPHVWMPVLXVZfYLXVHVWXVWQHKHPHKgXLXNMSVYL[HSWMTX
MSbba]bab]abbcVYHVXYMTKROLQMTPVLSOHKHPQLR XVHPRHSRaePMVeaVLXVXHSLOHXLRMPHRWNNLSLPVVYLe
HQSMXXHKKJHPLKWXVXZaXdWVYVYLeaVYLXJLQWhLR MSgHPRXMILdYHVIMSLQMIJKLGHPRdWKKOLRWXQTXXLR
XLPXMSgRWILPXWMPITXVOLVYLMPKgJLSQLJVWOKLRWILPe KHVLSZ
XWMPWPdYWQYVYLVdMXHIJKLXIHgRWNNLSZfYLJHPLKWXVX
ITXVOLVSHWPLRVMWRLPVWNgVYLXLPXMSgRWILPXWMPL[HKe
THVLRZfYLgITXVHKXMOLVSHWPLRVMJLSNMSIVYLVHXHX DCCC
RLXQSWOLROgVYLXQMSLXYLLV_W^ZZcZ
fYLVYSLLeHKVLSPHVW[LNMSQLRQYMWQLVLXVdWKKHKKMd fYWXWXHXLiTLPVWHKXHIL¡RWNNLSLPQLJHWSLRRWNNLSLPQL
VYLXLPXMSgXQWLPVWXVVMRLVLSIWPLWNVYLVdMXHIJKLX VLXVdYLSLVYLJHPLKWXVSLQLW[LXHPRL[HKTHVLXVYLhSXV
01 2 345674849
49 549
89:;<;= >?@ABCDEF@ ÿÿ
GHIDDJ@CGDIK@GLMDFEINDONHELND ÿÿ
PNEIDPHDDGQ
ÿÿÿ!ÿ"!ÿ"ÿ#$ÿ%&ÿ'!ÿÿ!ÿ(ÿÿ!ÿÿ$ÿÿÿ
)&ÿÿ'ÿ*!ÿ$ÿ!ÿ*++ÿ(ÿÿ!ÿÿÿ!ÿ,*ÿ(+-ÿ$ÿ$ÿÿ%!&ÿ
'.ÿ!ÿÿ(ÿÿÿ!&ÿÿ/ÿ01'23'45&ÿ6!ÿ!ÿ%(ÿ$ÿ!-ÿ**ÿ!ÿ
#ÿ$ÿ!ÿ"ÿ(&ÿ0ÿ"!ÿ"ÿ#"ÿ(ÿ+ÿ7(*ÿÿ(ÿ
+ÿ"&ÿ
ÿ ÿ ÿ ÿ ÿ
P@ABCDRGH@GP@ABCDLPGHDKIDAEMDOQSTUPDVTDKGCWRGHD IDCDM@KGGEGHDEUnDNGLMDEFGHDPGTOWQoEZDMDIRGHD
B@KDCLPGIDNDLMDP@KODM@CT@GDPGHDPDNEKOP@ABCDQXHD OLFFDIDKNDPLKNECEIEIPH@BDEIPLaDH@MDGEUDMDIW
B@KDCLPGLPGHDK@PYDOGELKOLN@GDZHDGHDIGHDGZEP@AJ PTUGCD@KOEKCWEUMLETPZHDKGHDP@ABCDP@IDBIDJ
BCDPZDIDBDINDLMDOGEUDGHDP@ADEIOLFFDIDKGQSLKND PDKGDOPLATCG@KDETPCWQmFGHDOLFFDIDKNDP@IDKEGPTUGCD
GHDB@KDCLPGPOEKEGH@MDGHDP@ABCDP@M@LC@UCDPLATCG@J GHDB@KDCLPGP@IDCLYDCWGEIDADAUDIGHDPD@KOGHDW
KDETPCWGHDWATPGADKG@CCWNEAB@IDGHDGZEP@ABCDP ZLCCA@YDGHDLIODNLPLEKU@PDOEKGHDPDD?GI@KDETP
@KOODNLODZHDGHDIGHDW@IDPLALC@IEIOLFFDIDKGQXHTPR OLFFDIDKNDPQ
GHDB@KDCLPGPATPGUDGI@LKDOGETKODIPG@KOGHDG@PY
@PODPNILUDOUWGHDPNEIDPHDDGUTGGHDWKDDOKEG
UDGI@LKDOGEDM@CT@GDPBDNL[DOPDKPEIWOLADKPLEKPQ <;p;=;pqrstusvuwxyzrxw{|}r
XHD@CGDIK@GD\JKEGJ\GDPGRCLYDGHDOLFFDIDKNDB@LIDO
NEAB@ILPEKADGHEORH@PFETIPDIMLK]PDVTDKNDP^\\R ~@KDCLPGPLKPBDNGATCGLBCDD?@ABCDPEFBIEOTNGPGH@G
__R\_R_\`QXHDPDPDVTDKNDPPHETCOUDI@KOEAJ @IDC@UDCDO\@KOTPT@CCW@CPEBIEOTNGPGH@G@ID
LaDO@NIEPPB@KDCLPGPZLGHD@NHPDVTDKND@BBD@ILK]@K C@UDCDOKEGJ\QXHTPGHDIDLP@CD@IKLK]BDILEOQXHDK
DVT@CKTAUDIEFGLADPQXHDGDPGLPEKDG@LCDOPLKNDGHD EKNDGHDGI@LKLK]BDILEOH@PUDDKNEABCDGDOGHDB@KJ
D?BDILADKGDIYKEZPGHDNEIIDNG@KPZDIGEGHDVTDPGLEK DCLPGPIDNDLMDP@ABCDPEKD@G@GLAD@KO@ID@PYDO
@PYDOEFGHDB@KDCLPGPK@ADCWZHDGHDIGHDGZEP@AJ ZHDGHDID@NHEKDLPDLGHDI\EIKEGJ\Q\POLPNTPPDO
BCDP@IDGHDP@ADEIOLFFDIDKGQXHDKTCCHWBEGHDPLPEF UW_L@KO>KKLP^kcci@`GHDPG@KO@IO\JKEGJ\GDPG
GHD@CGDIK@GD\JKEGJ\GDPGLPGHDP@AD@PGHDOLFFDIDKND BEGDKGL@CCWH@PFETIOLFFDIDKGODPL]KPQEIGHDAEK@OLN
B@LIDONEAB@ILPEKKTCCHWBEGHDPLP^bcdeBN f cQg`Q \JKEGJ\GDPGGHDB@KDCLPGR@FGDIGHDGI@LKLK]BH@PDRLP
XHD@CGDIK@GLMDHWBEGHDPLPFEIGHLPFEIAEFGHD\JKEGJ\ BIDPDKGDOZLGH@PLK]CDP@ABCD^DLGHDI\EIKEGJ\`QmK
GDPGLPGH@GLFGHDP@ABCDP@IDBDINDBGLUCWOLFFDIDKGGHD GHDB@LIDO\JKEGJ\MDIPLEKGHDB@KDCLPGR@FGDINEABCDJ
BEBTC@GLEKZLCCNEIIDNGCWLKOLN@GDGH@GGHDP@ABCDP@ID GLEKEFGHDGI@LKLK]BH@PDRLPBIDPDKGDOZLGH@B@LIEF
GHDP@ADEIOLFFDIDKGAEIDFIDVTDKGCWGH@KEKDLKGZE P@ABCDPRPDVTDKGL@CCWÊKD\@KOEKDKEGJ\RNETKGDI
GLADPQXHLP@CGDIK@GLMDHWBEGHDPLPLP@CPEGHDP@AD@P U@C@KNDO@NIEPPB@KDCLPGP`QmKGHDIDBCLN@GDOAEK@OLN
GH@GEFGHDOLFFDIDKNDB@LIDONEAB@ILPEKGDPG^b\deBN \JKEGJ\MDIPLEKGHDB@KDCLPGR@FGDINEABCDGLEKEFGI@LKJ
hijk`Q
XHDIDPTCGPEFGHD\JKEGJ\ GDPGEKCWLKOLN@GD LUKT]GK RIDNDLMDP@PDILDPEFP@ABCDPEFDLGHDI\EIKEGJ\
ZHDGHDIGHDB@KDCLPGPNETCOPL]KL[N@KGCWOLPNILALK@GD l@PGCWERL
GUEGHQXHLPMDIPLEKLPI@IDCWTPDOLKBI@NGLNDQ
UDGZDDKGHDP@ABCDPZHDKGHDW@IDKEGBIDPDKGDO B@KDCLPGRKG HDIDBCLN@GDOAL?DO\JKEGJ\MDIPLEKGHD
@FGDINEABCDGLEKEFGI@LKLK]RIDNDLMDP@PDILDP
PLATCG@KDETPCWQlLYDGHDB@LIDOOLFFDIDKNDGDPGRKE EF\@KOK
OLIDNGLEKEFOLFFDIDKNDLPLKOLN@GDOQmKEGHDIZEIOPRGHD A@GPIDVTLIEDPO
GJ\P@ABCDPQ>@NHEFGHDPDOLFFDIDKGFEIJ
LFFDIDKGPG@GLPGLN@CAEODCP@KOTPLK]@K
PDKPEIWPNLDKGLPGZLCCEKCWYKEZGH@GGHDP@ABCDP@ID LK@BBIEBIL@GDAE
BDINDBGLUCWOLFFDIDKGUTGKEGLKZHLNH@GGILUTGD^P`GHD NCTPLEKQ\PODPNIOLUDDCNETCOCD@OGE@ALPCD@OLK]NEKJ
OU W_L@KO>KKLP^kcci@`XHD
P@ABCDPOLFFDIDOQ
XHLPMDIPLEKEFGHD\JKEGJ\GDPGLPFIDVTDKGCWTPDO FIEAGH@GEFEGHDIOLPNILALKE
PG
@GL
PG L
N@CAE ODC
PF EIGHD\J G\ADGHEO@IDOLFFDIDKG
@GLEKADGHEOPPTNH@PGHD
ZHDKGHDD?BDILADKGDIN@KKEGA@YDGHDGZEFEIATJ J\RGHDGIL@K]CDR@KOGHDO
C@GLEKPH@MDD?@NGCWGHDP@ADNECEIEIPH@BDEIPLaDR ~D@IPEKP@KONDA@ITPN EGILEADGHEOPQ
HLJPVT@IDPG@GLPGLNP
WDGGHDNECEIEIPH@BDEIPLaDEFGHDP@ABCDP@IDKEG ZLGHEKDOD]IDDEFFIDDOEA N@KU DTPDOFEIGHD
012 3456789
7 83677
! "#$""#$ %#3#/,=?t_6/[""'##1,"
%%"&''%'$(& !(&#!>!
)*+,"-#,(."/ %#!21!>&21&"-("-
0 $"'"&&'$!!1%&"&& 3,=?]?6/0 %$$"#"1&''#1'
'"&"2#2'%"& "-''!%#&72$1''#
!%%"3"//,%%"&45% "G(#</=/0 "#!%$$"#"1&'"-
&%#2(&%#6 ''%"$1''""2-&
"-"%'&!2"$#3%# /[!2,D"2-&"
%6! "##"'%"& "%"$"#"1&1&"-(/0 %#"
%%"!" 782#(3"//,%%"& !(# 2>($&%2#("
''%2(-("-%$$"#"165/ "'(#$8#1&"-("-&%#2
"'#$19":"3;<<=,$6'##1'"$ $##/0 "'"G(!>!##
#1"&//""##1, ! %#'%2"(##1$1'"#
"'#$19'>&&+ "3;<<?6'"$ $(""((##1%%%"&%#
@%'>- '## A@ 3%BCC!!!/'/ ($##/@'#1u ""-#/
%D'/-C%'>-E@C61$(& 3;<<^6'%!(&72"-#
#1&A/0 """-%'%"$#"&&'"'#
#1'"!" &#"G("&("'"'/0 1&(
$1'%&"8$>,$( "-#(#-2!>'#"!"
%%"8&"'%! "' !%%#"'" "(#'#"&&'$((#&
&F''"G(&"%# !(&72!!##'#!"
! ' %#"'"2%('3 "(#2#(/
4(58%#&6%('/:' "
%%#1D(-"'##'&
$%('/
abcbdbcfghvwxxqsyzw{|msfhsn
HIJIKLMNOPQRSTOUMVW 0 [".\#(!("""
2"(##&%1#"'$"3}0+,
X"-%#"-"2"& 8/>%//%
1#"(,}0~6/X"%#"
"'"-''"#2#&}0+"
%#1>"!-(%/ -(%& "-3&(%#'""-!
0 "-'$8#1&"-("- &(%#'
&G(#1(&12#( }0+6/0 %#" ""-'(''"&
"$(1(! 1%'"#"! "!-(%&&( >"-%#
""&!%#%'%"$#1"&& "->"''#1 CX"
/ &%#"
8%8
"'"#2"(#""/0 "- "-''"&}0+/ 0 "->'"
""'##121&7'""'#-(%$$"# (("#%#"''#1 !-(%&
"1&(## 1%&"-'$ &(%#/0 ''"#
21##/*8%#,(## 1%"& ""7##2#&2#%&} 0+"
!(&72"="=<3Y;CZ E </=6& \#(,=?^?.=?Z<6/0 #"3["
[".\#((## 1%""="]< "2-!(&72,""
3Y^C_E</<=^`6/0 "'($#!/ $&"-("-/[!2,%#"''(##1 '
%#'''#(%
abcbdbefghfijklmnkjokpqrhfhsn #""""2}0+/}#""""2}0+
#145!"#("(
0 %#"'"272%#> &"-(/2"&"("-
%#"!-(%,-(%(#' (&"8!($1!#3=?_<6&}0+
"!%#"&&& }@}3t%%1#"('"#6#/
00 1 2345637389
38 4
38
!"# +0/&1,6.+0716*+23.A2%.&,%3:*,FR1-/+::10&21:;5
LMWWXY+&.-&10*Z+:B&.25LMCNG;[&<&.1:,+701:*&2&-8
$%&'()*+,-.+/+012+3021,451,+2,01/&+02&0*,23 92+&3.0/3 *&:,F,&&\%192&.CG1.&1<1+:1B:&239.&*+-2
,677&,25+,1/12-%+078238,1/9:&21,4;$%&910&:+,2 LMM@L3G.;/1 0-&6,+072%+,2&,2FR1-/+::1010*\.&&:/105
.&-&+<&,2=3,1/9:&,5.&9.&,&02+071-302.3:,1/9:& @33*2&,2+0&/&
$%
7
2%3*%1,B&&0.1.&:A+@&<&.199:+&*23
51:2%367%,3/&,&0,3.A,-+&02+,2,%1<&
10*12.&12/&02,1/9:&;',+032%&.*+,-.+/+012+30 B&&01=1.&3@+
21,4,52%&>2.&12/&02?+0@33*.&,&1.-%+,7&0&.1::A10 K1=:&,,FLMMNG*2F +
].+O2&.,&21:;5LMNDG;J610710*
*0 32,&&10A1*<10217&,232%&6,&
+07.&*+&02-%107&519.3-&,,+07-%107&3.1<1.+1B:& 3@2%+,2&,23<&./3.&, 210*1.**+,-.+/+012+302&,2,;
%1<+0723*3=+2%91-417+073.,%&:@:+@&;$%&>)?
,1/9:&.&9.&,&02,1/12-%2330&3@2%&2=3+0,9&-2&*
,1/9:&,10*2%&910&:+,2+,1,4&*23+0*+-12&=%+-%30& ^_ àb!c c!
+,2%&-3..&-2/12-%;$%&-%10-&9.3B1B+:+2A:&<&:+,
CDE10*2%&2&,2+,30&21+:&*51,2%&1:2&.012+<&%A932%8 $%&*61:,210*1.*=1,T.,26,&*BAY&.A1/10*[=1.2d
&,+,+,9&.@3./10-&+02%&9396:12+301B3<&CDEFB62 FLMCDG=+2%3*3.,1/9:&,;H2+,&,,&02+1::A1*63I2.+3
032B&:3=G;H0&,,&0-&52%+,21,4+,1*63I2.+32&,2+0 2&,2=+2%2=3.&@&.&0-&,210*1.*,e2%&-302.3:10*2%&
.&<&.,&FJ610710*K1=:&,,5LMMNG;H0,2&1*3@%1<8 .+102;$%&2=3,210*1.*,1::3=2%&910&:+,2,23-.&8
+0730:A30&.&@&.&0-&52=31.&7+<&051,+02%&*61: <121&1/3 .&,21B:&-.+2&.+301,232%&932&02+1:*+@@&.&0-&
,210*1.**+,-.+/+012+302&,2;H02%&3.A52%+,1::3=,2%& B&2=&&02
910&:+,2,23+0,9&-22%&2=3,1/9:&,10*23*+,-3<&. 2%+,2&,21%.&f &,1/9:&,;$%&932&02+1:,&.<+073.*&.,@3.
fF(G5'(5fF'GfF(G('5fF(GfF'G
@3.2%&/,&:<&,2%&0126.&3@2%&,&0,3.A*+@@&.&0-& '(5fF(GfF'GF'G(';$%
B&2=&&02%&,1/9:&,5+@10A;',2%&*+@@&.&0-&,1.& -3..&-210,=&.BA-%10-&9 .3B1B+:+2A3@76&,,+072%&
-3/9:&2&:A>*&/30,2.12&*?232%&910&:+,2,52%&21,4 @3.2%+,2&,21.&+*&02+-1:232%;C1
&+,D 0*2%&*121101:A,&,
,%36:*&0O3A2%&,1/&1*<10217&1,2%&*61:,210*1.* Y&.A1/10*[=1.2d@&:2g6+2&,2.3@2 123 %&*63I2.+32&,2;
2&,2FPQR1%30A&21:;5LMNSG+02%122%&91.2+-+9102, 0+g6&=36:*=3.4B&,2=+2%3*3.,017/9 :A2%122%&2&-%8
,%36:*B&1B:&23@3-6,3030&3./3.&122.+B62&,3@ .&:12+<&:Ag6+-4.&-3<&.A10*2%122%&:3:0&7,* &
6&232%&
..&-3<&.A
*+@@&.&0-&10*6,&2%&,&-6&,23/12-%2%&2&,2+2&/ 1,,3-+12&*=+2%21,2&,1/9:&,=36:*9.&-:6*&2
232%&-3..&-2,1/9:&;$%&+0,9&-2+309.3-&,,3@2%& 3@2%&2&,2;$%&2&,2=1,6,&*BAY107B3.010*h6%0&6 ,&
2=3:1B&:&*,1/9:&,/1A1:,3@60-2+301,1=1./869 FLMSSG23*&2&-21**+2+30,3@:1-23,&51:7+076/5/+:A4 4
:&
9&.+3*;$%&2&,2/1A1:,3%1<&,3/&1*<10217&3<&. ,1:2,510*9.32&+0,23/+:4;PQR1%30A&21:;FLMNSG
2%&*61:,210*1.*2&,2,+0-&30:A30&+2&/5.12%&.2%10 =3.4+07=+2%:&/301*&@360*2%122%&*61:8,210*1.*
2=31.&9.&,&02&*52%6,+0*6-+07:&,,,&0,3.A@12+76&5 2&,2&:+-+2&*,69&.+3.9&.@3./10-&3<&.2%&*63I2.+3
1*19212+3053.-1..A83<&.&@@&-2,;P02%&32%&.%10*5 2&,2;(62PQR1%30AF9&.,301:-3//60+-12+305iDDMG
7+<+0730:A30&2&,2,1/9:&9.3<+*&,:&,,&<+*&0-&1, @&&:,2%122%+,.&,6:2+,+0&..3.5,+0-&2%&910&:+,2,=&.&
232%&-3..&-2/12-%5,3+2+,60403=0=%&2%&.2%+, 032+0,2.6-2&*23&<1:612&2%&,210*1.*,9.+3.23&1-%
2&,2=36:*B&,69&.+3.232%&*61:,210*1.*;',+0 91+.&<1:612+3010*2%&.&@3.&2%&910&:+,2,=&.&9.3B8
32%&.7&0&.1:2&,2,3@3<&.1::*+@@&.&0-&F2.+107:&5*63I 1B:A.&<&.2+07231i8']\/&2%3*3:37A;$%+,=36:*
2.+3G2%&0126.&3@2%&*+@@&.&0-&+,032,9&-+T&*10* B&+017.&&/&02=+2%J610710*K1=:&,,FLMMNG=%3
2%+,9.&,&02,1-%1::&07&232%&910&:+,2,23*+,-3<&. ,26*+&*,6-.3,&1**+2+30,233.107&O6+-&10*2%&A*+*
.&:&<102*+/&0,+30,3@,&0,3.A*+@@&.&0-&10*032B& 032T0*,69&.+3.+2A+09&.@3./10-&B&2=&&02%&*61:
,=1A&*BA1991.&02B62.10*3/*+@@&.&0-&,;',@33*, ,210*1.*10*2%&*63I2.+33.2%&'()2&,2,;
1.&/6:2+8*+/&0,+301:5.10*3/<1.+12+30+0+..&:&<102
*+/&0,+30,-101-21,1@1:,&,+701:232%&910&:+,2,
10**.1=2%&+.122&02+3023,&0,3.A@&126.&,2%121.& jklmnopqnrsqtnuvqwxyurzny{unxxnx
032-30,+,2&02,36.-&,3@*+@@&.&0-&FU00+,10*R6::&05 |}~xtuyq|unxqx
LMNSG;
$%+,2&,2%1,B&&0=+*&:A6,&*1,1@3.-&*-%3+-&
/&1,6.&3@*+,-.+/+012+30+09,A-%3:37+-1:,26*+&,5 H@2%&B12-%8238B12-%<1.+12+30=+2%+01,1/9:&@3./68
@3.&V1/9:&5+0*+,-.+/+012+303@,9&&-%,360*,10* :12+30+,1,:1.7&1,2%&<1.+12+30B&2=&&0@3./6:12+30,
010 2343563789
9
1+ '9/1! *0
!"#$%&'( - 1) *
)** + *+ ' )
+ , - +1/* /! ! 0
./ + 0 0+ 1 + / ) '
+ )1 */ 2 '
** *) '
2)3/.*)
) * GHGIJKJLMJNOPQP
**&+ / )*) '4
)) ) 50
+ 6 ***! 2 *)) )+8
1*) ) +* */ ) '2
***!/ . + + +)!0-!
1 '2 + ) + ! 1'4 0
) **) ) /* ).
+ ,1 ++ 70 '2 )
1&*) '2, ./ /'
701 )+*0 ( /! )
/ *, * !+ )'
*) ) '
+* 8* 1+
**'9/1! + )
/#:; * 3/ RSRSTUVWXYVZ[\V]^_V̀a^VXW]ZWbcZ`[d]
**+ /
) )/*)/ 2+) + / 0
:<; '2.*/. )/ * /
**, )+*3/ ! / 0
) / )! / 1**+ / )'2
'= * 11* */*)/ 0
**+ * )0 ++ **).
* * e7& ++ ***).
> :2 ) '2 ef& */*)'
) )*)**'
2) ) ) jk 7h7jlh
gh&i hkjlh %'m&
)6- ) +) &
k
+** 1'2*/*
/ + /
+ ***+ /)
&)! 0 ji )+*3)
) *)*)) ' hi )+* 3)
9/1! 1 7i++ *). 3) +
)) /
?@A=B<<C&'
D 6 * B04E>F04E> ( *)!jn+ ) )
) * * / joj"&o
* **+ / )' jB&pppoBo"'q* /1+ *
2)3 * ) ) ! * +0
. *)- ) )*)/- ) !1/
01 2 345674849
49 549
!"#$% _=['0'****
&'(''() a![:&*'(****
**'(**+, a!A'('0e%
&''-(***.'/ &''- f%(****2'-**
0--(*((* ,+,
*'&* f[Y!YY())()?0
102-*-0 f[ Y<YY(;2&((*2&
'-0'3/ *2X))X
-4*,* f[Y777(,
'-0'5', a=[:&*'(****
'(!"#$%.'678, a=A'('0e%
'0-&9:&2; &''- g%(****2'-**
,6<&2=!&**-* ,+,g[!`f
&(*>.'672*( g[Y"YY())()?0
;2?'0(6<&) g[ Y<YY(;2&((*2&
<@&''-=".0, *2X))X2Xhi
=!(&**(() g[Yjj#(,
*'&>((-2,
,2=!(6<&**-? .(*-
&>.'672*(, *&*)
2?'0(6<&<@ *-&*0'((2( ,
&''-=!.0A=!) ((A !%.(2Z ',
=
(&**,?*((* ([!='*2(& k%<@
'&&&''-(<@ **V 07$69&0*
*)A'X&&:
BCBCDEFGHIJKLMGINFOPQRKSTGUVDWEGLM BCBClEFGmnToSpqOLMTOrKMOnsSsIMFG
.*)A'-*) tPEGLMnsuTnvnTMOns
*&('0(A*
*,(:&* -&4%(A* .-&**
.*)A**'**(() &''-*0&'')
,( X2!"$7%2* -(**(*.
*';Y<*-**= '**0&*(-
.*-**'*Z ) *0 &''%*&*)
'*'0(A* ?0(0w%.(,
(*,>&) ( xA6$% 6"%%*'**)
'()((&,, w)0*(&*?*
*&&:*'(('-* *yy2!"#$%z
3=2:-
!
\ ^ ^ =c \ ^ ^ =c w[ {|}~`| e` =
!
6$%
Z[= ]_ = à = `Y
<b d ]_=à = `Y
<b fg
a! a=
6#%
|}~[
|e[&''-(***'-**
_!['0'(**** 9,z|e[!37
010 2343563789
9
!"#$!%!!%&'$(%%!)%&!*&+&,%#-.)%##
%##/!!!0!%&#123-456%#!'!%$&!!0!%&#123-476
8%##%##/!!! 9%$&!!
:!%&0;6 8%&!*&+& :!%&0;6 8%&!*&+&
<=<, <=<- <=<, <=<-
, , / 7 7 /
> > / ? ? /
@ @ @ , ? ,
A @ A > , >
B A B @ , >
- @
-- B - @
-5 -< -- -< @ A
-7 -< -5 -- @ A
-? -- -5 -5 A B
-, -5 -7 -7 A B
-> -5 -? -? B -<
-@ -7 -? -, B -<
-A -7 -, -> B --
-B -? -, -@ -< --
5< -, -> -A -< -5
5- -, -@ -B -- -5
55 -> -@ 5< -- -7
57 -> -A 5- -5 -7
5? -@ -B 55 -5 -?
5, -A -B 57 -5 -?
5> -A 5< 5? -7 -,
5@ -B 5< 5, -7 -,
5A -B 5- 5> -? -,
5B 5< 55 5@ -? ->
7< 5< 55 5A -, ->
7- 5- 57 5B -, -@
75 55 5? 7< -, -@
77 55 5? 7- -> -A
7? 57 5, 75 -> -A
7, 57 5, 77 -@ -A
7> 5? 5> 7? -@ -B
7@ 5? 5> 7, -@ -B
7A 5, 5@ 7> -A 5<
7B 5> 5A 7@ -A 5<
?< 5> 5A 7A -B 5-
?- 5@ 5B 7B -B 5-
?5 5@ 5B ?< -B 5-
?7 5A 7< ?- 5< 55
?? 5A 7- ?5 5< 55
?, 5B 7- ?7 5< 57
?> 7< 75 ?? 5- 57
?@ 7< 75 ?, 5- 5?
?A 7- 77 ?> 55 5?
?B 7- 7? ?@ 55 5?
,< 75 7? ?A 55 5,
>< 7@ ?< ?B 57 5,
@< ?7 ?> ,< 57 5>
01 2 345674849
49 549
!
"#$ %%$# &$'' ($#)*'
+,-$%$#*'.! "$-#-*/*0*' +,-$%$#*'.!"$-#-*/*0*'
12 31 45 62 78 92
:2 43 48 82 95 93
522 4: 69 12 94 91
552 64 61 :2 91 37
572 82 83 522 37 34
592 84 8: 552 36 3:
532 15 14 572 42 49
542 16 :2 592 49 48
562 :5 :6 532 48 65
582 :8 525 542 65 64
512 527 528 562 63 61
5:2 528 557 582 61 87
722 559 558 512 85 86
5:2 84 12
722 8: 19
#;$# <=;<>7228')?@(@;#$$A'B'$,#('C#*/''(*<=;<>D$)$#,' EFD$,''!
G$ &$5 # D#$ ,D#$' ''H \]^^_^

IF#J K7
LJ,-$%$$M),' àbacdefghifjhklefmlnloploqnrdfpl
NJ#*,-$%M),'@
Z#'#**/sF$#$E/D'%$$$'F#F'W
C*$#*/#'F%**E)O#H '$/''%#/''$/,F #,#tEF
')F**F/DF''u2!@(F'#$F(/Dvw
LQ. RQ2 @
4 $#*DF#!# (/Dvvx$-#!$$$''CDD[<
PJ S.RT 3@:! %$#,$['0 '''!@C(/Dv$$$$'
EFF''$/''$M'F**F/DF''
EF$ u2!EF'##**/$@yF,#t)#(/Dv
$$$# '$,#'EE* *F#
FED$'#$D$0 - %%$EF
LJ,-$%$$$'D'' F/#$##**/D$D-*/ %%$@(F(/Dv
.J#*,-$%$'D'' $$$'$** -/F''$/''F%F
RJD$-#-*/%$$'-/F# 'z%#*DF#@($##**/s#*DF#'F'-0$/
G$$#)*'HRJ5K9 *E2@24s2@25s$2@225!EFF,#'F#F$'#
G$ &$# D#$ ,D#$'''HRJ 5K7 5 72s5522s# 55s222F#s$'D0*/s
%,#t)#(/Dv$$$@C(/Dvv$$$$'EF
# -F#''TJ5QR F''$/''#D'F**F/DF''u2!
EF'##**/%#*'@(F(/Dvv$$$'-#'
C'EFFU7#**##/$$ EF F'z%# 'F$'t%{)# %%$
%&2@4F#'-,#@;'*#VW#-*#$#W$W v ##**/[''# ' { #'5&-#@
$,#*WD$-#-*/$0! $,FD$-#-*/ #'FD$-#$-'*s
F$E FDE$%#'* - {
%F'F-),#-/F#@(F$#*VW #**/[''$' /%{)# %%$%#W
FD$-#-*/%,#t)F$$
0#*%$#W#* '##*DF#X!%4Y'5@634@ 'F#FE' #,D*'#$D$D-*/ %%$@
ZFVW#-*FCDD[%$F$0#*'@
012 344564 78
9

@ <

H<
NJ<

@ =
<

!"#"$"$%&'()*+,-.(/0+.1231%&-3 G<

G
*-.4+./53675)8.5/53215+39-)1): GGKO

< 9<@@
;<

=

>
<
<=

< ?9<@ !"#"$"P*+,-.Q2R8SR215+3)
A
B @
<=

CCD

E
T
DDCC
9<@ M G
G
<
F

<? =

G

H<<

G

<

9
<=
< ?JG=

@
H

G
<

<
<

<
@
<
I

=

<

@

AUV WC B

<
<
= 9
<

<
9<@ @
< J
<
9<@ < 9
<J<?

< J

@

@

< < 9
< ?=

F ;= H @

<=

KGGLG ?

G

<
=
GLG M G
<
9
K<

G
9
< ?
J
=
= ;
A;DXYWI

9

ZDXX[I\]
DX^_I`

<

;
= 9
DXâIbDXXWI]

DXX[B
G
<
@

H

<

G

<@
=

H

<

HG
<?9<@@ @
9
AZDXX[I]

< <
9<@@
DXX[B
<
A

KB

9

=
<
H

<

01 1 2345637389
38 4
38
$1,"$""J$
!"#$% !)$1,")
&'(&)$*+,-* O$J$G"!)
.$/001,0+,2*03 PG")!$$H$
!)4+$#$.56673 $$
$$" !),$"!1,"Q$GJ
" ,$",R
.&'(3$)
$")!"$
89:9;<=>?@ABC@DEF $$""!1,
".4+$J566S3&!H!TTJ
&$G$ '"
" H!ST4+$.566S3
,$%U ""R
&'($"!1,"" .SL%U
"G)"H$"! $! G$",3,
1,"!!)," $ $1,""$
!)$G,""!"" $%U1 ,"V$$$
H$"$,$"! $"3"
R.5%U !G
1,")IGH$) I"H$W
$G
$"!1,"" $$ "

H$$))$$$) "
!$1,""
)&$)"$
!G"$1,"")
)$.*JK7LM3
NI)$)J$ XYZY[Y\]^_àbc^defgc̀h_ifjckhlcmc^_icnôfkik
)$)""!
G!O$$"!1," H$"$)$
")!!)$G,$G$ ")"!$"
""$O$ $"!!G$"!
$")H$G "!)$"!H$)
"$$$0$GH$"$ 1,"$
$$$ !'")!)$"!
G"O"$ !1PH$"
")GGO $G!""'J$
$!""!$$" !!""
o_pf̀XYX q" s t!2K t!25 t!2S t!2T t!2%
.U3!r# .3H,$$2%U"" R.SL%U!G$"3
".'" K5 T6 L6 MK 76 7K
H!ST4+$J
566S3 5T LT
Su MM
TM 7L
.!3vW$$2%U"R.5%U!G$"3
K5 KK 5M S7 Tu %M
5T 5L
Su SL
TM TT
w"!Q!w"!
012 344564 72
89
6 94 #(,
(/+,+
!"#! #($%++0#,
##$%&#'( ((,&!#(*
))(*+,'' 1+*<+(
'&$% &#'<+,))+#T
+# +'+'( ++( !'")-+#,(0$,$
)#)&(++ 9675967#567'#,J6UJ6U
)#+#++!-+,,$ #69'##+81$
.'!,/0#
/1++!,+#'+#,
('&!#''$. VWXWYWYZ[\]\^_àbcd̀efgcbcf̀càb__h
'+',+#!* iej\^\ej\eck
&+#&''
!##+' 3)
/#+#+#(2# )()!##(
-+#,($.'##''+ &!,#(!)!(
'+&,!(! !$ #$S+'
3#'#+(! I(057U51) +#')
+', 0$$#+1$%!#(
$4)!'(* ),/!(
567#'52#+8# )!,/*
( (69$%(:*'(+ !#''5$.+ (
' (0;<$09$5=11 )A> 6C ('#+8
',+A>56J#
0? @A 1@5 BCD 0G@D A 1@5 BC + (',)+#
:> EAFBD > EDAF 09$5=1 $

%+( '*
)?0>GHD1G m
0l5lC1###,0m >l5nlC1$
++( D+( '-+#,A l5 +( '0D1#
!#F>5@A$ lC0)&,1+#
I&,'G'+'D#+,A>567 lC(,
+ '+#!$.'l5
',J*3K),'), <+#', ((+(+( '-+#,(
/#''*
'L>5$M9N0AO=$=N#1P '
))
E
G>DB7Q=$N5R DQ=$N 09$551 ;(+
,
/ ###$
!"###!$
.'
#'A> 69'#+###+8 +( '-+#,(
5 ,
'l5 #lC((+(
P <+#',/
E #''' )
G>DB9Q=$R5C DQ=$N 09$5C1 ),/#*
# #$%( ##
%! ,((## !"#'-+#,()(# !#''
+#',,&+':'!+) $
+&+'G'* T(!)2#
!&$S'+!+(++#+&+'G #(')2'(+
T,,!++,#* 0)+)#)
&#+$%()&& *)1)!#'''(
!#))#! $%!###+
01 2 345674849
49 549
S90R9
!"# $%&' '(/0B
(#!)*#+,&' #>1('(
( (@9'##9
#+-&'(! ## /0#(
.(/ >'##R@!T1
#'*#0 1#
/1! /09/ ##00
2'345 6+,7+-6 #9'#34!
8,9:;6 +, ':;:+! S/('
./8!;<=68, ##4 /0
/&'# # !)'0/('
>?@%A $;!.9+, $; >U9+--V9#!;;-@W
##!.B Z Y[ ]
/8, ;
X6 Y; \>;\@=; >8!;$@
#C->+D$%@!.(
EBF#//B
(*'!2'GH!>8!+@98,
/#9C-
/#9IH;J+9E6+!K,,;!
.EB/F#//( [Y6 /#
6#
0'/(-!--+,!.# \6 ##'#>1'
#0/ #40 ##
! /@
=6/&'![\*
LM
NM4984
OPQO W
^= `_aY
.0/(3 \6 _6;= >8!;8@
#0#
#!)##( `_6/&'
>!'!9##''/9(B ##!
''(@( 3b/#!
'B'*/(349( [*
(/# !.
/B/ 9 c =
1#(>0!@ [6 >>`_aY@\@+ >8!;%@
'#!"'34##B _6;
#'/#9
'###B d'0!e
'*0!. /B/ */
/B/# #9/ #/!.('
1&/ 'R!.0
100 '(/B/ /
(#! 0'!)/>./d@9
./B/ #B /#' /
#/ 1//B &'9# /
!./#9/( H9/&'>/
/R#'! #/#@!./&
012 344564 78
9
99 c$ &9
9&9

9 9
9 9!9<9$!9
!99 " & %99 99$
#99$
9
9
9 99%9$ $ 9
$&$ & 9
9
&$ '99( )**+,,-". ]\W
![9 ,X;+"&@d
012 3 9 &? ]\W
!9,X;;"<
'/ 4)2331," -,+" 9&
999! 9$$
9
9
&
! 9 9C99
$
0 9
& %
#
$ B9
]\W
!,XX*"&
5 2 7 13 8") 9 &A9A 99$ 9
0/ 8,18" , -,9" &99$

$ 9 9 &9

9
$ 9
9$$
6/,
8 9
9 9% 9 &%99$ & 9
99!
D99)***"
$&&9&$
52 &9
%99 &99 $ 9
e
8/ 76:23 -,;" $$ 9

!B ?a 9$

& 9 9B
$%9
6/,
&9
$f 9)**+"9
#
9& 9
'99 9
9
9 &9
9 `9 `
!% %9'9
9
9
99
&&9999)Agd99 $

9
9999
99
! $99 9 &!9
@d$9"
& 99 9
$9&9
$999

$
99!9 9 &9
9 9
9!
<9
$= & 9A9)**9"99
99! &&&99& %9 9!$ 9
9 !
9
><&9
'99 %99
9 )Agd99h9
9 9 9?99
,XXX"9&&9& B"
9@99
9$!
&&9 ! && =9 $ 9
9&
9
= $99
9 99#

&&99

99AB&9

& & =9 9
C 99D$
)**E" 9!
#99
9
9 99
9$
FGHGIJKLMNOPQRRSTUV 9
9
%9#
!
9 9
$
9
%9 99

B99
9 &9
&B9
W
$ 99 9

99 $9
99 "9
9!
9& $& 9 9
9!9 9
99(!9,XXYZW9[ 9
)**9"A$99! 999
99 &9
9$ &&9 &9
9 FGHGFijMMjklmVUKnSVlKoSmkUpS
$ 9
9
9\ $9
99
! &&9 qkUSLPLSUKUmjkjRrmVTLmMmkKUmjk
]&999$!"9
!99!9 sSVUV
%99
$!$! $

9
9 &
&#
9 & =9
!&9
9999$ 9
^$ <& 9 99
&!
_ 9 B9 &99 A $ 9
9$9
! 9 9%
$$ &99&9
9
99
9$$99 ! &&9
&9
`9 99a!b$9=,XE*" 9
9
9 & 9
01 2 345674849
49 549
-/5.V$5W
!"#$#%& 1#$2-&.
''& "
% (%$') &1""
& ''
&%$ 1"& 5'/
#%&)#& /)&&Q
$#*$# $"%'5&
"+$# 3X&
''#"& , $&%"%'V&
&&-%. W&V##
%'&&/ &&"'&-&.W,
'"0 '&/)&&
1$%#&/ '%$'%
&0&"0/ V#Y*,&&W1
%, %&&'#ZY
2&&2&)#(%$ !25$/5[&
&& /#'$%#&
3%#' V\$]$^$&_
''#%#&&#'& )#$WR
)#4+$5#
&')#$#&
&"''
67789:;<=6>;?7@8677ABCDE
FBGC9:@;9HFE86IJBFK6I
C9:>C?8LM;NN8A89FO8PFP
Q%$
,
&'#&'#4
+$#2RS3
'#&'&&/ Y$"%"
$3&32T '%#",#/
2"&U,# "#&
)&&&# &,#&`\&`_'%
(%$# ,)"0#
&2V0&W/ `]V-
&V% 5.&-5.W
&2#&+#&&& &%`^ 3%'
W $$
3&"&'$"% %&& Q
55$/5 -/5.&#"a
&V-.WX&, "&+5&'2$
)' %%&"
-/5." "&,&%'
0121314516 77
89
9999
9!9!"#$
9!%& #$9!%'
9!(
)*+##,# +,-.
/#
) 9 09 9&9
12!3!!$4 9
0567&
! 89
:
9; 4(
<!0:191:09!98=

3 > 9 9 (99>
!:? ?! @
50!$9
(
HIJIKILMIN
:&:&O0$9&P P9&QR-SB8
O99@3!9: &
T9;&T
:&: P9&QR-S6U&9@
3!9&89 Q 9U'V&8V9&
W:
:99&XQ&Y9&< <3&:P-SB6
!959O!998;&
B&-7C-B7
:99&X-SZS<9902!9!9!9
9@9 99 090@95
9W8 V39!&7&-8C-B-
:0!9&E&"&'18 E[9@&==Z89@ 1
&93 9
1!V99 \!@
O&-6&-S=C-S8
R$@&E&9;&P&<
&X&W&X E[9@&
-SSV9! 099
$
90! 99091
*+7=,- = +,- .
X9!989@8! &6&7CBZ
) 7=/-= A-B6
067 R9;V&9< -S6=O9@09 9@ 0X9!
9 \!@&&-C-=
99! 05 R&:
X Q&<==-89 9:19
9 X9!989@8! &-B&-8CZ
$
99 R&X Q&<==-$9
9:19:
; 9$
9 R&X 9 X9!989@8! &-B&7C8S
C 99 8==B8

O
9@<9 !(
&O9 ! $R;

$ (
D !
99 R9;9&OR9==9
O !$
0O &:&D:
9

&9$$@ 939 9 R9V;999 \! @ O&-7&7=8C7-Z
D"E99 9 &OR W&P'R==S!9
@9!00512!$!F
90 99 9@ 90
9&G90
90
@9!! @ 9(-=-=-B]99 92!!9
X V99 \!@ O&
==S=7==
>@$@9!
9 &94?:9& <9 9&W&8!309&V '<!@&===Q
999- 7&91 0 94395@9
1!9
X9!989@8! &-8&-8-C-B
$!99;9 0!0!& Q&< !&^-S6B99
9@$@39F@G94 9@ 9W8&--&8-C8
011 2 345674849
49 549
!"!#$%!&'(&() (%,,*T/I""%("'7&("((7&'()
$*+!',"-!#-(+$./012 !',(('!$'$$$(($+)(!('(
3!4(!*516'!'("((!'7+,'!&($'$ $""!%((*+!',"3$%,7#2I01
(!,'("(+$&',!8'8,(%&$, Q%,,H<:,7$,V>9(B,%';'!U+($
((!)',(!'(9"!%$%%7',$(%((! :9!$7T>:/1B!("-%(P%>9(7'(
:!(*+!',"'(&'(%','$-('((%',;#%,7# !$$(6('7:EQ L!E1
<<<0<= JG'#'$F,$(C5.I""%(9")
3!4(!*5.=-!#$""!%((7'$( !#$""!%((?-U+(',(9(#'',#"!,U+$
&'+!&("!#$%!&'8,(#>?*5;77(( "$(&+,*+!',"3$-%%///20//
@$A-!#B',#"3$,9!B,$-%% JG'#T7-L'$J$8!(Q.I-!#$")
;+8,%'(C$10=2 "!%((?-&!(E7%%!7(7!',!+,
3!4(!*5D(!'B'$6!4E;3F.2 ('(&!(9((+" !(&+,C8&((,
'8,"$G"!((!'7+,'!&($'$()B3H7) T%'"(+$%,70/
',$(%(!%$+!;!%('$;#%#%<1 JG'#&S'$F,$(C5..
1I01. '!&)+""%(''&'"%!'7($%!&'8,)
F'%+,'H'$-7*.=-('((%',&($"$ (#"!#$""!%((*+!',"3$-%%/
'$%+&!!'!%B%'$&%J!,'$3C .=.0./2
K'!!K'$D',&+K=&'+!&("('( ;'78!5'$+E,#TCII-!#$%!&')
(9(# ( #,(+!'B',"+7%/ ("&,E',($',#W'8,%((+('$',77+&
<=0 &,E*+!',"'!#-%%=0I
K+'7L)'$C' ,K.-(9(#"(B:M ;!#'&5/.-!#$""!%((*+!',"3$
$%!&'(((*+!',"-!#-(+$<<0<N %,7#<<0?O!&(',%P!&') ;!%*5'$F,8!(Q/.JBM'$B:M,&
("%7(9(!'(7"!$""!%(('$""%( *+!',"(B%+(%',-%(#"B&!%'2/0//
"!%(+',9'!'%*+!',"-!#-(+$< 5,!:;'78!5-$,*C'$-(K
=I/0=.= 1.O'$$('((%',('8,"!(&'(77P%'%
D!'&!KH'$&'-.1K &'#+84%(? '!$)!"!%'!$$""!%$+)(!'$(!'7,
-('((%', !'',#!'!%-'7Q 8+!#;'!E ((*+!',"3$-%%==20=
HB 5+'+:57'+O'$JG'#+('!$
C' ,K.2B%&'!"$""!(&($+$ $%!&'(8#'&)$""!('$(!'7,((?'%(
('(9(#((('("#,(%'!8'&$ "!!('(&&!#'$('+%!(!'3$R+',(#'$
@;HAH&%',-/<=10</I ;!"!%210.=
CK)-'$JG'#<2219,+("'&$,? -%,%;5E('8,"!$%!&'(((*+!',"
B!9 "+!(''$%$(',(&+,+""%( 3$R+',(#'$;!"!%==0/
$""!%((73$R+',(#'$;!"!%. -&(FC.-('((%',!!("&,!#$""!)
I0. %((?HP$%,&('$7P%'%((*+!',
C77((5B'$,%*3<22/8(')8&', "(-%%"3$'$B7!%+,(+!</0/<2
&$,?6'!'8,(#9!)$!'%!&($'$ -(K'$-$,*C1.H&+(7O'%(!8'8,(
9!(&*+!',"-!#-(+$<2=.0I !#$%!&'(((*+!',"3$-%%=
'%&,,'QBD','KC'$H!,&'H11 2<.02<
#%#%"%'(7!%',!%(;#%,7%', -(K'$-$,*C<22=-!#9',+'(;!'%(%
59 =/<0=1 !$$(B%'$&%,9!Q L!E
'%&,,'QB'$H!,&'H(%(!#? &S'$JG'#2$P%'((!#
BS!GF+$S9!(#;!H'&8!$7SD $""!%((!(%,? '!&$)+'!$%&'!)
'%!'BT/HP$%(!9',"!((!'7,(( (7,('$'!$$+)(!'$(B()B ((
%'79!'+!'%('(!$+%('!&,'!3$R+',(# &$P$"!!8'*+!',"-!#-(+$/
'$;!"!%II0I1 /01I
'(')F'!%'B7+,J'$JG'#<221J T(75+!!'#-H'(%X'$,,D<22=
'!&)+*+!',"-!#-(+$<<.10 +"!#$""!%(((9(7'(!%(8,
,7''!$H9,,H6'$H'!!:<22I-!# %,+!)$""!%'%&(%!$+%(H,!5'!%'$
9',+'(%U+3+!($(H5H:%'5'( B,%'(<<02=
3C
Chapter 6
Measurement of Sensory Thresholds
Abstract This chapter discusses the concept of threshold and contrasts the
conceptual notion with the idea of threshold as a statistically derived quantity. A
simple method for determining detection thresholds based on ASTM method E-679
is illustrated with a worked example. Other methods for determining thresholds are
discussed as well as alternative analyses.
A light may be so weak as not sensibly to dispel the darkness, a sound so low as not to be heard, a
contact so faint that we fail to notice it. In other words, a finite amount of the outward stimulus is
required to produce any sensation of its presence at all. This is called by Fechner the law of the
threshold—something must be stepped over before the object can gain entrance to the mind.
—(William James, 1913, p. 16)
Contents 6.11 Conclusions . . . . . . . . . . . . . . . . . 142

Appendix: MTBE Threshold Data for Worked
Example . . . . . . . . . . . . . . . . . . 143
6.1 Introduction: The Threshold Concept . . . . 125 References . . . . . . . . . . . . . . . . . . . . 145
6.2 Types of Thresholds: Definitions . . . . . . . 127
6.3 Practical Methods: Ascending Forced
Choice . . . . . . . . . . . . . . . . . . . 128
6.4 Suggested Method for Taste/Odor/Flavor
Detection Thresholds . . . . . . . . . . . . 129 6.1 Introduction: The Threshold Concept
6.4.1 Ascending Forced-Choice Method
of Limits . . . . . . . . . . . . . . . 129
6.4.2 Purpose of the Test . . . . . . . . . . 129 One of the earliest characteristics of human sensory
6.4.3 Preliminary Steps . . . . . . . . . . . 130 function to be measured was the absolute threshold.
6.4.4 Procedure . . . . . . . . . . . . . . . 131
6.4.5 Data Analysis . . . . . . . . . . . . . 131
The absolute or detection threshold was seen as an
6.4.6 Alternative Graphical Solution . . . . . 131 energy level below which no sensation would be pro-
6.4.7 Procedural Choices . . . . . . . . . . 133 duced by a stimulus and above which a sensation
6.5 Case Study/Worked Example . . . . . . . . . 133 would reach consciousness. The concept of threshold
6.6 Other Forced Choice Methods . . . . . . . . 134
6.7 Probit Analysis . . . . . . . . . . . . . . . 136
was central to Fechner’s psychophysics. His integra-
6.8 Sensory Adaptation, Sequential Effects, tion of Weber’s law produced the first psychophysical
and Variability . . . . . . . . . . . . . . . 136 relationship. It depended upon the physical inten-
6.9 Alternative Methods: Rated Difference, sity being measured with the threshold for sensing
Adaptive Procedures, Scaling . . . . . . . . . 137
6.9.1 Rated Difference from Control . . . . . 137
changes as the unit (Boring, 1942). Early physiolo-
6.9.2 Adaptive Procedures . . . . . . . . . . 138 gists like Weber and Fechner would use the classical
6.9.3 Scaling as an Alternative Measure method of limits to measure this point of disconti-
of Sensitivity . . . . . . . . . . . . . 140 nuity, the beginning of the psychophysical function.
6.10 Dilution to Threshold Measures . . . . . . . 140
6.10.1 Odor Units and Gas-Chromatography
In the method of limits, the energy level would be
Olfactometry (GCO) . . . . . . . . . . 140 raised and lowered and the average point at which
6.10.2 Scoville Units . . . . . . . . . . . . . 142 the observer changed response from “no sensation”
126 6 Measurement of Sensory Thresholds
to “yes, I perceive something” would be taken as the usefulness of a threshold is in defining a threshold
threshold. This specification of the minimum energy for taints or off-flavors in a product.
product Such a value
level required for perception was one of the first operat- has immediate practical implications for what may
ing characteristics of sensory function to be quantified. be acceptable versus unacceptable levels of undesired
Historically, the other common property to be mea- flavor components. Turning from the product to the
sured was the difference threshold or minimal increase sensory panelists, a third application of thresholds is
in energy needed to produce a noticeable increase in as one means of screening individuals for their sensi-
sensation. Together, these two measures were used to tivity to key flavor components. The measurement of a
specify the psychophysical function, which to Fechner person’s sensitivity has a long history in clinical test-
was a process of adding up difference thresholds once ing. Common vision and hearing examinations include
the absolute (minimum) threshold had been surpassed. some measurements of thresholds. In the chemical
In practice, some complications arise in trying to senses, threshold measurements can be especially use-
apply the threshold idea. First, anyone who attempts to ful, due to individual differences in taste and smell
measure a threshold finds that there is variability in the acuity. Conditions such as specific anosmia, a selec-
point at which observers change their response. Over tive olfactory deficit, can be important in determining
multiple measurements there is variability even within who is qualified for sensory test panel participation
a single individual. In a sequence of trials, even within (Amoore, 1971).
the same experimental session, the point at which a Another appealing aspect of the threshold concept
person changes his or her responses will differ. An old is that the values for threshold are specified in physical
story has it that S.S. Stevens, one of the pioneers of intensity units, e.g., moles per liter of a given com-
twentieth century psychophysics, used the following pound in a product. Thus many researchers feel com-
classroom demonstration at Harvard: Students were fortable with threshold specification since it appears
asked to take off their wristwatches and hold them at to be free from the subjective units of rating scales or
about arm’s length, then count the number of ticks sensory scores. However, threshold measurements are
they heard in 30 s (back in the day when spring- no more reliable or accurate than other sensory tech-
wound watches still made ticking sounds). Assuming niques and are usually very labor intensive to measure.
the watch of one of these Harvard gentlemen made Perhaps most importantly, thresholds represent only
uniform ticking sounds, the common result that one one point on a dose–response curve or psychophysi-
would hear some but not all of the ticks illustrated the cal function, so they tell us little about the dynamic
moment-to-moment variation in auditory sensitivity. characteristics of sensory response as a function of
Of course, there are also differences among individu- changes in physical concentration.
concentration How the sensory
als, especially in taste and smell sensitivity. This led to system behaves above threshold requires other kinds
the establishment of common rules of thumb for defin- of measurements.
ing a threshold, such as the level at which detection In this chapter, we will look at some threshold def-
occurs 50% of the time. initions and approaches and their associated problems.
The empirical threshold (i.e., what is actually mea- Next, we will examine some practical techniques for
sured) remains an appealing and useful concept to threshold measurement and discuss a few applications.
many workers involved in sensory assessments. One Throughout, we will pay special attention to the prob-
example is in the determination of flavor chemicals lems of variability in measurement and the challenges
that may contribute to the aromatic properties of a nat- that this poses for researchers who would use thresh-
ural product. Given a product like apple juice, many olds as practical measures of peoples’ sensitivities to
hundreds of chemical compounds can be measured a given stimulus, or conversely, of the potency or bio-
through chemical analysis. Which ones are likely to logical activity of that stimulus in activating sensory
contribute to the perceived aroma? A popular approach perceptions. Most of the examples chosen come from
in flavor analysis assumes that only those com- olfaction and taste, as the chemical senses are espe-
pounds that are present in concentrations above their cially variable and are prone to difficulties due to
thresholds will contribute. A second example of the factors such as sensory adaptation.
6.2 Types of Thresholds: Definitions 127
6.2 Types of Thresholds: Definitions It is not a sudden step function. There is a probability
function determined by an empirical method of mea-
surement on which we define some arbitrary point as
What is a threshold? The American Society for Testing
the threshold.
and Materials (ASTM) provides the following defini-
Recognition thresholds are also sometimes mea-
tion that captures the essence of the threshold concept
sured. These are the minimum levels that take on
for the chemical senses: “A concentration range exists
the characteristic taste or smell of the stimulus and
below which the odor or taste of a substance will
are often a bit higher than detection thresholds. For
not be detectable under any practical circumstances,
example, dilute NaCl is not always salty, but at low
and above which individuals with a normal sense of
concentrations just above the detection threshold is
smell or taste would readily detect the presence of the
perceived as sweet (Bartoshuk et al., 1978). The con-
substance.”—ASTM method E-679-79 (2008a, p. 36).
centration at which a salty taste is apparent from NaCl
Conceptually, the absolute or detection threshold is
is somewhat higher. In food research, it is obvious that
the lowest physical energy level of a stimulus or lowest
the recognition threshold for a given flavor in a food
concentration in the case of a chemical stimulus that is
would be a useful thing to know, and perhaps more
perceivable. This contrasts with empirical definitions
useful than detection thresholds, since both the percept
of threshold. When we try to measure this quantity, we
and the appropriate label have been made consciously
end up establishing some practical rule to find an arbi-
available and actionable to the taster. In the case of off-
trary value on a range of physical intensity levels that
flavors or taints, recognition may have strong hedonic
describes a probability function for detection. In 1908,
correlates in predicting consumer rejection.
the psychologist Urban recognized the probabilistic
To be recognized and identified, discrimination
nature of detection and called such a function a psycho-
from the diluent is only one requirement. In addition,
metric function, as shown in Fig. 6.1 (Boring, 1942).
the observer must assign the appropriate descriptor
We portray this function as a smooth curve in order
word to the stimulus. However, it is difficult to set up
to show how the original concept of a fixed threshold
a forced-choice experiment for identification in some
boundary was impossible to measure in practice. That
modalities. In taste, for example, you can have the
is, there is no one energy level below which detection
observer pick from the four (or five) taste qualities,
never occurs and above which detection always occurs.
but there is no assurance that these labels are suffi-
cient to describe all sapid substances (O’Mahony and
A Psychometric Function in
Threshold Measurement Ishii, 1986). Furthermore, one does not know if there
100 is an equal response bias across all four alternatives.
90
Thus the expected frequencies or null hypothesis for
statistical testing or difference from chance responding
PERCENT POSITIVE RESPONSE
80
is unclear. In an experiment on bitter tastes, Lawless
70 (1980) attempted to control for this bias by embed-
60 ding the to-be-recognized bitter substances in a series
that also included salt, acid, and sugar. However, the
50
success of such a procedure in controlling response
40
biases is unclear and at this time there are no estab-
30 lished methods for recognition thresholds that have
20 adequately addressed this problem.
50% positive response The difference threshold has long been part of clas-
10 = threshold concentration
sical psychophysics (see Chapter 2). It represents the
0 minimum physical change necessary in order for a per-
0 2 4 6 8 10 12 14 16 18 20
son to sense the change 50% of the time. Traditionally,
Concentration, millimolar
it was measured by the method of constant stimuli
Fig. 6.1 A psychometric function. (a method of comparison to a constant reference) in
which a series of products were raised and lowered the bitter side taste of saccharin. At high levels, the
around the level of the reference. The subject would bitterness will overtake the sweet sensation for some
be asked to say which of member of the pair was individuals. This makes it difficult to find a sweetness
stronger and the point at which the “stronger” judg- match of saccharin to other sweeteners at high levels
ment occurred 75% (or 25%) of the time was taken as (Ayya and Lawless, 1992). Further increases in con-
the difference threshold or “just-noticeable-difference” centration only increase bitterness and this additional
(JND). sensation has an inhibiting effect on sweet perception.
One can think of sensory discrimination tests (tri- So although saturation of response seems physiolog-
angles and such) as a kind of difference threshold ically reasonable, the complex sensations evoked by
measurement. The main difference between a psy- very strong stimuli mediate against any measurement
chophysical threshold test and a sensory discrimination of this effect in isolation.
test is that the psychophysical procedure uses a series Recently, a new type of threshold has been pro-
of carefully controlled and usually simple stimuli of posed for consumer rejection of a taint or off-flavor.
known composition. The sensory product test is more Prescott et al. (2005) examined the levels at which
likely to have only two products, and the pair is either consumers would show an aversion to cork taint from
deemed different or not, based on a criterion of sta- trichloroanisole in wines. Using a paired preference
tistical significance. But clearly the two kinds of tests tests with increasing levels of trichloroanisole, they
are related. Along these lines, one can think of the defined the rejection threshold as the concentration at
absolute threshold as a special case of a difference which there was a statistically significant preference
threshold, when the standard happens to be some blank for an untainted sample. This novel idea may find wide
or baseline stimulus (such as pure air or pure water). application in flavor science and in the study of specific
In addition to detection, recognition, and difference commodities (like water) in which the chemistry and
thresholds, a fourth category is the terminal thresh- origins of taints are fairly well understood (for another
old or region in which no further increase in response example, see Saliba et al., 2009). The method in its
is noted from increasing physical stimulus intensity original form requires some refinement as to the cri-
(Brown et al., 1978). In other words, the sensory terion for threshold because statistical significance is
response has reached some saturation level, beyond a poor choice. As they noted in their paper, the level
which no further stimulation is possible due to maxi- of statistical significance depends upon the number of
mal responding of receptors or nerves or some physical judges, a more or less arbitrary choice of the experi-
process limiting access of the stimulus to receptors. menter (not a function of the sensory response of the
This makes sense in terms of neurophysiology as well. participants). A better choice would be something akin
There are only a limited number of receptors and to the difference threshold, i.e., the concentration at
nerves and these have a maximal response rate. This which 75% preference was reached. Of course, con-
idea fits well with the notion of a threshold as a dis- fidence intervals can be drawn around any such level
continuity or inflection point in the psychophysical for those that need statistical assurance.
assurance
function (Marin et al., 1991).
However, in practice, this level is rarely approached.
There are few foods or products in which the satu-
ration level is a common level of sensation, although
6.3 Practical Methods: Ascending Forced
some very sweet confections and some very hot pep- Choice
per sauces may be exceptions. For many continua, the
saturation level is obscured by the addition of new In the early days of psychophysics, the method of
sensations such as pain or irritation (James, 1913). limits was the most common approach to measur-
For example, some odors have a down-turn in the ing thresholds. In this procedure, stimulus intensity
psychophysical function at high levels, as trigemi- would be raised in an ascending series and then low-
nal irritation begins to take place, that may in turn ered in a descending series to find points at which the
have an inhibiting effect on odor intensity (Cain, observer’s response changed from a negative to a posi-
1976; Cain and Murphy, 1980). Another example is in tive response or from positive to negative. Over several
6.4 Suggested Method for Taste/Odor/Flavor Detection Thresholds 129
Example of the Method of Limits the background level. A forced choice technique is
compatible with signal detection principles and is bias
32
free, since the observer does not choose whether or not
16 to respond—response is required on each trial.
Stimulus Intensity
1 6.4 Suggested Method for

0.5
Taste/Odor/Flavor Detection
.25
Thresholds
A D A D A D A D 6.4.1 Ascending Forced-Choice Method

Reversal
point: 2 1 4 4 2 2 4 4 of Limits
(log) (.3) (0) (.6) (.6) (.3) (.3) (.6) (.6)
This procedure is based on a standard method des-

Fig. 6.2 Method of limits example.
ignated ASTM E-679 (ASTM, 2008a). It follows the
classical method of limits, in which the stimulus inten-
ascending and descending runs, an average changing sity, in this case concentration of a taste or odor
point could be taken as the best estimate of threshold chemical, is raised in specified steps until the substance
(McBurney and Collings, 1977). This method is illus- is detected. The procedure adds a forced choice task in
trated in Fig. 6.2. which the substance to be detected is embedded in a set
Although this procedure seems straightforward, it of stimuli or products that includes other samples that
has several problems. First, the descending series do not contain any of the added substance. The stimu-
may cause such fatigue or sensory adaptation that lus or product with the taste or odor chemical is called
the observer fails to detect stimulus presentations that a “target” and the other items with no added chemical
would be clearly perceived if they were presented are often referred to as “blanks.” One can use various
in isolation. To avoid the adaptation or fatigue prob- combinations of targets and blanks, but it is common
lem that is common in the taste and smell senses, to have one target and in the case of E-679, two addi-
the method is usually performed only in an ascend- tional blanks. So the task is a three-alternative forced
ing series. A second difficulty is that different persons choice task (3-AFC), because the person being tested
may set different criteria for how much of a sensa- is forced to choose the one different sample in the set
tion they require before changing their response. Some of three. That is if they are uncertain, they are told to
people might be very conservative and have to be pos- guess.
itively sure before they respond, while others might
take any inkling at all as a reason to report a sensation.
Thus the classical method of limits is contaminated
by the panelist’s individual bias or criterion, which is 6.4.2 Purpose of the Test
not a function of their sensitivity, i.e., what the test
is actually trying to measure. This is a central issue This method is designed to find the minimum level
in the theory of signal detection (see Chapter 5). To (minimum concentration) of a substance that is
address the problem of uncontrolled individual crite- detected by 50% of the sample group. In practice, this
rion, later workers introduced a forced choice element is calculated as the geometric mean of the individual
to the trials at each intensity level or concentration step threshold estimates. The geometric mean is a reason-
(e.g., Dravnieks and Prokop, 1975). This combines able choice because it is often very close to the median
the method of limits with a discrimination test. The (50th percentile) of a positively skewed distribution.
task requires that the observer gives objective proof Threshold data tend to show high outliers, i.e., some
of detection by discriminating the target stimulus from insensitive individuals cause positive skew.
6.4.3 Preliminary Steps population, then the panel should be representative of

that population with respect to age, gender, and so on
Before the test is conducted, there are several tasks and and a larger panel of 100 or more is recommended. It is
some choices that must be made, as shown in Table 6.2. common practice to exclude people with known health
First, a sample of the substance of known purity must problems that could affect their sense of taste or smell
be obtained. Second, the diluent (solvent, base) or car- and individuals with obvious sensory deficits in the
rier must be chosen. For the detection threshold for modality being tested. Of course, all the appropriate
flavors, for example, it is common to use some kind setup work must be done that is associated with con-
of pure water such as deionized or distilled. Third, the ducting any sensory test, such as securing a test room
size of the concentration steps must be chosen. It is that is free from odors and distractions, scheduling the
common to use factors of two or three. In other words, panelists, setting up the questionnaire or answer sheet,
the concentrations will be made up in a geometric pro- writing instructions for the participants. See Chapter 3
gression, which are equal steps on a log scale. Fourth, for further details on good practices in sensory testing.
some sample concentrations should be set up for pre- For threshold work it is especially important to have
liminary or “benchtop screening” to estimate the range clean odor-free glassware or plastic cups that are abso-
in which the threshold is likely to occur. This can be lutely free of any odor that would contaminate the test
done by successive dilutions using factors of five or samples. In odor testing the sample vessels are usually
ten, but beware the effects of adaptation on reducing covered to preserve the equilibrium in the headspace
one’s sensitivity to subsequent test items. Exposure to a above the liquid. The covers are removed by each
strong sample early in this series may cause subsequent panelist at the moment of sniffing and then replaced.
samples to seem odorless or tasteless, when they might Finally, external sources of odor must be minimized
in fact be perceived when tasted alone. The outcome of or eliminated, such as use of perfumes or fragrances
the preliminary test should bracket the likely concen- by participants, hand lotions, or other fragranced cos-
tration range, so that most, if not all, of the people who metics that could contaminate the sample vessels or
participate in the formal test will find an individual the general area. Avoid using any markers or writing
threshold estimate somewhere within the test series. instruments that might have an odor when marking
It is common to use about eight to ten steps in this the samples. As always, sample cups or vessels should
procedure. be marked with blind codes, such as randomly cho-
Next, the panel should be recruited or selected. A sen three digit numbers. The experimenter must set up
sample group should have at least 25 participants. If random orders for the three items at each step and use
the goal is to generalize the result to some larger a different randomization for each test subject. This
Table 6.1 Types of thresholds

Detection (absolute) threshold: Point at which the substance is differentiated from the background
Recognition threshold: Point at which the substance is correctly named
Difference threshold: (just-noticeable-difference, JND) Point at which the change in concentration is noted
Terminal threshold: Point at which no further intensity increase is found with increasing concentration
Consumer rejection threshold: Point at which a consumer preference occurs for a sample not containing the substance
Table 6.2 Preliminary tasks 1. Obtain test compound of known purity (note source and lot number)
before threshold testing
2. Choose and obtain the solvent, carrier, or food/beverage system
3. Set concentration/dilution steps, e.g., 1/3, 1/9, 1/27
4. Begin benchtop screening to bracket/approximate threshold range
5. Choose number of dilution steps
6. Recruit/screen panelists. N ≥ 25 is desirable
7. Establish procedure and pilot test if possible
8. Write verbatim instructions for panelists
6.4 Suggested Method for Taste/Odor/Flavor Detection Thresholds 131
should be recorded on a master coding sheet showing of the product of the two values). One value is the
the randomized three-digit codes and which sample is concentration at which they first answered correctly
the correct choice or target item. and all higher concentrations were also correct. The
other value is the concentration just below that, i.e.,
the last incorrect judgment. This interpolation provides
6.4.4 Procedure some protection against the fact that the forced-choice
procedure will tend to slightly overestimate the indi-
vidual’s threshold (i.e., the concentration at which they
The steps in the test are shown in Table 6.3. The partic-
have a 0.5 probability of sensing that something is dif-
ipant or test subject is typically seated before a sample
ferent from the blanks). If the subject gets to the top
tray containing the eight or so rows of three samples.
of the series with an incorrect judgment, or starts at
Each row contains one target sample and two blank
the bottom with all judgments correct, then a value is
samples, randomized. The instructions, according to
extrapolated beyond the test series. At the top, it is the
E-679-04 (ASTM, 2008a) are the same as in the trian-
geometric mean of the highest concentration tested and
gle test, that is to pick out the sample which is different
the next concentration that would have been used in the
from the other two. The subject is told to evaluate the
series if the series had been continued. At the bottom,
three samples in each row once, working from left to
it is the geometric mean of the lowest concentration
right. The test proceeds through all the steps of the con-
tested and the next lower concentration that would have
centration series and the answers from the subject are
been used had the series been continued lower. This
recorded, with a forced guess if the person is uncer-
is an arbitrary rule, but it is not unreasonable. Once
tain. According to E-679, if a person misses at the
these individual best estimates are tabulated, the group
highest level, that level will be repeated. If a person
threshold is the geometric mean of the individual val-
answers correctly through the entire series, the low-
ues. The geometric mean is easily calculated by taking
est level will also be repeated for confirmation. If the
the log of each of the individual concentration values,
response changes in either case, it is the repeated trial
finding the average of the logs, and then taking the
that is counted.
antilog of this value (equivalent to taking the Nth root
of the product of N observations).
6.4.5 Data Analysis
6.4.6 Alternative Graphical Solution
Figure 6.3 shows an example of how the data are ana-
lyzed and the threshold value is determined. First, an An alternative analysis is also appropriate for this kind
individual estimated threshold is determined for each of data set. Suppose that 3-AFC tests had been con-
person. This is defined as the concentration that is ducted and the group percent correct calculated at
the geometric mean of two values (the square root each step. For examples, see Antinone et al. (1994)
Table 6.3 Ascending forced-choice testing steps

1. Obtain randomized or counterbalanced orders via software program or random number generator.
2. Setup trays or other staging arrangements for each participant, based on random orders.
3. Instruct participants in procedure per verbatim script developed earlier.
4. Show suprathreshold example (optional).
5. Present samples and record results. Force a choice if participant is unsure.
6. Tally results for panel as series of correct/incorrect answers.
7. Calculate estimated individual thresholds: Geometric mean of first correct answer with all higher
concentrations correct and last incorrect step.
8. Take geometric mean of all individual threshold estimates to get group threshold value.
9. Plot graphic results of proportion correct against log concentration. Interpolate 66.6% correct point and drop
line to concentration axis to get another estimate of threshold (optional).
10. Plot upper and lower confidence interval envelopes based on ±1.96(p (1–p)/N). Drop lines from the upper
and lower envelopes at 66.6% to concentration axis to convert envelope to concentration interval.
Concentration ug/L
Panelist 2 3.5 6 10 18 30 60 100 BET log(BET)
1 + o + + + + + + 2.6 0.415
2 o o o + + + + + 7.7 0.886
3 o + o o o o + + 42 1.623
. . . . . . . . . . .
. . . . . . . . . . .
. . . . . . . . . . .
N o o + + o + o + 77 1.886
Prop. Mean
Corr. 0.44 0.49 0.61 0.58 0.65 0.77 0.89 0.86 (log(BET)) 1.149
10^1.149 = 14.093
Fig. 6.3 Sample data analysis from ascending 3-AFC method calculated from the geometric mean of the BET values. In prac-
Notes: Correct choices indicated by + and incorrect by o. BET, tice, this is done by taking the logs of the BET values, finding
Best estimate of individual threshold, defined as the geometric the mean of the logs (x), then taking the antilog of that value
mean of the first correct trial with all subsequent trials cor- (or 10x ).
rect and the previous (incorrect) trial. The group threshold is
and Tuorila et al. (1981). We can take the marginal if the data are fairly linear and a curve is fit by eye).
count of the number of correct choices from the bot- A useful equation that can be fit to many data sets is
tom row in Fig. 6.3 and expressing it as the proportion based on logistic regression, shown in Eq. (6.3) (e.g.,
correct. As the concentration increases, this propor- Walker et al., 2003).
tion should go from near the chance level (1/3) to
p
nearly 100% correct. Often this curve will form an ln = b0 + b1 log C (6.3)
S-curve similar to the cumulative normal distribution. 1−p
The threshold can then be defined as the level at which where p is the proportion correct at concentration C
performance is 50% correct, once we have adjusted the and b0 and b1 are the intercept and slope. The quantity,
data for chance, i.e., the probability that a person could p/1–p, is sometimes referred to as the odds ratio. The
guess correctly (Morrison, 1978; Tuorila et al., 1981; interpolation is shown in Fig. 6.4. Note that this also
Viswanathan et al., 1983). This is done by Abbott’s for-
mula, a well-known correction for guessing as shown Example Threshold Curve
100
in Eqs. (6.1) and (6.2):
Pcorr = (Pobs − Pchance )/(1 − Pchance ) (6.1)

80
where Pcorr is the chance-corrected proportion, Pobs is

the observed proportion correct in the data, and Pchance
Percent Correct
60
is the chance probability, e.g., 1/3 for the 3-AFC.
Another form is
40
Preq = (Pchance − Pcorr )/(1 − Pchance ) (6.2)
Chance (=33%)
where Preq is the observed proportion that is required 20

in order to achieve a certain chance corrected level
of performance. So if one needed to get a chance
D10 D25 D50
corrected proportion of 0.5 (i.e., a threshold, 50% 0
detection) in a 3-AFC test, you would need to see –0.5 0 0.5 1 1.5 2 2.5
1/3 + 0.5 (1–1/3) or 2/3 (= 66.7%) correct. Log Concentration
Once a line or curve is fitted to the data, the con- Fig. 6.4 Sample threshold curve and interpolation. D10, D25
centration at which the group would achieve 66.6% and D50 show interpolated detection levels for 10%, 25% and
correct can be solved (or simply interpolated by eye 50% of persons, respectively.
6.5 Case Study/Worked Example 133
allows one to estimate percentages of the population Re-tasting could either confuse the subjects or it might
that would detect with other probabilities and not just help them get a better idea of which item is the target.
the arbitrary 50% detection that we use as the thresh- We would generally argue against re-tasting, because
old value. That is, one could interpolate at 10 or 90% that will introduce a variable that is left up to the
detection if that was of interest. A lower percentage of individual subject and will thus differ among people.
detection might be of interest, for example, in setting Some will choose to re-taste and others will not. So,
level to protect consumers from an off-flavor or taint. on the basis of maintaining a consistent test proce-
This graphical method has certain assumptions and dure across all participants, re-tasting is generally not
limitations that the user should be aware of. First, it recommended.
assumes that persons are either detecting or guess- Another important choice is that of a “stopping
ing (Morrison, 1978). In reality, every person has an rule.” In the published version of E-679, every subject
individual threshold gradient or gradually increasing must continue to the top of the series. There are some
probability of detection around their own threshold. pitfalls in this, because of the possibility that the senses
Second, the model does not specify what percent of the will become fatigued or adapted by the high levels at
time that a given percentage of the group will detect. the top of the series, especially for an individual with a
In the data set examined below, the ASTM method low personal threshold. For this reason, some threshold
and the graphical solution provide a good estimate of procedures introduce a “stopping rule.” For example,
when 50% of the group will detect 50% of the time. the panelist may be allowed to stop tasting after giving
More extensive statistical models have been developed three correct answers at adjacent levels (Dravnieks and
for this kind of data and an extensive paper on alter- Prokop, 1975). This prevents the problem of exposing
native statistical analyses is given in USEPA (2001), a sensitive individual to an overwhelming stimulus at
again using the data set we have chosen as an example high levels. Such an experience, if unpleasant (such
below. as a bitter taste), might even cause them to quit the
test. On the downside, the introduction of a stopping
rule can raise the false positive rate. We can think of a
false positive as finding a threshold value for an indi-
6.4.7 Procedural Choices vidual that is due to guessing only. In the most extreme
case, it would be a person who is completely insen-
Note that although the instructions are the same as sitive (e.g., anosmic to that compound if it is an odor
in the triangle test, all the possible combinations of threshold) finding a threshold somewhere in the series.
the three samples are not used, i.e., the test is not a With an eight-step series, for the ASTM standard rule
fully counterbalanced triangle. Only the three possible (everyone completes the series), the probability of find-
orders that are given by combinations of two blanks ing a threshold somewhere in steps one through eight,
and one target are used. In a fully counterbalanced tri- for a completely anosmic person who is always guess-
angle, the additional three combinations of two targets ing is 33.3%. For the three-in-a-row stopping rule, the
and one blank would have been used (thus a total of six chances of the anosmic person making three lucky
possible), but this is not done according to E-679. For guesses in a row somewhere rise above 50%. The sen-
taste or flavor, there is generally no rinsing between sory professional must weigh the possible negatives
samples, although testers may be instructed to rinse from exposing the participant to strong stimuli against
between rows (triads). If possible, it is wise to give the increased possibility of false positives creating a
the subject a preliminary sample at a detectable level, low-threshold estimate when using a stopping rule.
in order to show them the target item that they will
be trying to sense in the test. Of course, one must be
careful when using such an above-threshold sample so
that it does not adapt or fatigue the senses. An appro- 6.5 Case Study/Worked Example
priate waiting time and/or rinsing requirement should
be used to prevent any effect on the subsequent test For the ascending forced choice method of limits
samples in the formal test. The experimenters should (ASTM E-679), we can use a published data set for
also decide whether they will allow re-tasting or not. odor thresholds. The actual data set is reproduced in
the Appendix at the end of this chapter. The data are chance-adjusted level for 50% probability of detec-
from a study conducted to find the odor detection tion in the group. Confidence intervals (CI) for this
threshold for methyl tertiary butyl ether (MTBE), a level can be found by constructing upper and lower
gasoline additive that can contaminate ground water, curves form an envelope of uncertainty around the fit-
rendering some well waters unpotable (Stocking et al., ted curve. The standard error is given by the square
2001; USEPA, 2001). The ASTM procedure was fol- root of (p(1–p)/N) or in this case 0.062 for p = 1/3 and
lowed closely, including the triangle test instructions N = 57.
(choose the sample different from the other two), using The 95% CI is found by multiplying the z-score
the 3-AFC in eight concentration steps differing by for 0.95 ( = 1.96) times the standard error, in this
a factor of about 1.8. Individual best estimates were case equal to ± 0.062(1.96) or ± 0.122. Constructing
taken as the geometric mean of the last step missed curves higher and lower than the observed proportions
and the first step answered correctly, with all higher by this amount will then permit interpolation at the
steps also correct. Individuals who got the first and all 66.7% level to find concentrations for the upper and
subsequent steps correct (there were 10/57 or 17.5% of lower CI bounds. This method is simple, but it pro-
the group) had their estimated threshold assigned as the vides conservative (wider) estimate of the confidence
geometric mean of the first concentration and the hypo- intervals that found with some other statistical methods
thetical concentration one step below that which would such as bootstrap analysis (USEPA, 2001). Another
have been used had the series been extended down. method for error estimation based on the standard error
A similar extrapolation/estimation was performed at of the regression line is given in Lawless (2010).
the high end for persons that missed the target on the Note that by the graphical method, the interpo-
eighth (highest) level. lated value for 10% detection (= 40% correct by
The geometric mean of the individual threshold Abbott’s formula) will be at about 1–2 ppb. Similarly
estimates across a panel of 57 individuals, balanced the interpolated value for 25% detection (50% correct
for gender and representing a range of ages, was by Abbott’s formula), will be between 3 and 4 ppb.
14 μg/l (14 ppb). Figure 6.5 shows the graphical solu- These values are practically useful to a water company
tion, which gives a threshold of about 14 ppb, in who wanted to set lower limits on the amount of MTBE
good agreement with the geometric mean calculation. that could be detected by proportions of the population
This is the interpolated value for 66.7% correct, the below the arbitrary threshold value of 50% (Dale et al.,
1997).
100
90 Y = 36.2 + 26.4X, R2 = 0.933

6.6 Other Forced Choice Methods
80
Ascending forced-choice procedures are widely used
Percent Correct
70 techniques for threshold measurement in the experi-

mental literature on taste and smell. One early exam-
60
ple of this approach is in the method for determin-
ing sensitivity to the bitter compound phenylthiourea,
50
formerly called phenylthiocarbamide or PTC, and
40
the related compounds 6-n-propylthiouracil or PROP.
Approximately one-third of Caucasian peoples are
30 insensitive to the bitterness of these compounds, as a
L.C.I. U.C.I. function of several mutations in a bitter receptor that
20 usually manifests as a simple homozygous recessive
0 0.5 1 Thr. 1.5 2 2.5
status for this trait (Blakeslee, 1932; Bufe et al., 2005).
Log (C)
Early researchers felt the need to have a very stringent
Fig. 6.5 Interpolation of threshold from the data of Stocking test of threshold, so they intermingled four blank sam-
et al. (2001). ples (often tap water) with four target samples at each
6.6 Other Forced Choice Methods 135
concentration step (Harris and Kalmus, 1949). The variation was used by Stevens et al. (1988) in a land-
chance probability of sorting correctly is only 0.014, mark paper on the individual variability in olfactory
so this is a fairly difficult test. In general, the formula thresholds. In this case, five correct pairs were required
for the chance probability of sorting any one level of X to score the concentration as correctly detected, and
target samples among N total samples is given by Eq. this performance was confirmed at the next highest
(6.4): concentration level. The most striking finding of this
study was that among the three individuals tested 20
p = X!/[N!/(N − X)!] (6.4) times, their individual thresholds for butanol, pyridine,
and phenylethylmethylethyl carbinol (a rose odorant)
Obviously, the larger the number of target and blank varied over 2,000- to 10,000-fold in concentration.
samples, the more stringent the test and the higher Variation within an individual was as wide as the varia-
the eventual threshold estimate. However, arbitrarily tion typically seen across a population of test subjects.
increasing X and N may make the task tedious and This surprising result suggests that day-to-day varia-
may lead to other problems such as fatigue and flag- tion in olfactory sensitivity is large and that thresholds
ging motivation among the participants. The rigor of for an individual are not very stable (for an exam-
the test estimate must be weighed against undue com- ple, see Lawless et al., 1995). More recent work using
plexity that could lead to failures to complete the series extensive testing of individuals at each concentration
or poor quality data. step suggests that these estimates of variability may
Another example of a threshold test for olfaction be high. Walker et al. (2003) used a simple yes/no
is Amoore’s technique for assessing specific anos- procedure (like the A, not-A test, or signal detection
mia (Amoore, 1979; Amoore et al., 1968). Specific test) with 15 trials of targets and 15 trials of blanks at
anosmia describes a deficit in the ability to smell a each concentration level. Using a model for statistical
compound or closely related family of compounds significant differences between blank and target trials,
among people with otherwise normal olfactory acuity. they were able to get sharp gradients for the individual
Being classified as anosmic was operationally defined threshold estimates.
by Amoore as having olfactory detection thresholds In summary, an ascending forced-choice method is
more than two standard deviations above the popula- a reasonably useful compromise between the need to
tion mean (Amoore et al., 1968). The test is sometimes precisely define a threshold level and the problems
called a “two-out-of-five” test because at each concen- encountered in sensory adaptation and observer fatigue
tration level there are two target stimuli containing the when extensive measurements are made. However, the
odorant to be tested and three diluent or blank con- user of an ascending forced-choice procedure should
trol samples. The tester must sort the samples correctly be aware of the procedural choices that can affect the
in this two-out-of-five test, and the chance probability obtained threshold value. The following choices will
of obtaining correct sorting by merely guessing is one affect the measured value: the number of alternatives
in ten. Performance is normally confirmed by testing (both targets and blanks), the stopping rule, or the
the next highest concentration (an example of a “stop- number of correct steps in a row required to estab-
ping rule”). The chance occurrence of sorting correctly lish a threshold, the number of replicated correct trials
on two adjacent levels is then 1 in 100. This makes required at any one step, and the rule to determine at
the test somewhat difficult but provides a good deal of what level of concentration steps the threshold value is
insurance against a correct answer by guessing. assigned. For example, the individual threshold might
Another way to reduce the chance performance on be assigned at the lowest level correct, the geomet-
any one level is to require multiple correct answers at ric mean between the lowest level correct and highest
any given concentration. This is part of the rationale level incorrect. Other specific factors include the cho-
behind the Guadagni multiple pairs test (Brown et al., sen step size of concentration units (factors of two or
1978) in which up to four pairs may be given for a three are common in taste and smell), the method of
two-alternative forced choice test in quadruplicate at averaging or combining replicated ascending runs on
any one concentration. Brown et al. commented upon the same individual and finally the method of aver-
the user-friendliness of this technique, i.e., how simple aging or combining group data. Geometric means are
it was to understand and administer to participants. A commonly used for the last two purposes.
6.7 Probit Analysis 97.5 7.0
95
(90)
It is often useful to apply some kind of transforma- 90
tion or graphing method to the group data to linearize (80) (85) 6.0
the curve used to find the 50% point in a group. Both 80
Cumulative Percentage
the psychometric curve that represents the behavior of 70
(57)
an individual in multiple trials of a threshold test and 60
PROBIT
the cumulative distribution of a group will resemble 50 5.0
(40)
an S-shaped function similar to the cumulative nor- 40
mal distribution. A number of methods for graphing 30
such data are shown in the ASTM standard E-1432 (18)
20
4.0
(ASTM, 2008b). One simple way to graph the data is
10
simply to plot the cumulative percentages on “prob- (6)
ability paper.” This pre-printed solution provides a 5
graph in which equal standard deviations are marked 2.5 3.0
off along the ordinate, effectively stretching the per- 1 2 4 8 16 32 64
Threshold Concentration (log scale)
centile intervals at the ends and compressing them in
the midrange to conform to the density of the normal Fig. 6.6 An example of Probit analysis. Numbers in paren-
distribution. Another way to achieve the straightening theses are the cumulative percentages of panelists reaching
of the S-shaped response curve is to transform the data threshold at each concentration step. Note the uneven scale on
by taking z-scores. Statistical packages for data anal- the left axis. Probits mark off equal standard deviations and are
based on the z-score for any proportion plus the constant, 5.
ysis often provide options for transformation of the Interpolation at the 50% or Probit 5.0 gives the threshold value.
data.
A related method was once widely used in threshold
measurement, called Probit analysis (ASTM, 2008b; 6.8 Sensory Adaptation, Sequential
Dravnieks and Prokop, 1975; Finney, 1971). In this
approach, the individual points are transformed relative
Effects, and Variability
to the mean value, divided by the standard deviation
and then a constant value of +5 is added to translate all Individual variability, both among a group of people
the numbers to positive values for convenience. A lin- and within an individual over repeated measurements
ear fitted function can now be interpolated at the value presents a challenge to the idea that the threshold
of 5 to estimate the threshold as in Fig. 6.6. The conver- is anything like a fixed value. For example, stable
sion (to a z-score +5) tends to make an S-shaped curve olfactory thresholds of an individual are difficult to
more linear. An example of this can be found in the measure. The test–retest correlation for individual’s
paper by Brown et al. (1978), using data from a mul- olfactory threshold is often low (Punter, 1983). Even
tiple paired test. First the percent correct is adjusted within an individual, threshold values will generally
for chance. Then the data are transformed from the decrease with practice (Engen, 1960; Mojet et al.,
percent correct (across the group) at each concentra- 2001; Rabin and Cain, 1986), and superimposed upon
tion level by conversion to z-scores and a constant of this practice effect is a high level of seemingly ran-
5 is added. The mean value or Probit equal to 5 can dom variation (Stevens et al., 1988). Individuals may
be found by interpolation or curve fitting. An exam- become sensitive to odorants to which they were for-
ple of this technique for estimating threshold from a merly anosmic, apparently through simple exposure
group of 20 panelists is shown in Meilgaard et al. (Wysocki et al., 1989). Increased sensitivity as a func-
(1991) and in ASTM (2008b). Probit plots can be tion of exposure may be a common phenomenon
used for any cumulative proportions, as well as ranked among women of childbearing age (Dalton et al., 2002;
data and analysis of individuals who are more exten- Diamond et al., 2005).
sively tested than in the 3-AFC method example shown Sensory adaptation and momentary changes in
earlier. sensitivity due to sequences may have occurred in
6.9 Alternative Methods: Rated Difference, Adaptive Procedures, Scaling 137
the experiments of Stevens et al. (1988) and could Brown et al. (1978) stated that for any test compound, a
have contributed to some instability in the measure- number of insensitive individuals would likely be seen
ments. As predicted by sequential sensitivity analysis in the data set, when 25 or more persons were tested to
(Masuoka et al., 1995; O’Mahony and Odbert, 1985) determine an average threshold. Among any given pool
the specific stimulus sequence will render discrimina- of participants, a few people with otherwise normal
tion more or less difficult. After the stronger of two smell acuity will have high thresholds. This is poten-
stimuli, an additional second strong stimulus presented tially important for sensory professionals who need to
next may be partially adapted and seem weaker than screen panelists for detection of specific flavor or odor
normal. Stevens et al. remarked that sometimes sub- notes such as defects or taints. In an extensive survey
jects would get all five pairs correct at one level with of thresholds for branched-chain fatty acids, Brennand
some certainty that they “got the scent” but lost the sig- et al. (1989) remarked that “some judges were unable
nal at the next level before getting it back. This report to identify the correct samples in the pairs even in the
and the reversals of performance in threshold data are highest concentrations provided” and that “ panelists
consistent with adaptation effects temporarily lessen- who were sensitive to most fatty acids found some
ing sensitivity. The sensory impression will sometimes acids difficulty to perceive” (p. 109). Wide variation
“fade in and out” at levels near threshold. in sensitivity was also observed to the common fla-
In attempts to avoid adaptation effects, other vor compound, diacetyl, a buttery-smelling by-product
researchers have gone to fewer presentations of the of lactic bacteria fermentation (Lawless et al., 1994).
target stimulus. For example, Lawless et al. (1995) Also, simple exposure to some chemicals can modify
used one target among three blank stimuli, a 4-AFC specific anosmia and increase sensitivity (Stevens and
test that has appeared in previous studies (e.g., Engen, O’Connell, 1995).
1960; Punter, 1983). This lowers the chance perfor-
mance level and lessens the potential adaptation at any
one concentration step. To guard against the effects 6.9 Alternative Methods: Rated
of correct guessing, threshold was taken as the low- Difference, Adaptive Procedures,
est concentration step with a correct choice when all Scaling
higher concentrations were also correct. Thresholds
were measured in duplicate ascending runs in a test
session, and a duplicate session of two more ascend- 6.9.1 Rated Difference from Control
ing runs was run on a second day. Correlations across
the four ascending series ranged from 0.75 to 0.92 for Another practical procedure for estimating threshold
cineole and from 0.51 to 0.92 for carvone. For car- has involved the use of ratings on degree-of-difference
vone, thresholds were better duplicated within a day scales, where a sample containing the to-be-recognized
(r = 0.91 and 0.88) than across days (r from 0.51 stimulus is compared to some control or blank stimu-
to 0.70). This latter result suggests some drift over lus (Brown et al., 1978; Lundahl et al., 1986). Rated
time in odor thresholds, in keeping with the variabil- difference may use a line scale or a category scale,
ity seen by Stevens et al. (1988). However, results ranging from no difference or “exact same” to a large
with this ascending method may not be this reliable difference, as discussed in Chapter 4. In these pro-
for all compounds. Using the ascending 4-AFC test cedures ratings for the sensory difference from the
and a sophisticated olfactometer, Punter (1983) found control sample will increase as the intensity of the tar-
median retest correlations for 11 compounds to be only get gets stronger. A point on the plot of ratings versus
0.40. The sense of taste may fare somewhat better. In a concentration is assigned as threshold. In some vari-
study of electrogustometric thresholds with ascending ations on this method, a blind control sample is also
paired tests requiring five correct responses, retest cor- rated. This provides the opportunity to estimate a base-
relations for an elderly population were 0.95 (Murphy line or false alarm rate based on the ratings (often
et al., 1995). nonzero) of the control against itself. Identical samples
In many forced-choice studies, high variability in will often get nonzero difference estimates due to the
smell thresholds is also noted across the testing pool. moment-to-moment variability in sensations.
In one application of this technique for taste and of the individual thresholds, due to the larger num-
smell thresholds, a substance was added in various lev- ber of observations, another oddity of using statistical
els to estimate the threshold in a food or beverage. significance to determine the threshold.
In each individual trial, three samples would be com- Instead of using statistical significance as a crite-
pared to the control sample with no added flavor—two rion, Marin et al. determined the point of maximum
adjacent concentration steps of the target compound curvature on the dose–response curve as the threshold.
and one blind control sample (Lundahl et al., 1986). Such an approach makes sense from consideration
Samples were rated on a simple 9-point scale, from of the general form of the dose–response (psychophys-
zero (no difference) to eight (extreme difference). This ical) curve for most tastes and odors. Figure 6.7 shows
provided a comparison of the control to itself and a a semi-log plot for the Beidler taste equation, a widely
cost-effective way of making three comparisons in one applied dose–response relationship in studies of the
set of samples. Since sample concentrations within the chemical senses (see Chapter 2). This function has two
three rated test samples were randomized, the proce- sections of curvature (change in slope, i.e., accelera-
dure was not a true ascending series and was dubbed tion) when plotted as a function of log concentration.
the “semi-ascending paired difference method.” There is a point at which the response appears to be
How is the threshold defined in these procedures? slowly increasing out from the background noise and
One approach is to compare the difference ratings for a then rises steeply to enter the middle of the dynamic
given level with the difference ratings given to the con- range of response. The point of maximum curvature
trol sample. Then the threshold can be based on some can be estimated graphically or determined from curve
measure of when these difference ratings diverge, such fitting and finding the maximum rate of change (i.e.,
as when they become significantly different by a t-test maximum of the second derivative) (Marin et al.,
(see Brown et al., 1978). Another approach is simply to 1991).
subtract the difference score given to the blind control
from the difference score give to each test sample and
Psychophysical function (semi-hyperbolic)
treat these adjusted scores as a new data set. In the orig-
100
inal paper of Lundahl et al. (1986), this latter method
was used. In the analysis, they performed a series of t-
Response (percent of maximum)
tests. Two values were taken to bracket the range of the 80

threshold. The upper level was the first level yielding
a significant t-test versus zero, and the lower level was 60
the nearest lower concentration yielding a significant t- R = (Rmax C) / (K + C)
K = 1000, Rmax = 100
test versus the first. This provided an interval in which
40
the threshold (as defined by this method) lies between
the two bracketing concentrations.
One problem with this approach is that when the 20
zone of inflection
threshold is based on the statistical significance of = threshold range
t-statistics (or any such significance test), the value of
0
threshold will depend upon the number of observa- 0 2 4 6 8
tions in the test. This creates a nonsensical situation log Concentration
where the threshold value will decrease as a function
Fig. 6.7 Beidler curve.
of the number of panelists used in the test. This is
an irrelevant variable, a choice of the experimenter,
and has nothing to do with the physiological sen-
sitivity of the panelist or the biological potency of
the substance being tested, a problem recognized by 6.9.2 Adaptive Procedures
Brown et al. (1978) and later by Marin et al. (1991).
Marin et al. also pointed out that a group threshold, Popular methods for threshold measurement for visual
based on a larger number of observations than an and auditory stimuli have been procedures in which
individual threshold, would be lower than the mean the next stimulus intensity level to be tested depends
6.9 Alternative Methods: Rated Difference, Adaptive Procedures, Scaling 139
upon detection or non-detection at the previous inter- the part of the respondent. Psychophysical researchers
val. In these procedures, the subject will track around have found ways to undo this sequential dependence
the threshold level, ascending in intensity when perfor- to counteract observer expectancies. One example is
mance is incorrect (or non-detection is the response) the double random staircase procedure (Cornsweet,
and descending in physical intensity when perfor- 1962) in which trials from two staircase sequences are
mance is correct (or detection is indicated). A common randomly intermixed. One staircase starts above the
example of this procedure is in some automated hear- threshold and descends, while the other starts below
ing tests, in which the person being tested pushes a the threshold and ascends. On any given trial, the
button as long as the signal is not audible. When the observer is unaware which of the two sequences the
button is depressed, the tone will increase in inten- stimulus is chosen from. As in the simple staircase
sity and when the button is released, the tone will procedure, the level chosen for a trial depends upon
decrease in intensity. This automated tracking pro- detection or discrimination in the previous trial, but
cedure leads to a series of up and down records, of that particular sequence. Further refinements of the
and an average of reversal points is usually taken to procedure involve the introduction of forced-choice
determine the threshold. Adaptive procedures may be (Jesteadt, 1980) to eliminate response bias factors
more efficient than a traditional procedure like the involved in simple yes/no detection.
method of limits. They focus on the critical range Another modification to the adaptive methods has
around threshold and do not waste time testing inten- been to adjust the ascending and descending rules so
sity levels very much higher or very much lower than that some number of correct or incorrect judgments is
the threshold (McBurney and Collings, 1977). Further required before changing intensity levels, rather than
information on these methods can be found in Harvey the one trial as in the simple staircase (Jesteadt, 1980).
(1986). An example is the “up down transformed response”
With discrete stimuli, rather than those that are rule or UDTR (Wetherill and Levitt, 1965). Wetherill
played constantly as in the example of the hearing and Levitt gave an example where two positive judg-
test, the procedure can be used for the taste and smell ments were required before moving down, and only
modalities as well. This procedure is sometimes called one negative judgment at a given level before mov-
a staircase method, since the record of ascending and ing up. An example is shown in Fig. 6.9. Rather than
descending trials can be connected on graph paper to estimating the 50% point on a traditional psycho-
produce a series of step intervals that visually resemble metric function, this more stringent requirement now
a staircase. An example is shown in Fig. 6.8. The pro- tends to converge on the 71% mark as an average of
cedure creates a dependence of each trial on previous the peaks and valleys in the series. A forced choice
trials that may lead to some expectations and bias on can be added to an adaptive procedure. Sometimes
SAMPLE STAIRCASE RECORD SAMPLE UDTR RECORD

64
CONCENTRATION STEP
CONCENTRATION STEP
32
16
4
R
2
R R R
1
R R
1 2 3 4 5 6 7 8 9 10
TRIAL TRIAL
Fig. 6.8 Staircase example. Fig. 6.9 Staircase example.

the initial part of the test sequence is discarded in examines the data across a highly variable group as is
analysis as it is unrepresentative of the final thresh- the case with specific anosmia or tasting PTC bitter-
old and derives from a time when the observer may ness. The correlation is negative since higher thresh-
still be warming up to the test procedure. Examples olds indicate lower sensitivity and thus lower rated
of the up-down procedure can be found in the liter- intensity.
ature on PTC/PROP tasting (e.g., Reed et al., 1995).
Recent advances in adaptive methods have shown that
thresholds may be estimated in very few trials using
these procedures, a potential advantage for taste and 6.10 Dilution to Threshold Measures
smell measurement (Harvey, 1986; Linschoten et al.,
1996). 6.10.1 Odor Units and
Gas-Chromatography
Olfactometry (GCO)
6.9.3 Scaling as an Alternative Measure
of Sensitivity In this section, several applied methods will be
described that make use of the threshold concept in
Threshold measurements are not the only way to trying to determine the sensory impact of various fla-
screen individuals for insensitivity to specific com- vors and odor materials. The first group of methods
pounds like PTC or to screen for specific anosmia. Do concerns the olfactory potency of volatile aroma com-
thresholds bear any relation to suprathreshold respond- pounds as they are found in foods or food components.
ing? While it has been widely held that there is no nec- The issue here becomes one of not just threshold deter-
essary relationship between threshold sensitivity and mination, but determination of both the threshold and
suprathreshold responding (Frijters, 1978; Pangborn, the actual concentration present in a food sample. The
1981), this assertion somewhat overstates the case. ratio of these concentrations (actual concentration to
Counter-examples of good correlations can be seen in threshold concentration) can help indicate whether or
tests involving compounds like PTC where there are not a given flavor substance is likely to contribute
insensitive groups. For example, there is a –0.8 correla- to the overall sensory impression in a food. These
tion between simple category taste intensity ratings for ratios are commonly called “odor units.” The sec-
PTC and the threshold, when the rated concentration ond much older method is similar in logic and was
is near the antimode or center between the modes of developed to determine the point at which the irrita-
a bimodal threshold frequency distribution (Lawless, tive or heat sensations from pepper compounds would
1980). Thus ratings of a carefully chosen level can be be first detectable when diluted to a given extent,
used for a rapid screening method for PTC taster status the Scoville procedure. Both of these techniques then
(e.g., Mela, 1989). use dilution-to-threshold as a measure of sensory
Similar results have been noted for smell. Berglund impact.
and Högman (1992) reported better reliability of When a complex natural product like a fruit extract
suprathreshold ratings than threshold determinations is analyzed for its chemical components, hundreds or
in screening for olfactory sensitivity. Stevens and even thousands of chemicals may be identified, many
O’Connell (1991) used category ratings of perceived of which have odor properties. The number of poten-
intensity as well as qualitative descriptors as a screen- tial flavor compounds identified in any product seems
ing tool before threshold testing for specific anosmia. only to be limited by the resolution and sensitivity
Threshold versus rating correlations were in the range of the current methods in analytical chemistry. These
of –0.6 for cineole, –0.3 for carvone, and –0.5 for methods are always improving leading to longer and
diacetyl (Lawless et al., 1994, 1995). The correla- longer lists of possible contributing flavor materials
tions were obtained after subtraction of ratings to a (Piggott, 1990). Flavor scientists need to find a way
blank stimulus, in order to correct for differences in to narrow the list or to separate those compounds
scale usage. Thus there is a moderate negative cor- which are most likely contributing to the overall
relation of sensitivity and rated intensity when one flavor from those compounds that are present in
6.10 Dilution to Threshold Measures 141
such low concentrations that they are probably not of testing. Unless the same techniques are used and
important. Obviously, a sensory-based method is the same medium was used as the carrier (rarely the
needed in conjunction with the analytical chemistry case) the values may not be necessarily comparable for
to provide a bioassay for possible sensory impact different compounds.
(Acree, 1993). A second approach is to actually measure the dilu-
Thresholds can be useful in addressing this kind tions necessary to reach threshold for each compound,
of problem. The reasoning goes that only those com- starting with the product itself. This necessitates the
pounds that are present in the product in concentrations use of a separatory procedure, so that each compound
above their threshold are likely to be contributors may be individually perceived. The combination of gas
to the flavor of the product. There are a number chromatography with odor port sniffing of a dilution
of potential flaws in this thinking discussed below, series is a popular technique (Acree, 1993). Various
but for now let us see how this can be put to use. catchy names have been applied to such techniques in
Given a concentration C present in a natural product, the flavor literature, including Aroma Extract Dilution
a dimensionless quantity can be derived by divid- Analysis (for examples, see Guth and Grosch, 1994;
ing that concentration by the threshold concentration Milo and Grosch, 1993; Schieberle and Grosch, 1988),
Ct , and the ratio C/Ct defines the number of odor CHARM analysis (Acree et al., 1984) or more gener-
units (or flavor units) for compounds assessed by ically, gas chromatography olfactometry or GCO. The
smell. According to this logic, only those compounds basis of these techniques is to have subjects respond
with odor units greater than one will contribute to when an odor is perceived when sniffing the exit port
the aroma of the product. This reasoning is extended during a GC run. In recent years, the effluent has
sometimes to include the idea that the greater the been embedded in a cooled, humidified air stream to
number of odor units, the greater the potential con- improve the comfort of the observer and to increase
tribution. However, it is now widely recognized that sensory resolution of the eluting compounds. Over sev-
the odor unit is a concentration multiple and not a eral dilutions, the response will eventually drop out,
measure of subjective magnitude. Only direct scaling and the index of smell potency is related to the recip-
methods can assess the actual magnitude of sensation rocal of the dilution factor. The sniffer’s responses
above threshold and the psychophysical relationship occur on a time base that can be cross-referenced to
between concentration and odor intensity (Frijters, a retention index and then the identity of the com-
1978). Furthermore, this idea ignores the possibility of pound can be determined by a combination of reten-
subthreshold additivity or synergy (Day et al., 1963). tion index, mass spectrometry, and aroma character.
A closely related group of chemical compounds might In practice, these techniques considerably shorten the
all be present below their individual thresholds, but list of potential aroma compounds contributing to fla-
together could stimulate common receptors so as to vor in a natural product (e.g., Cunningham et al.,
produce an above-threshold sensation. Such additivity 1986).
is not predicted by the odor unit approach and such The method has also been used as an assessment
a group of compounds could be missed in dilution technique for measuring the sensitivity of human pan-
analysis. elists, as opposed to being a tool to determine the sen-
Nonetheless, thresholds provide at least one iso- sory impact of flavor compounds (Marin et al., 1988).
intense reference point on the dose response curve, so In this approach, the gas chromatograph once again
they have some utility as a measure of potency used serves as an olfactometer. Known compounds can be
to compare different odor compounds. In analyzing a presented as a dilution series of mixtures. Variation in
food, one could look up literature values for all the individual thresholds can readily be assessed for a vari-
identified compounds in the product in one of the pub- ety of compounds since they can be combined in a GC
lished compendia of thresholds (e.g., ASTM, 1978; run as long as they have different retention times, i.e.,
van Gemert, 2003). If the concentration in the product do not co-elute on the column of choice. The potential
is determined, then the odor unit value can be calcu- use of GCO for screening panelists, for assessing odor
lated by simply dividing by threshold. However, it is responses in general, and for assessing specific anos-
important to remember that the literature values for mias has also been attempted (Friedrich and Acree,
thresholds depend upon the method and the medium 2000; Kittel and Acree, 2008).
6.10.2 Scoville Units modification defined a detection threshold at mean

scale value of 3.5. Almost all these methods specify
mandatory rest periods between samples due to the
Another example of a dilution method is the tra- long lasting nature of these sensations. The measure-
ditional Scoville procedure for scoring pepper heat ments are still difficult. One problem is that capsaicin,
in the spice trade. This procedure was named for the active heat principle in red pepper, is prone to
W. Scoville who worked in the pharmaceutical indus- desensitize observers within a session and also regu-
try in the early twentieth century. He was interested lar consumers of hot spices also become less sensitive,
in the topical application of spice compounds like leading to wide individual differences in sensitivity
pepper extracts as counterirritants, and he needed to among panelists (Green, 1989; Lawless et al., 1985).
establish units that could be used to measure their Alternative procedures have been developed based on
potency. His procedure consisted of finding the num- rating scales with fixed physical references (Gillette
ber of dilutions necessary for sensations to disappear et al., 1984) and these have been endorsed by ASTM
and then using this number of dilutions as an esti- (2008c) as standard test methods. These rating scale
mate of potency. In other words, potency was defined procedures show good correlations with instrumental
as a reciprocal threshold. A variation of this proce- measures of capsaicin content in pepper samples and
dure was adopted by the Essential Oil Association, can be cross-referenced to Scoville units for those who
British Standards Institution, International Standards prefer to do business in the traditional units (Gillette
Organization, American Spice Trade Association et al., 1984).
(ASTA), and adopted as an Indian standard method
(for a review, see Govindarajan, 1987).
The procedure defined units of pungency as the
highest dilution at which a definite “bite” would be per- 6.11 Conclusions
ceived and thus contains instructions consistent with a
recognition threshold. Scoville units were dilution fac- Threshold measurements find three common uses in
tors, now commonly given as mL/g. ASTA Method sensory analysis and flavor research. First, they can be
21 (ASTA, 1968) is widely used and contains some used to compare the sensitivities of different panelists.
modifications in an attempt to overcome problems with Second, they can be used as an index of the biological
the original Scoville procedure. In brief, the method potency of a flavor compound. Third, they can provide
proceeds as follows: Panelists are screened for acuity useful information regarding the maximum tolerable
relative to experienced persons. Dilution schedules are levels of an off-flavor or taint. A variety of differ-
provided which simplify calculations of the eventual ent techniques have been used to find thresholds or
potency. Solutions are tested in 5% sucrose and negli- have employed the threshold concept in practical fla-
gible amounts of alcohol. Five panelists participate and vor work. Examples of different threshold methods are
concentrations are given in ascending order around the given in Table 6.4. In spite of their practical appli-
estimated threshold. Threshold is defined as the con- cations, the usefulness of threshold measures is often
centration at which three out of five judges respond questioned in sensory evaluation. One criticism is that
positively. thresholds are only one point on an intensity function
This method is difficult in practice and a number of and thus they do not tell us anything about above-
additional variations have been tried to improve on the threshold responding. There are some good examples
accuracy and precision of the method (Govindarajan, in which thresholds do not predict or do not correlate
1987). Examples include the following: (1) substitu- very well with suprathreshold responses. For exam-
tion of other rules for 3/5, e.g., mean + SD of 20–30 ple, patients irradiated for cancer may lose their sense
judgments, (2) use of a triangle test, rather than simple of taste temporarily and thresholds return to normal
yes/no at each concentration, (3) requiring recogni- long before suprathreshold responsiveness is recov-
tion of pungency (Todd et al., 1977), (4) reduction ered (Bartoshuk, 1987). However, as we have seen
of sucrose concentration in the carrier solution to both in the case of PTC tasting and in specific anos-
3%, and (5) use of a rating scale from 1 (definitely mias, insensitive individuals (as determined by their
not detectable) to 6 (definitely detectable). This latter threshold) will also tend to be less responsive above
Appendix 143
Table 6.4 Threshold procedures

Method Citations/examples Response Threshold
Ascending forced choice ASTM E-679-79 3-AFC Geometric mean of individual threshold
Ascending forced choice Stevens et al. (1988) 2-AFC, 5 replicates Lowest correct set with confirmation at next
concentration
Semi-ascending paired Lundahl et al. (1986) Rated difference from t-test for significant difference from zero with
difference control all blank trials (blind control scores)
subtracted
Adaptive, up-down Reed et al. (1995) 2-AFC Average of responses ascending after one
transformed response rule incorrect, descending after two correct at
each level
Double Random Staircase Cornsweet (1962) Yes/No Average of reversal points
CHARM analysis Acree et al. (1986) Yes/No Nonresponse on descending concentration
runs (implied)
threshold. These correlations are strong when compar- as a potentially useful construct of our methods and
ing individuals of very different sensitivities, but the informational requirements.
threshold–suprathreshold parallel may not extend to all Finally, because of the problems mentioned above,
flavor compounds. A more complete understanding of sensory professionals need to keep the following prin-
the whole dynamic range of dose–response, as found ciples firmly in mind when working with threshold
in scaling studies, would be more informative. procedures: First, changes in method will change the
Other shortcomings need to be kept in mind by obtained values. Literature values cannot be trusted
sensory workers who would use threshold measure- to extend to a new product or a new medium or
ments for making product decisions. First, thresholds if changes are made in test procedure. Next, thresh-
are statistical constructs only. Signal detection theory old distributions do not always follow a normal bell
warns us that the signal and noise diverge in a con- curve. There are often high outliers and possibly cases
tinuous fashion and discontinuities in perception may of insensitivity due to inherited deficits like specific
be an idealized construction that is comfortable but anosmia (Amoore, 1971; Brown et al., 1978; Lawless
unrealistic. There is no sudden transition from non- et al., 1994). Threshold values for an individual are
detection to 100% detection. Any modern concept of prone to high variability and low test–retest reliabil-
threshold must take into account that a range of values, ity. An individual’s threshold measure on a given day
rather than a single point is involved in the speci- is not necessarily a stable characteristic of that person
fication. Thresholds depend upon the conditions of (Lawless et al., 1995; Stevens et al., 1988). Practice
measurement. For example, as the purity of the dilu- effects can be profound, and thresholds may stabilize
ent increases, the threshold for taste will go down. So a over a period of time (Engen, 1960; Rabin and Cain,
measure of the true absolute threshold for taste (if such 1986). However, group averaged thresholds are reli-
a thing existed) would require water of infinite purity. able (Brown et al., 1978; Punter, 1983) and provide a
The threshold exists not in this abstract sense, but only useful index of the biological activity of a stimulus.
Appendix: MTBE Threshold Data for Worked Example
Concentration (μg/L)
1 + o + + + + + + 4.6 0.663
2 o o o + + + + + 7.7 0.886
3 o + o o o o + + 42 1.623
4 o o o o o o + + 42 1.623
5 + o + o + + o + 77 1.886
6 o o + + + + + + 4.6 0.663
Concentration (μg/L)
7 o + + o + + + + 13 1.114
8 + + o + + + + + 7.7 0.886
9 o o + o + + + + 13 1.114
10 o o o o o o o + 77 1.886
11 + o + + + + + + 4.6 0.663
12 o o o o + o + o 132 2.121
13 + + + + + + + + 1.4 0.146
14 + + + + + + + + 1.4 0.146
15 o + + o + + + + 13 1.114
16 o o + o o o + o 132 2.121
17 + o + + + + + + 4.6 0.663
18 o o + + + + + + 4.6 0.663
19 + + + + + + + + 1.4 0.146
20 + o + + o + o o 132 2.121
21 + + + + + + + + 1.4 0.146
22 + + + + + + + + 1.4 0.146
23 + + o o + + + + 13 1.114
24 + + + + + + + + 1.4 0.146
25 o + + + o + + + 23 1.362
26 o + + + o + + + 23 1.362
27 + o o o + o o + 77 1.886
28 o o + + + + + o 132 2.121
29 o + o o + + + + 13 1.114
30 o + + o + + + + 13 1.114
31 + o o + o o + o 132 2.121
32 + + + + o o o + 77 1.886
33 + + + o + o + + 42 1.623
34 o o o o o o + o 132 2.121
35 o o o + o + + o 132 2.121
36 o o o + o + + + 23 1.362
37 + + + + + + + + 1.4 0.146
38 + o + + + + + + 1.4 0.146
39 + + o o + + + + 13 1.114
40 o o + o + + + + 13 1.114
41 + + + + + + + + 1.4 0.146
42 o + o + + + + + 7.7 0.886
43 o o o o o o + + 42 1.623
44 o + + o + o + + 42 1.623
45 o + + o + + + + 13 1.114
46 + + + + + + o + 77 1.886
47 o + o o o + + o 132 2.121
48 o + o + o + + + 23 1.362
49 o o o + o + + + 23 1.362
50 o o + + + + + + 4.6 0.663
51 o o o + + + + + 7.7 0.886
52 + + + + + + + + 1.4 0.146
53 + + + + + + + + 1.4 0.146
54 + o o o o + + + 23 1.362
55 o o + + + + + + 4.6 0.663
56 o o o o o + + + 23 1.362
57 o o + + o + o + 77 1.886
Prop. Corr. 0.44 0.49 0.61 0.58 0.65 0.77 0.89 0.86 Mean (log(BET)) 1.154
10ˆ1.154= 14.24
References 145
References Boring, E. G. 1942. Sensation and Perception in the

History of Experimental Psychology. Appleton-Century-
Crofts, New York, pp. 41–42.
Acree, T. E. 1993. Bioassays for flavor. In: T. E. Acree Brown, D. G. W., Clapperton, J. F., Meilgaard, M. C. and Moll,
and R. Teranishi (eds.), Flavor Science, Sensible Principles M. 1978. Flavor thresholds of added substances. American
and Techniques. American Chemical Society Books, Society of Brewing Chemists Journal, 36, 73–80.
Washington, pp. 1–20. Bufe, B., Breslin, P. A. S., Kuhn, C., Reed, D., Tharp, C.
Acree, T. E., Barnard, J. and Cunningham, D. G. 1984. A pro- D., Slack, J. P., Kim, U.-K., Drayna, D. and Meyerhof,
cedure for the sensory analysis for gas chromatographic W. 2005. The molecular basis of individual differences in
effluents. Food Chemistry, 14, 273–286. phenylthiocarbamide and propylthiouracil bitterness percep-
American Spice Trade Association 1968. Pungency of capsicum tion. Current Biology, 15, 322–327.
spices and oleoresins. American Spice Trade Association Cain, W. S. 1976. Olfaction and the common chemical
Official Analytical Methods, 21.0, 43–47. sense: Some psychophysical contrasts. Sensory Processes, 1,
Amoore, J. E. 1971. Olfactory genetics and anosmia. In: L. M. 57–67.
Beidler (ed.), Handbook of Sensory Physiology. Springer, Cain, W. S. and Murphy, C. L. 1980. Interaction between
Berlin, pp. 245–256. chemoreceptive modalities of odor and irritation. Nature,
Amoore, J. E. 1979. Directions for preparing aqueous solu- 284, 255–257.
tions of primary odorants to diagnose eight types of specific Cornsweet, T. M. 1962. The staircase method in psychophysics.
anosmia. Chemical Senses and Flavor, 4, 153–161. American Journal of Psychology, 75, 485–491.
Amoore, J. E.,Venstrom, D. and Davis, A. R. 1968. Cunningham, D. G., Acree, T. E., Barnard, J., Butts, R. M. and
Measurement of specific anosmia. Perceptual and Motor Braell, P. A. 1986. CHARM analysis of apple volatiles. Food
Skills, 26, 143–164. Chemistry, 19, 137–147.
Antinone, M. A., Lawless, H. T., Ledford, R. A. and Johnston, Dale, M. S., Moylan, M. S., Koch, B. and Davis, M. K. 1997.
M. 1994. The importance of diacetyl as a flavor compo- Taste and odor threshold determinations using the flavor
nent in full fat cottage cheese. Journal of Food Science, 59, profile method. Proceedings of the American Water Works
38–42. Association Water-Quality Technology Conference, Denver,
ASTM. 1978. Compilation of Odor and Taste Threshold CO.
Values Data. American Society for Testing and Materials, Dalton, P., Doolittle, N. and Breslin, P. A. S. 2002. Gender-
Philadelphia. specific induction of enhanced sensitivity to odors. Nature
ASTM. 2008a. Standard practice for determining odor and taste Neuroscience, 5, 199–200.
thresholds by a forced-choice ascending concentration series Day, E. A., Lillard, D. A. and Montgomery, M. W. 1963.
method of limits, E-679-04. Annual Book of Standards, Autooxidation of milk lipids.III. Effect on flavor of the
Vol. 15.08. ASTM International, Conshocken, PA, additive interactions of carbonyl compounds at subthreshold
pp. 36–42. concentrations. Journal of Dairy Science, 46, 291–294.
ASTM. 2008b. Standard practice for defining and calculat- Diamond, J., Dalton, P., Doolittle, N. and Breslin, P. A. S. 2005.
ing individual and group sensory thresholds from forced- Gender-specific olfactory sensitization: Hormonal and cog-
choice data sets of intermediate size, E-1432-04. ASTM nitive influences. Chemical Senses, 30(suppl 1), i224–i225.
International Book of Standards, Vol. 15.08. ASTM Dravnieks, A. and Prokop, W. H. 1975. Source emission odor
International, Conshocken, PA, pp.82–89. measurement by a dynamic forced-choice triangle olfac-
ASTM. 2008c. Standard test method for sensory evaluation of tometer. Journal of the Air Pollution Control Association, 25,
red pepper heat, E-1083-00. ASTM International Book of 28–35.
Standards, Vol. 15.08. ASTM International, Conshocken, Engen, T. E. 1960. Effects of practice and instruction on olfac-
PA, pp. 49–53. tory thresholds. Perceptual and Motor Skills, 10, 195–198.
Ayya, N. and Lawless, H. T. 1992. Qualitative and quantita- Finney, D. J. 1971. Probit Analysis, Third Edition. Cambridge
tive evaluation of high-intensity sweeteners and sweetener University, London.
mixtures. Chemical Senses, 17, 245–259. Friedrich, J. E. and Acree, T. E. 2000. Design of a standard set
Bartoshuk, L. M. 1987. Psychophysics of taste. American of odorants to test individuals for specific anosmia. Frontiers
Journal of Clinical Nutrition, 31, 1068–1077. in Flavour Science [Proc. 9th Weurman Flavour Res. Symp.],
Bartoshuk, L. M., Murphy, C. and Cleveland, C. T. 1978. Sweet 230–234.
taste of dilute NaCl: Psychophysical evidence for a sweet Frijters, J. E. R. 1978. A critical analysis of the odour unit num-
stimulus. Physiology and Behavior, 21, 609–613. ber and its use. Chemical Senses and Flavour, 3, 227–233.
Berglund, B. and Högman, L. 1992. Reliability and validity of Gillette, M., Appel, C. E. and Lego, M. C. 1984. A new method
odor measurements near the detection threshold. Chemical for sensory evaluation of red pepper heat. Journal of Food
Senses, 17, 823–824. Science, 49, 1028–1033.
Blakeslee, A. F. 1932. Genetics of sensory thresholds: Taste for Green, B. G. 1989. Capsaicin sensitization and desensitiza-
phenylthiocarbamide. Proceedings of the National Academy tion on the tongue produced by brief exposures to a low
of Sciences, 18, 120–130. concentration. Neuroscience Letters, 107, 173–178.
Brennand, C. P., Ha, J. K. and Lindsay, R. C. 1989. Aroma prop- Govindarajan, V. S. 1987. Capsicum – production, technology,
erties and thresholds of some branched chain and other minor chemistry and quality. Part III. Chemistry of the color, aroma
volatile fatty acids occurring in milkfat and meat lipids. and pungency stimuli. CRC Critical Reviews in Food Science
Journal of Sensory Studies, 4, 105–120. and Nutrition, 24, 245–311.
Guth, H. and Grosch, W. 1994. Identification of the character material. Journal of Agricultural and Food Chemistry, 41,
impact odorants of stewed beef juice by instrumental anal- 2076–2081.
ysis and sensory studies. Journal of Agricultural and Food Mojet, J., Christ-Hazelhof, E. and Heidema, J. 2001. Taste
Chemistry, 42, 2862–2866. perception with age: Generic or specific losses in thresh-
Harris, H. and Kalmus, H. 1949. The measurement of taste sen- old sensitivity to the five basic tastes. Chemical Senses 26,
sitivity to phenylthiourea (P.T.C.). Annals of Eugenics, 15, 854–860.
24–31. Morrison, D. G. 1978. A probability model for forced binary
Harvey, L. O. 1986. Efficient estimation of sensory thresholds. choices. American Statistician, 32, 23–25.
Behavior Research Methods, Instruments and Computers, Murphy, C., Quiñonez, C. and Nordin, S. 1995. Reliability and
18, 623–632. validity of electrogustometry and its application to young and
James, W. 1913. Psychology. Henry Holt and Co., New York. elderly persons. Chemical Senses, 20(5), 499–515.
Jesteadt, W. 1980. An adaptive procedure for subjective judg- O’Mahony, M. and Ishii, R. 1986. Umami taste concept:
ments. Perception and Psychophysics, 28(1), 85–88. Implications for the dogma of four basic tastes. In:
Kittel, K. M and Acree, T. E. 2008. Investigation of olfactory Y. Kawamura and M. R. Kare (eds.), Umami: A Basic Taste.
deficits using gas-chromatography olfactometry. Manuscript Marcel Dekker, New York, pp. 75–93..
submitted, available from the authors. O’Mahony, M. and Odbert, N. 1985. A comparison of sensory
Lawless, H. T. 1980. A comparison of different methods used to difference testing procedures: Sequential sensitivity analysis
assess sensitivity to the taste of phenylthiocarbamide PTC. and aspects of taste adaptation. Journal of Food Science, 50,
Chemical Senses, 5, 247–256. 1055–1060.
Lawless, H. T. 2010. A simple alternative analysis for thresh- Pangborn, R. M. 1981. A critical review of threshold, intensity
old data determined by ascending forced-choice method of and descriptive analyses in flavor research. In: Flavor ’81.
limits. Journal of Sensory Studies, 25, 332–346. Walter de Gruyter, Berlin, pp. 3–32.
Lawless, H. T., Antinone, M. J., Ledford, R. A. and Johnston, Piggott, J. R. 1990. Relating sensory and chemical data to
M. 1994. Olfactory responsiveness to diacetyl. Journal of understand flavor. Journal of Sensory Studies, 4, 261–272.
Sensory Studies, 9(1), 47–56. Prescott, J., Norris, L., Kunst, M. and Kim, S. 2005. Estimating a
Lawless, H. T., Rozin, P. and Shenker, J. 1985. Effects of “consumer rejection threshold” for cork taint in white wine.
oral capsaicin on gustatory, olfactory and irritant sensations Food Quality and Preference, 18, 345–349.
and flavor identification in humans who regularly or rarely Punter, P. H. 1983. Measurement of human olfactory thresh-
consumer chili pepper. Chemical Senses, 10, 579–589. olds for several groups of structurally related compounds.
Lawless, H. T., Thomas, C. J. C. and Johnston, M. 1995. Chemical Senses, 7, 215–235.
Variation in odor thresholds for l-carvone and cineole and Rabin, M. D. and Cain, W. S. 1986. Determinants of measured
correlations with suprathreshold intensity ratings. Chemical olfactory sensitivity. Perception and Psychophysics, 39(4),
Senses, 20, 9–17. 281–286.
Linschoten, M. R., Harvey, L. O., Eller, P. A. and Jafek, B. W. Reed, D. R., Bartoshuk, L. M., Duffy, V., Marino, S. and Price,
1996. Rapid and accurate measurement of taste and smell R. A. 1995. Propylthiouracil tasting: Determination of under-
thresholds using an adaptive maximum-likelihood staircase lying threshold distributions using maximum likelihood.
procedure. Chemical Senses, 21, 633–634. Chemical Senses, 20, 529–533.
Lundahl, D. S., Lukes, B. K., McDaniel, M. R. and Henderson, Saliba, A. J., Bullock, J. and Hardie, W. J. 2009. Consumer rejec-
L. A. 1986. A semi-ascending paired difference method for tion threshold for 1,8 cineole (eucalyptol) in Australian red
determining the threshold of added substances to background wine. Food Qualithy and Preference, 20, 500–504.
media. Journal of Sensory Studies, 1, 291–306. Schieberle, P. and Grosch, W. 1988. Identification of potent fla-
Masuoka, S., Hatjopolous, D. and O’Mahony, M. 1995. Beer vor compounds formed in an aqueous lemon oil/citric acid
bitterness detection: Testing Thurstonian and sequential sen- emulsion. Journal of Agricultural and Food Chemistry, 36,
sitivity analysis models for triad and tetrad methods. Journal 797–800.
of Sensory Studies, 10, 295–306. Stevens, D. A. and O’Connell, R. J. 1991. Individual differences
Marin, A. B., Acree, T. E. and Barnard, J. 1988. Variation in threshold and quality reports of subjects to various odors.
in odor detection thresholds determined by charm analysis. Chemical Senses, 16, 57–67.
Chemical Senses, 13, 435–444. Stevens, D. A. and O’Connell, R. J. 1995. Enhanced sensitiv-
Marin, A. B., Barnard, J., Darlington, R. B. and Acree, T. E. ity to adrostenone following regular exposure to pemenone.
1991. Sensory thresholds: Estimation from dose-response Chemical Senses, 20, 413–419.
curves. Journal of Sensory Studies, 6(4), 205–225. Stevens, J. C., Cain, W. S. and Burke, R. J. 1988. Variability of
McBurney, D. H. and Collings, V. B. 1977. Introduction to olfactory thresholds. Chemical Senses, 13, 643–653.
Sensation and Perception. Prentice-Hall, Englewood Cliffs, Stocking, A. J., Suffet, I. H., McGuire, M. J. and Kavanaugh,
NJ. M. C. 2001. Implications of an MTBE odor study for set-
Meilgaard, M., Civille, G. V. and Carr, B. T. 1991. Sensory ting drinking water standards. Journal AWWA, March 2001,
Evaluation Techniques, Second Edition. CRC Press, Boca 95–105.
Raton, FL. Todd, P. H., Bensinger, M. G. and Biftu, T. 1977. Determination
Mela, D. 1989. Bitter taste intensity: The effect of tastant and of pungency due to capsicum by gas-liquid chromatography.
thiourea taster status. Chemical Senses, 14, 131–135. Journal of Food Science, 42, 660–665.
Milo, C. and Grosch, W. 1993. Changes in the odorants of boiled Tuorila, H., Kurkela, R., Suihko, M. and Suhonen. 1981.
trout (Salmo fario) as affected by the storage of the raw The influence of tests and panelists on odour detection
References 147
thresholds. Lebensmittel Wissenschaft und Technologie, 15, Walker, J. C., Hall, S. B., Walker, D. B., Kendall-Reed, M. S.,
97–101. Hood, A. F. and Nio, X.-F. 2003. Human odor detectability:
USEPA (U.S. Environmental Protection Agency). 2001. New methodology used to determine threshold and variation.
Statistical analysis of MTBE odor detection thresh- Chemical Senses, 28, 817–826.
olds in drinking water. National Service Center for Wetherill, G. B. and Levitt, H. 1965. Sequential estimation
Environmental Publications (NSCEP) # 815R01024, avail- of points on a psychometric function. British Journal of
able from http://nepis.epa.gov. Mathematical and Statistical Psychology, 18(1), 1–10.
Van Gemert, L. 2003. Odour Thresholds. Oliemans, Punter & Wysocki, C. J., Dorries, K. M. and Beauchamp, G. K. 1989.
Partners, Utrecht, The Netherlands. Ability to perceive androstenone can be acquired by osten-
Viswanathan, S., Mathur, G. P., Gnyp, A. W. and St. Peirre, sibly anosmic people. Proceedings of the National Academy
C. C. 1983. Application of probability models to threshold of Science, USA, 86, 7976–7989.
determination. Atmospheric Environment, 17, 139–143.
Chapter 7
Scaling
Abstract Scaling describes the application of numbers, or judgments that are

converted to numerical values, to describe the perceived intensity of a sensory expe-
rience or the degree of liking or disliking for some experience or product. Scaling
forms the basis for the sensory method of descriptive analysis. A variety of methods
have been used for this purpose and with some caution, all work well in differentiating
products. This chapter discusses theoretical issues as well as practical considerations
in scaling.
The vital importance of knowing the properties and limitations of a measuring instrument can
hardly be denied by most natural scientists. However, the use of many different scales for sensory
measurement is common within food science; but very few of these have ever been validated. . . .
—(Land and Shepard, 1984, pp. 144–145)
Contents 7.7 Issues . . . . . . . . . . . . . . . . . . . . 168

7.7.1 “Do People Make Relative Judgments”
Should They See Their Previous
7.1 Introduction . . . . . . . . . . . . . . . . . 149 Ratings? . . . . . . . . . . . . . . . 168
7.2 Some Theory . . . . . . . . . . . . . . . . 151 7.7.2 Should Category Rating Scales Be Assigned
7.3 Common Methods of Scaling . . . . . . . . . 152 Integer Numbers in Data Tabulation? Are
7.3.1 Category Scales . . . . . . . . . . . . 152 They Interval Scales? . . . . . . . . . 169
7.3.2 Line Scaling . . . . . . . . . . . . . 155 7.7.3 Is Magnitude Estimation a Ratio
7.3.3 Magnitude Estimation . . . . . . . . . 156 Scale or Simply a Scale with Ratio
7.4 Recommended Practice and Practical Instructions? . . . . . . . . . . . . . 169
Guidelines . . . . . . . . . . . . . . . . . . 158 7.7.4 What is a “Valid” Scale? . . . . . . . . 169
7.4.1 Rule 1: Provide Sufficient 7.8 Conclusions . . . . . . . . . . . . . . . . . 170
Alternatives . . . . . . . . . . . . . . 159 Appendix 1: Derivation of Thurstonian-Scale Values
7.4.2 Rule 2: The Attribute Must for the 9-Point Scale . . . . . . . . . . . . . 171
Be Understood . . . . . . . . . . . . 159 Appendix 2: Construction of Labeled Magnitude
7.4.3 Rule 3: The Anchor Words Should Make Scales . . . . . . . . . . . . . . . . . . . . 172
Sense . . . . . . . . . . . . . . . . . 159 References . . . . . . . . . . . . . . . . . . . . 174
7.4.4 To Calibrate or Not to Calibrate . . . . . 159
7.4.5 A Warning: Grading and Scoring are Not
Scaling . . . . . . . . . . . . . . . . 160
7.5 Variations—Other Scaling Techniques . . . . 160 7.1 Introduction
7.5.1 Cross-Modal Matches and Variations on
Magnitude Estimation . . . . . . . . . 160
7.5.2 Category–Ratio (Labeled Magnitude) People make changes in their behavior all the time
Scales . . . . . . . . . . . . . . . . 162 based on sensory experience and very often this
7.5.3 Adjustable Rating Techniques: Relative involves a judgment of how strong or weak something
Scaling . . . . . . . . . . . . . . . . 164 feels. We add more sugar to our coffee if it is not sweet
7.5.4 Ranking . . . . . . . . . . . . . . . 165
7.5.5 Indirect Scales . . . . . . . . . . . . 166 enough. We adjust the thermostat in our home if it is
7.6 Comparing Methods: What is a Good too cold or too hot. If a closet is too dark to find your
Scale? . . . . . . . . . . . . . . . . . . . . 167 shoes you turn the light on. We apply more force to
150 7 Scaling
chew a tough piece of meat if it will not disintegrate dose–response curve or psychophysical function. Such
to allow swallowing. These behavioral decisions seem a study would perhaps be done on a student sample or a
automatic and do not require a numerical response. But consumer population. A second kind of scaling is done
the same kinds of experiences can be evaluated with a when trained panelists are used as if they were measur-
response that indicates the strength of the sensation. ing instruments, as in descriptive analysis. In this case
What was subjective and private becomes public data. they may be trained to insure a uniform understanding
The data are quantitative. This is the basis of scaling. of the attribute involved (e.g., astringency) and often
The methods of scaling involve the application of they are calibrated with reference standards to illustrate
numbers to quantify sensory experiences. It is through what is a little and what is a lot of this attribute. In this
this process of numerification that sensory evalua- case the focus is in on the products being tested and not
tion becomes a quantitative science subject to statis- the more basic process of specifying a psychophysical
tical analysis, modeling, prediction, and hard theory. function that has general application.
However, as noted in the quote above, in the practi- Note that these are “cheap data” (a term used by
cal application of sensory test methods, the nature of my advisor in graduate school, but perhaps “cost-
this process of assigning numbers to experiences is effective” sounds a little less negative). One stimulus
rarely questioned and deserves scrutiny. Clearly num- gives at least one data point. This is in contrast to indi-
bers can be assigned to sensations by a panelist in a rect methods like forced choice tests. Many responses
variety of ways, some by mere categorization, or by on a triangle test are needed to give one data point, i.e.,
ranking or in ways that attempt to reflect the intensity the percent correct. Fechner and others referred to scal-
of sensory experience. This chapter will illustrate these ing as “the method of single stimuli” and considered it
techniques and discuss the arguments that have been less reliable than the choice methods that were used to
raised to substantiate the use of different quantification generate difference thresholds. However, direct scaling
procedures. came into its own with the advent of magnitude estima-
Scaling involves sensing a product or stimulus and tion, an open-ended numerical response method. One
then generating a response that reflects how the person or another type of scaling forms the basis for virtually
perceives the intensity or strength of one or more of all descriptive analysis techniques. In descriptive anal-
the sensations generated by that product. This process ysis, panelists generate scaled responses for various
is based on a psychophysical model (see Chapter 2). sensory attributes to reflect their subjective intensity.
The psychophysical model states that as the physical There are two processes involved in scaling as
strength of the stimulus increases (e.g., the energy of a shown in Fig. 7.1. The first is the psychophysical chain
light or sound or the concentration of a chemical stim- of events in which some energy or matter impinges
ulus) the sensation will increase in some orderly way. upon receptors and the receptors send signals to the
Furthermore, panelists are capable of generating dif-
ferent responses to indicate these changes in what they
experience. Thus a systematic relationship can be mod-
eled of how physical changes in the real world result in
changing sensations.
Scaling is a tool used for showing differences and
degrees of difference among products. These differ-
ences are usually above the threshold level or just-
noticeable difference. If the products are very similar
and there is a question of whether there is any differ-
ence at all, the discrimination testing methods are more
suitable (Chambers and Wolf, 1996). Scaling is usually Fig. 7.1 The two processes involved in scaling. The first, phys-
done in one of the two scenarios. In the first, untrained iological process is the psychophysical translation of energy in
observers are asked to give responses to reflect changes the outside world into sensation, i.e., conscious experience. The
in intensity and it is presumed that (1) they understand second is the translation of that experience into some response.
The psychophysical process can be modified by physiological
the attribute they are asked to scale, e.g., sweetness and processes such as adaptation and masking. The judgment func-
(2) there is no need to train or calibrate them to use tion can be modified by cognitive processes such as contextual
the scale. This is the kind of scaling done to study a effects, number usage, and other response biases.
7.2 Some Theory 151
brain. These signals are interpreted in conscious per- of the sexes. They merely serve as convenient labels.
ception as a sensation with some intensity or strength. The meals at which a food might be eaten could be
The translation can be modified (i.e., the experience coded with numbers as categories—one for breakfast,
will change) by processes like adaptation or mask- two for lunch, three for supper, and four for snacks.
ing from another stimulus. The second process is the The assignment of a number for analysis is merely a
translation of that experience into an overt response label, a category or pigeonhole. The appropriate analy-
(the data). This judgment function is influenced by the sis of such data is to make frequency counts. The mode,
nature of the scaling task that the panelist is asked to the most frequent response, is used as a summary
perform. Factors such as contextual effects, the choice statistic for nominal data. Different frequencies of
of comparison products, and response biases of that response for different products or circumstances can be
particular person can modify the process. The bet- compared by chi-square analysis or other nonparamet-
ter the data reflect the experience, the more valid is ric statistical methods (Siegel, 1956; see Appendix B).
the scaling method. The sensory professional must be The only valid comparisons between individual items
careful to avoid response methods that introduce biases with this scale is to say whether they belong to the
or non-sensory influences on the response output. For same category or to different ones (an equal versus not
example, I might be asked to generate some open- equal decision).
ended numerical response to reflect my perception, but In ordinal scaling, numbers are assigned to recog-
I might have some “favorite” numbers I find easy to use nize the rank order of products with regard to some
(e.g., integers or multiples of 2, 5, and 10) so this num- sensory property, attitude, or opinion (such as pref-
ber bias interferes to some degree with the translation erence). In this case increasing numbers assigned to
of my experience into a truly accurate response. the products represent increasing amounts or intensi-
This chapter will focus on various methods that have ties of sensory experience. So a number of wines might
been used in sensory evaluation and in psychophysics be rank ordered for perceived sweetness or a num-
for scaling. Theory, principles, and issues will be dis- ber of fragrances rank ordered from most preferred to
cussed to provide depth of understanding. For the least preferred. In this case the numbers do not tell us
student who wishes to learn just the basic practices, anything about the relative differences among the prod-
Sections 7.3 and 7.4 are the most practically rele- ucts. We cannot draw conclusions about the degree of
vant sections. Section 7.5 illustrates some alternative difference perceived nor the ratio or magnitude of dif-
methods that have appeared in the sensory evaluation ference. In an analogy to the order of runners finishing
literature, but have not at this time enjoyed widespread in a race, we know who placed first, second, third, etc.
adoption in industrial practice. The sensory scientist But this order does indicate neither the finishing dis-
should be aware of these additional methods for the tances between contestants nor the differences in their
potential advantages they may provide. elapsed times. In general, analyses of ranked data can
report medians as the summary statistic for central ten-
dency or other percentiles to give added information.
As with nominal data, nonparametric statistical analy-
7.2 Some Theory ses (see Appendix B) are appropriate when ranking is
done (Siegel, 1956).
Measurement theory tells us that numbers can be The next level of scaling occurs when the subjec-
assigned to items in different ways. This distinction tive spacing of responses is equal, so the numbers
was popularized by S. S. Stevens, the major propo- represent equal degrees of difference. This is called
nent of magnitude estimation (1951). At least four interval-level measurement. Examples in the physi-
ways of assigning numbers to events exist in common cal sciences would be the centigrade and Fahrenheit
usage. These are referred to as nominal scaling, ordinal scales of temperature. These scales have arbitrary zero
scaling, interval scaling, and ratio scaling. points but equal divisions between values. The scales
In nominal scaling, numbers are assigned to events are inter-convertible through a linear transformation,
merely as labels. Thus gender may be coded as a for example, ◦ C = 5/9 (◦ F–32). Few scales used in sen-
“dummy variable” in statistical analysis by assigning sory science have been subjected to tests that would
a zero to males and a one to females; no assumption is help establish whether they achieved an interval level
made that these numbers reflect any ordered property of measurement and yet this level is often assumed.
152 7 Scaling
One scale with approximately equal subjective spacing instructed to assign numbers in relative proportions
is the 9-point category scale used for like–dislike judg- that reflect the strength of their sensations (Stevens,
ments, the 9-point hedonic scale (Peryam and Girardot, 1956). However, ratio instructions are easy to give,
1952). The phrases are shown below: but whether the scale has ratio properties in reflecting
a person’s actual subjective experiences is difficult to
Like extremely
determine, if not impossible.
Like very much
Because of these different measurement types with
Like moderately
different properties, the sensory professional must be
Like slightly
careful about two things. First, statements about dif-
Neither like nor dislike
ferences or ratios in comparing the scores for two
Dislike slightly
products should not be made when the measurement
Dislike moderately
level is only nominal or ordinal. Second, it is risky
Dislike very much
to use parametric statistics for measurements that
Dislike extremely
reflect only frequency counts or rankings (Gaito, 1980;
These response choices are commonly entered as Townsend and Ashby, 1980). Nonparametric methods
data by assignment of the numbers one through nine. are available for statistical analyses of such data.
Extensive research was conducted to find the apparent
spacing of various adjective labels for the scale points
(Jones and Thurstone, 1955; Jones et al., 1955). The
7.3 Common Methods of Scaling
technique for deciding on the subjective spacing was
the use of one kind of Thurstonian scaling method.
The original data do not fully support the notion of Several different scaling methods have been used to
equality of spacing, as discussed below. However, the apply numbers to sensory experience. Some, like mag-
scale worked well in practice, so this tradition of inte- nitude estimation, are adapted from psychophysical
ger assignment has persisted. The method is illustrated research, and others, like category scaling have become
in Appendix 1 at the end of this chapter. Thurstonian popular through practical application and dissemi-
theory is discussed further in Chapter 5. nation in a wide variety of situations. This section
The advantage of interval-level measurement is that illustrates the common techniques of category scales,
the data allow added interpretation. In a horse-racing line marking, and magnitude estimation. The next
example, we know the order the horses finished in section discusses the less frequently used techniques
and about how many “lengths” separated each horse. of hybrid category–ratio scales, indirect scales, and
A second advantage is that more powerful statistical ranking as alternatives. Two other methods are illus-
methods may be brought to bear the parametric meth- trated. Intensity matching across sensory modalities,
ods. Computation of means, t-tests, linear regression, called cross-modality matching, was an important psy-
and analysis of variance are appropriate analyses. chophysical technique and a precedent to some of the
Another even more desirable level of measurement category–ratio scales. Finally, adjustable rating tech-
is ratio measurement. In this case the zero level is niques in which panelists make relative placements and
fixed and not arbitrary and numbers will reflect relative are able to alter their ratings are also discussed.
proportions. This is the level of measurement com-
monly achieved in the physical sciences for quantities
like mass, length, and temperature (on the absolute 7.3.1 Category Scales
or Kelvin scale). Statements can be made that this
item has twice as much length or mass than that Perhaps the oldest method of scaling involves the
item. Establishing whether a sensory scaling method choice of discrete response alternatives to signify
actually assigns numbers to represent the relative pro- increasing sensation intensity or degrees of liking
portions of different sensation intensities is a dif- and/or preference. The alternatives may be presented
ficult matter. It has been widely assumed that the in a horizontal or vertical line and may offer choices of
method of magnitude estimation is a priori a ratio scal- integer numbers, simple check boxes, or word phrases.
ing procedure. In magnitude estimation, subjects are Examples of simple category scales are shown in
7.3 Common Methods of Scaling 153
A) INTENSITY
1 2 3 4 5 6 7 8 9
Weak Strong
B) Oxidized ___ not noticeable C) Difference from reference

___ trace, not sure
___ faint ___ No difference
___ slight ___ Very slight difference
___ mild ___ Slight difference
___ moderate ___ Moderate difference
___ definite ___ Large difference
___ strong ___ Very large difference
___ very strong
D) Sweetness
Not sweet Extremely

at all sweet
E) Sweetness
Weaker R Stronger
F) Hedonic scale for children
Maybe Good
Super Good Really Good Good or Bad Really Bad Super Bad
Maybe Bad
Fig. 7.2 Examples of category scales. (a) a simple integer scale check-box scale for perceived intensity; (e) a simple check-box
for sensation strength (after Lawless and Malone, 1986b); (b) a scale for difference in intensity from some reference sample,
verbal scale for degree of oxidized flavor (after Mecredy et al., marked R (after Stoer and Lawless, 1993); (f) a facial scale
1974; (c) a verbal scale for degree of difference from some ref- suitable for use with children, after Chen et al. (1996).
erence or control sample (after Aust et al., 1985), (d) a simple
Fig. 7.2. The job of the consumer or panelist is to of intensity levels will often improve and more scale
choose the alternative that best represents their reac- points may be added to allow the panel to make
tion or sensation. In a category scale the number of finer distinctions. A key idea is to present an eas-
alternative responses is limited. Seven to 15 categories ily understandable word like “sweetness” and ask the
are commonly used for intensity scaling depending participant to evaluate the perceived intensity of that
upon the application and the number of gradations that attribute. A second important factor concerns the ver-
the panelists are able to distinguish in the products. bal labels that appear along the alternatives. At the very
As panel training progresses, perceptual discrimination least, the low and high ends of the scale must be labeled
154 7 Scaling
with words that make sense, e.g., “not sweet at all” to referring to all products (a so-called universal scale)?
“extremely sweet.” Or is the scale anchored in the panelists’ minds only
A wide variety of these scales have been used. to products in this category? In that case, extremely
A common version is to allow integer numerical sweet for a salt cracker refers to something different
responses of approximately nine points (e.g., Lawless than extremely sweet for a confectionary product or
and Malone, 1986a, b). Further gradations may be ice cream. Or is the high end of the scale the most
allowed. For example, Winakor et al. (1980) allowed extreme attribute for this product? That would yield a
options from 1 to 99 in rating attributes of fabric hand- product-specific scale in which comparisons between
feel. In the Spectrum method (Meilgaard et al., 2006) a different attributes could be made, e.g., this cracker is
15-point category scale is used, but allows intermediate much sweeter than it is salty, but not comparisons to
points in tenths, rendering it (at least in theory) a 150- another type of product. These are important concerns
point scale. In hedonic or affective testing, a bipolar for a descriptive panel leader.
scale is common, with a zero or neutral point of opin- However, some experimenters nowadays avoid any
ion at the center (Peryam and Girardot, 1952). These extra instructions, allowing the subject or panelist to
are often shorter than the intensity scales. For exam- distribute their ratings along the scale as they see
ple, in the “smiling face” scale used with children, only fit. In fact, most people will have a tendency to dis-
three options may be used for very young respondents tribute their judgments along most of the scale range,
(Birch et al., 1980, 1982), although with older children although some avoid the end categories, reserving
as many as nine points may be used (Chen et al., 1996; them in case extreme examples show up. However,
Kroll, 1990). Lately there has been a move away from panelists do not like to overuse one part of the scale
using labels or integers, in case these may be biasing and will tend to move these judgments into adjoining
to subjects. People seem to have favorite numbers or response categories (Parducci, 1965). These tendencies
tendencies to use some numbers more than others (e.g., are discussed in Chapter 9.
Giovanni and Pangborn, 1983). A solution to this prob- In practice, simple category scales are about as
lem is to use an unlabeled check-box scale as shown in sensitive to product differences as other scaling tech-
Fig. 7.2. niques, including line marking and magnitude estima-
In early applications of category scaling, the pro- tion (Lawless and Malone, 1986a, b). Due to their
cedure specifically instructed subjects to use the cate- simplicity, they are well suited to consumer work.
gories to represent equal spacing. They might also be In addition, they offer some advantages in data cod-
instructed to distribute their judgments over the avail- ing and tabulation for speed and accuracy as they
able scale range, so the strongest stimulus was rated are easier to tabulate than measuring lines or record-
at the highest category and the weakest stimulus at ing the more variable magnitude estimates that may
the lowest category. This use of such explicit instruc- include fractions. This presumes, of course, that the
tions surfaces from time to time. An example is in data are being tabulated manually. If the data are
Anderson’s (1974) recommendation to show the sub- recorded online using a computer-assisted data collec-
ject specific examples of bracketing stimuli that are tion system, this advantage vanishes. A wide variety of
above and below the anticipated range of items in scales with fixed alternatives are in current use. These
the set to be judged. A related method is the relative include Likert-type scales used for opinions and atti-
scaling procedure of Gay and Mead (1992) in which tudes, which are based on the degree to which a person
subjects place the highest and lowest stimuli at the agrees or disagrees with a statement about the prod-
scale endpoints (discussed below). The fact that there uct. Examples of such scales used in consumer tests
is an upper boundary to the allowable numbers in a cat- are found in Chapter 14. Lately the term “Likert scale”
egory scaling task may facilitate the achievement of a has been used to refer to any category type of scale,
linear interval scale (Banks and Coleman, 1981). but we prefer to reserve the name of Likert to the
A related issue concerns what kind of experience the agree/disagree attitude scale in keeping with his origi-
high end anchor refers to. Muñoz and Civille (1998) nal method (Likert, 1932). The flexibility of categori-
pointed out that for descriptive analysis panels, the cal alternatives for many different situations is thus one
high end-anchor phrase could refer to different situ- important aspect of the appeal of this kind of response
ations. For example, is the term “extremely sweet” measurement.
7.3.2 Line Scaling a)

very poor excellent
A second widely used technique for intensity scaling AROMA

involves making a mark or slash on a line to indi-
cate the intensity of some attribute. Marking a line is b)
also referred to as using a graphic-rating scale or a weak moderate strong
visual analog scale. The response is recorded as the
distance of the mark from one end of the scale, usu- Pepper Heat
ally whatever end is considered “lower.” Line marking O
c)
differs from category scales in the sense that the per-
son’s choices seem more continuous and less limited. threshold slight moderate strong
In fact, the data are limited to the discrete choices mea- Sweetness
surable by the data-encoding instrument, such as the d)
resolution of a digitizer or the number of pixels resolv-
weaker reference stronger
able on a computer screen. The fundamental idea is
that the panelist makes a mark on a line to indicate overall opinion
the intensity or amount of some sensory characteris- e)
tic. Usually only the endpoints are labeled and marked dislike neither like
with short line segments at right angles to the main moderately moderately
line. The end-anchor lines may be indented to help
avoid end effects associated with the reluctance of sub- f) X X
jects to use the ends of scale. Other intermediate points least most
liked liked
may be labeled. One variation uses a central reference
point representing the value of a standard or baseline Fig. 7.3 Examples of line-marking scales: (a) with endpoints
product on the scale. Test products are scaled relative labeled (after Baten, 1946); (b) with indented “goal posts” (after
to that reference. Some of these variations are shown in Mecredy et al., 1974); (c) with additional points labeled as in
ASTM procedure E-1083 (ASTM, 2008b); (d) a line for ratings
Fig. 7.3. These techniques are very popular in descrip-
relative to a reference as in Stoer and Lawless (1993); (e) hedo-
tive analysis in which multiple attributes are evaluated nic scaling using a line; (f) the adjustable scale of Gay and Mead
by trained panels. (1992) as pictured by Villanueva and D Silva (2009).
The first use of line scales for sensory evaluation
appears in an experiment from the Michigan State
Agricultural Experiment Station conducted during
An important historical trend to use lines in descrip-
World War II (Baten, 1946). Various storage temper-
tive analysis was instrumental in the popularization of
atures of apples were tested. A simple category scale
line marking. Stone et al. (1974) recommended the use
for fruit appeal was used (ratings from very desirable
of line marking for Quantitative Descriptive Analysis
to very undesirable with seven alternatives) and then a
(QDA), then a relatively new approach to specifying
6 in. line scale was used, with the words “very poor”
the intensities of all important sensory attributes. It
over the left end and “excellent” over the right end.
was important to have a scaling method in QDA that
Responses on the line were measured in inches. A poll
approximated an interval scale, as analysis of variance
of the participants revealed a strong preference for the
was to become the standard statistical technique for
category scale over the line marking scale. However,
comparing products in descriptive analysis. The justi-
Baten reported that the t-values comparing apples were
fication for the application in QDA appears to rest on
about twice as large using the line-marking technique,
the previous findings of Baten regarding the sensitivity
implying a greater ability to statistically differentiate
of the method and the writings of Norman Anderson
the items using the line scale. Unfortunately, Baten
on functional measurement theory (Anderson, 1974).
did not report any numerical values for the t-statistics,
In his approach, Anderson used indented end anchors
so it is difficult to evaluate the size of the advantage
(see Weiss, 1972, for another example). Anderson
he saw.
also showed his subjects examples of the high and
156 7 Scaling
low stimuli that one might encounter in the experi- with a cursor that moves up and down via the com-
ment, and sometimes even more extreme examples, in puter mouse. Time–intensity methods are reviewed
order to orient and stabilize their scale usage. Whether more fully in Chapter 8.
such examples make sense for sensory evaluation is
questionable.
Since the advent of QDA, the line-marking tech-
niques have been applied in many different situa- 7.3.3 Magnitude Estimation
tions requiring sensory response. In an early sen-
sory application, Einstein (1976) successfully used a 7.3.3.1 The Basic Techniques
line-marking scale with consumers to evaluate beer
attributes of flavor intensity, body, bitterness, and after- A popular technique for scaling in psychophysical
taste. By “successful” we mean that statistically sig- studies has been the method of magnitude estimation.
nificant differences were obtained among test samples. In this procedure, the respondent is instructed to assign
The use of line marking is not limited to foods and con- numbers to sensations in proportion to how strong the
sumer products. Measurement of pain and pain relief sensation seems. Specifically, the ratios between the
has employed line marking in clinical settings, using numbers are supposed to reflect the ratios of sensation
both vertical and horizontal lines (Huskisson, 1983; magnitudes that have been experienced. For example,
Sriwatanakul et al., 1983). Lawless (1977) used a line- if product A is given the value of 20 for sweetness
marking technique along with ratio instructions for intensity and product B seems twice as sweet, B is
both intensity and hedonic judgments in taste and odor given a magnitude estimate of 40. The two critical parts
mixture studies. This was a hybrid procedure in which of the technique are the instructions given to the partic-
subjects were instructed to mark lines as if perform- ipant and the techniques for data analysis. Two primary
ing magnitude estimation. For example, if one product variations of magnitude estimation have been used. In
were twice as sweet as a previous item, a mark would one method, a standard stimulus is given to the sub-
be made twice as far down the line (which could be ject as a reference and that standard is assigned a fixed
extended if the panelist ran out of room). Villanueva value such as 10. All subsequent stimuli are rated rel-
and colleagues used a scale with equally spaced dots ative to this standard, sometimes called a “modulus.”
along the line and obtained good results for acceptabil- It is often easier for panelists if the reference (i.e., the
ity scaling (Villanueva and Da Silva, 2009; Villanueva item used as the modulus) is chosen from somewhere
et al., 2005). In comparisons of category ratings, line near the middle of the intensity range.
marking, and magnitude estimation, the line-marking In the other variation of magnitude estimation, no
method is about as sensitive to product differences standard stimulus is given and the participant is free
as other scaling techniques (Lawless and Malone, to choose any number he or she wishes for the first
1986a, b). sample. All samples are then rated relative to this first
Marking a point on a line has also been used widely intensity, although in practice people probably “chain”
in time–intensity scaling methods. The simplest ver- their ratings to the most recent items in the series.
sion of this is to move a pointer along a scale while a Because people can choose different ranges of num-
moving roll of paper is marked to see the continuous bers in this “non-modulus” magnitude estimation, the
changes in sensation over time. Originally this could data have to be treated to bring all judgments into the
be done with a simple marking pen held by the partic- same range, an extra step in the analysis. Variations on
ipant (e.g., Moore and Shoemaker, 1981). The record magnitude estimation and guidelines for data analysis
of the pen track would usually be obscured from the are found in ASTM Standard Test Method E 1697–05
person’s view, so as not to exert any influence on their (ASTM, 2008a).
response. The pen-tracking method has also been used In the psychophysical laboratory, where magnitude
with ratio instructions (Lawless and Skinner, 1979). estimation has found its primary usage, generally only
In some cases, the participant has turned a dial or one attribute is rated at a time. However, rating mul-
other response device while observing a linear display tiple attributes or profiling has been used in taste
(Lawless and Clark, 1992). Often the time–intensity studies (McBurney and Bartoshuk, 1973; McBurney
scale will look much like a vertical “thermometer” and Shick, 1971; McBurney et al., 1972) and this can
naturally be extended to foods with multiple taste and intensity. Please rate all other samples in proportion to
aromatic attributes. Magnitude estimation has not been this reference. For example, if the next sample is twice as
sweet, assign it a value of “20”, if half as sweet, assign it
used very often for descriptive analysis, but in princi-
a value of “5” and if 3.5 times as sweet, assign in a value
ple there is no reason why it could not be used for that of 35. In other words, rate the sweetness intensity so that
purpose. your numbers represent the ratios among the intensities of
Participants should be cautioned to avoid falling sweetness. You may use any positive numbers including
fractions and decimals.
into previous habits of using bounded category scales,
e.g., limited ranges of numbers from zero to ten. This The other major variation on this method uses no
may be a difficult problem with previously trained pan- reference. In this case the instructions may read as
els that have used a different scaling method, as people follows:
like to stick with a method they know and feel com-
Please taste the first sample and note its sweetness. Please
fortable with. Panelists who show such behavior may rate all other samples relative to this reference, applying
not understand the ratio nature of the instructions. It numbers to the samples to represent the ratios of sweet-
is sometimes useful with a new panelist to have the ness intensity among the samples. For example, if the
participant perform a warm-up task to make sure they next sample was twice as sweet, you would give it a num-
ber twice as big as the rating assigned to the first sample,
understand the scaling instructions. The warm-up task if half as sweet, assign it a number half as big and if 3.5
can involve estimation of the size or area of differ- times as sweet, assign it a number 3.5 times as big. You
ent geometric figures (Meilgaard et al., 2006) or the may use any positive numbers including fractions and
length of lines (McBurney et al., 1972). Sometimes decimals.
it is desired to have panelists rate multiple attributes
at the same time or to break down overall intensity
into specific qualities. If this kind of “profiling” is 7.3.3.3 Data Treatment
needed, the geometric figures can include areas with
different shading or the lines can be differently col- In non-modulus methods, participants will generally
ored. A practice task is highly recommended so that the choose some range of numbers they feel comfort-
sensory scientist can check on whether the participant able with. The ASTM procedure suggests having them
understands the task. pick a value between 30 and 100 for the first sample,
Values of zero are allowed in this method as some and avoiding any number that seems small. If partici-
of the products may in fact have or no sensation for a pants are allowed to choose their own number range,
given attribute (like no sweetness in our example). Of it becomes necessary to re-scale each individual’s data
course, the rating of zero should not be used for the ref- to bring them into a common range before statistical
erence material. While the value of zero is consistent analysis (Lane et al., 1961). This will prevent subjects
with common sense for products with no sensation of who choose very large numbers from having undue
some attributes, it does complicate the data analysis as influence on measures of central tendency (means)
discussed below. and in statistical tests. This rescaling process has been
referred to as “normalizing” (ASTM, 2008a) although
it has nothing to do with the normal distribution or
7.3.3.2 Instructions Z-scores. A common method for rescaling proceeds as
follows:
The visual appearance of the ballot in magnitude esti-
mation is not critical; it is the instructions and the (1) Calculate the geometric mean of each individual’s
participant’s comprehension of the ratio nature of the ratings across their data set.
judgments that are important. Some ballots even allow (2) Calculate the geometric mean of the entire data set
the subject/participant to view all previous ratings. (of all subjects combined).
Here are sample instructions for the use of magnitude (3) For each subject, construct a ratio of the grand geo-
estimation with a reference sample or modulus with a metric mean of the entire data set to each person’s
fixed number assigned to it: geometric mean. The value of this ratio provides
Please taste the first sample and note its sweetness. This a post hoc individual rescaling factor for each
sample is given the value of “10” for its sweetness subject. In place of the grand geometric mean,
158 7 Scaling
any positive numerator may also be chosen in of intensity, but an open-ended magnitude estimation
constructing this factor, e.g., a value of 100. procedure would allow panelists more freedom to dif-
(4) Multiply each data point for a given person by ferentiate and report variations among very intense
their individual rescaling factor. Do this for all sensations.
participants using their own individual re-scaling With hedonic scaling of likes and dislikes, there is
factors. an additional decision in using magnitude estimation
scaling. Two options have been adopted in using this
These re-scaled data are then analyzed. Note that technique, one employing a single continuum or unipo-
due to the extra data treatment step in this method, it lar scale for amount of liking and the other applying a
is simpler to use the modulus-based variation with a bipolar scale with positive and negative numbers plus
standard reference item. a neutral point (Pearce et al., 1986). In bipolar magni-
Magnitude estimation data are often transformed to tude scaling of likes and dislikes, positive and negative
logs before data analysis (Butler et al., 1987; Lawless, numbers are allowed in order to signify ratios or pro-
1989). This is done primarily because the data tend portions of both liking and disliking (e.g., Vickers,
to be positively skewed or log-normally distributed. 1983). An alternative to positives and negatives is to
There tends to be some high outlying values for any have the respondent merely indicate whether the num-
given sample. Perhaps this is not surprising because ber represents liking or disliking (Pearce et al., 1986).
the scale is open-ended at the top, and bounded by In unipolar magnitude estimation only positive num-
zero at the bottom. Transformation into log data and/or bers (and sometimes zeros) are allowed, with the lower
taking geometric means presents some problems, how- end of the scale representing no liking and higher num-
ever, when the data contain zeros. The log of zero is bers given to represent increasing proportions of liking
undefined. Any attempt to take a geometric mean by (Giovanni and Pangborn, 1983; Moskowitz and Sidel,
calculating the product of N items will yield a zero 1971). It is questionable whether a unipolar scale is
on multiplying. Several approaches have been taken to a sensible response task for the participant, as it does
this problem. One is to assign a small positive value to not recognize the fact that a neutral hedonic response
any zeros in the data, perhaps one-half of the smallest may occur, and that there are clearly two modes of
rating given by a subject (ASTM, 2008a). The resulting reaction, one for liking and one for disliking. If one
analysis, however, will be influenced by this choice. assumes that all items are on one side of the hedo-
Another approach is to use the median judgments in nic continuum—either all liked to varying degrees or
constructing the normalization factor for non-modulus all disliked to varying degrees then the one-sided scale
methods. The median is less influenced by the high makes sense. However, it is a rare situation with foods
outliers in the data than the arithmetic mean. or consumer product testing in which at least some
indifference or change of opinion was not visible in at
least some respondents. So a bipolar scale fits common
7.3.3.4 Applications sense.
For practical purposes, the method of magnitude esti- 7.4 Recommended Practice
mation may be used with trained panels, consumers,
and even children (Collins and Gescheider, 1989).
and Practical Guidelines
However, the data do tend to be a bit more variable
than other bounded scaling methods, especially in the Both line scales and category scales may be used effec-
hands of untrained consumers (Lawless and Malone, tively in sensory testing and consumer work. So we
1986b). The unbounded nature of the scale may make will not expend much effort in recommending one
it especially well suited to sensory attributes where of these two common techniques over another. Some
an upper boundary might impose restrictions on the practical concerns are given below to help the stu-
panelists’ ability to differentiate very intense sensory dent or practitioner avoid some potential problems.
experiences in their ratings. For example, irritative or The category–ratio or labeled magnitude scales may
painful sensations such as chili pepper intensity might facilitate comparisons among different groups, and this
all be rated near the upper bound of a category scale issue is discussed below in Section 7.5.2.
7.4 Recommended Practice and Practical Guidelines 159
7.4.1 Rule 1: Provide Sufficient agree. Bear in mind that most early psychophysics was
Alternatives done on simple attributes like the loudness of a sound
or the heaviness of a weight. In the chemical senses,
with their diverse types of sensory qualities and fuzzy
One major concern is to provide sufficient alternatives consumer vocabulary, this is not so straightforward.
to represent the distinctions that are possible by pan- Other problems to avoid include mixing sensation
elists (Cox, 1980). In other words, a simple 3-point intensity (strength) and hedonics (liking), except in
scale may not suffice if the panel is highly trained the just-right scale where this is explicit. An example
and capable of distinguishing among many levels of of a hedonically loaded sensory term is the adjective
the stimuli. This is illustrated in the case of the flavor “fresh.” Whatever this means to consumers, it is a poor
profile scale, which began with five points to repre- choice for a descriptive scale because it is both vague
sent no sensation, threshold sensation, weak, moderate, and connotes some degree of liking or goodness to
and strong (Caul, 1957). It was soon discovered that most people. Vague terms are simply not actionable
additional intermediate points were desirable for pan- when it comes to giving feedback to product devel-
elists, especially in the middle range of the scale where opers about what needs to be fixed. Another such
many products would be found. However, there is a vague term that is popular in consumer studies is “nat-
law of diminishing returns in allowing too many scale ural.” Even though consumers might be able to score
points—further elaboration allows better differentia- products on some unknown basis using this word, the
tion of products up to a point and then the gains dimin- information is not useful as it does not tell formula-
ish as the additional response choices merely capture tors what to change if a product scores low. A similar
random error variation (Bendig and Hughes, 1953). problem arises with attempting to scale “overall qual-
A related concern is the tendency, especially in ity.” Unless quality has been very carefully defined, it
consumer work, to simplify the scale by eliminating cannot be scaled.
options or truncating endpoints. This brings in the dan-
ger caused by end-use avoidance. Some respondents
may be reluctant to use the end categories, just in case
7.4.3 Rule 3: The Anchor Words Should
a stronger or weaker item may be presented later in
the test. So there is some natural human tendency to
Make Sense
avoid the end categories. Truncating a 9-point scale to
a 7-point scale may leave the evaluator with what is In setting up the scales for descriptive analysis or for a
functionally only a 5-point scale for all practical pur- consumer test, the panel leader should carefully con-
poses. So it is best to avoid this tendency to truncate sider the nature of the verbal end anchors for each
scales in experimental planning. scale as well as any intermediate anchors that may
be needed. Should the scale be anchored from “very
weak” to “very strong” or will there be cases in which
7.4.2 Rule 2: The Attribute Must Be the sensory attribute is simply not present? If so, it
Understood makes sense to verbally anchor the bottom of the scale
with “not at all” or “none.” For example, a sweetness
scale could be anchored with “not at all sweet” and a
Intensity ratings must be collected on an attribute
scale for “degree of oral irritation” could be anchored
which the participants understand and about which
using “none.”
they have a general consensus and agreement as to its
meaning. Terms like sweetness are almost universal
but a term like “green aroma” might be interpreted
in different ways. In the case of a descriptive panel, 7.4.4 To Calibrate or Not to Calibrate
a good deal of effort may be directed at using refer-
ence standards to illustrate what is meant by a specific If a high degree of calibration among the panelists
term. In the case of consumer work, such training is not is desired, then physical standards can be given for
done, so if any intensity ratings are collected, they must intensity. Often this is done with end examples as
be about simple terms about which people generally discussed above, but it may be advantageous to give
160 7 Scaling
examples of intermediate points on the scales as well. across different sensory experiences, mixing qualities.
An example of this kind of calibration is found in the An example is the pseudo-scale used for baked prod-
ASTM procedures for evaluating pepper heat, where ucts, where the number 10 is assigned for perfect
three points on the scale are illustrated (weak, mod- texture, 8 for slight dryness, 6 for gumminess, and 4 if
erate, and approaching strong) (ASTM, 2008b). The very dry (AACC, 1986). Gumminess and dryness are
traditional texture profile technique (Brandt et al., two separate attributes and should be scaled as such.
1963) used nine scale points for most texture attributes This is also an example of quality grading, which is
like hardness and would give examples of common not a true scaling procedure. When the numbers shift
products representing each point on the scale. In the among different sensory qualities, this violates the psy-
Spectrum descriptive method, the scales for intensity chophysical model for scaling the intensity of a single
are intended to be comparable across all attributes and attribute. Although numbers may be applied to the
all products so scale examples are given from various grades, they cannot be treated statistically, as the aver-
sensory domains representing points on the 15-point age of “very dry ” (4) and slightly dry (8) is not gummy
scale for intensity (Meilgaard et al., 2006). Whether (6) (see Pangborn and Dunkley, 1964, for a critique
or not this degree of calibration is required for a spe- of this in the dairy grading arena). The numbers in a
cific project should be considered. There may also be quality-grading scheme do not represent any kind of
a limit to the ability to stabilize the scale usage of unitary psychophysical continuum.
respondents. There are limits on the abilities of humans
to be calibrated as measuring instruments (Olabi and
Lawless, 2008) and in spite of decades of research in
scaling, this is not well understood. People differ in 7.5 Variations—Other Scaling
their sensitivities to various tastes and odors and thus Techniques
may honestly differ in their sensory responses.
Another decision of the test designer will be
An important idea in scaling theory is the notion that
whether to assign physical examples to intermediate
people may have a general idea of how weak or strong
scale points. Although reference items are commonly
sensations are, and that they can compare different
shown for the end categories, this is less often done
attributes of a product for their relative strength, even
with intermediate categories. The apparent advantage
across different sensory modalities. So, for example,
is to achieve a higher level of calibration, a desirable
someone could legitimately say that this product tastes
feature for trained descriptive panelists. A potential
much more salty than it is sweet. Or that the trum-
disadvantage is the restriction of the subject’s use of
pets are twice as loud as the flutes in a certain passage
the scale. What appears to be equal spacing to the
in a symphony. Given that this notion is correct, peo-
experimenter may not appear so to the participant. In
ple would seem to have a general internal scale for
that case, it would seem wiser to allow the respon-
the strength of sensations. This idea forms the basis
dent to distribute his or her judgments among the scale
for several scaling methods. It permits the compari-
alternatives, without presuming that the examples of
son of different sensations cross-referenced by their
intermediate scale points are in fact equally spaced.
numerical ratings, and even can be used to compare
This choice is up to the experimenter. The decision
word responses. Methods derived from this idea are
reflects one’s concerns as to whether it is more desir-
discussed next.
able to work toward calibration or whether there is
more concern with potential biasing or restriction of
responses. 7.5.1 Cross-Modal Matches
and Variations on Magnitude
7.4.5 A Warning: Grading and Scoring Estimation
are Not Scaling
The method of magnitude estimation has a basis in
In some cases pseudo-numerical scales have been set earlier work such as fractionation methods and the
up to resemble category scales, but the examples cut method of sense ratios in the older literature (Boring,
7.5 Variations—Other Scaling Techniques 161
1942) where people would be asked to set one stimulus brightness of the lights about in the same proportions
in a given sensation ratio to another. The notion of as the loudness of the sounds. Stevens (1969) pro-
allowing any numbers to be generated in response posed that these experiments could validate the power
to stimuli, rather than adjusting stimuli to represent law, since the exponents of the cross-modality match-
fixed numbers appeared somewhat later (Moskowitz, ing function could be predicted from the exponents
1971; Richardson and Ross, 1930; Stevens, 1956). An derived from separate scaling experiments. Consider
important outcome of these studies was the finding the following example:
that the resulting psychophysical function generally For one sensory attribute (using the log transform,
conformed to a power law of the following form: Eq. (7.2)),
R = kI n (7.1) log R1 = n1 log I1 + log k1 (7.3)
or after log transformation: and for a second sensory attribute
log R2 = n2 log I2 + log k2 (7.4)

log(R) = n log(I) + log(k) (7.2)
Setting R1 = R2 in cross-modality matching gives
where R was the response, e.g., perceived loudness
(mean or geometric mean of the data) and I was the
n1 log I1 + log k1 = n2 log I2 + log k2 (7.5)
physical stimulus intensity, e.g., sound pressure, and
k was a constant of proportionality that depends upon and rearranging,
the units of measurement. The important characteristic
value of any sensory system was the value for n, the log I1 = (n2 /n1 ) log I2 + a constant (7.6)
exponent of the power function or slope of the straight
line in a log–log plot (Stevens, 1957). The validity If one plots log I1 against log I2 from a cross-
of magnitude estimation then came to hang on the modality matching task, the slope of the function can
validity of the power law—the methods and resulting be predicted from the ratio of the slopes of the individ-
functions formed an internally consistent theoretical ual exponents (i.e., n2 /n1 , which you can derive from
system. Stevens also viewed the method as provid- two separate magnitude estimation tasks). This predic-
ing a direct window into sensation magnitude and did tion holds rather well for a large number of compared
not question the idea that these numbers generated sensory continua (Steven, 1969). However, whether it
by subjects might be biased in some way. However, actually provides a validation for the power law or
the generation of responses is properly viewed as for magnitude estimation has been questioned (e.g.,
combining at least two processes, the psychophysical Ekman, 1964).
transformation of energy into conscious sensation and For practical purposes, it is instructive that people
the application of numbers to those sensations. This can actually take disparate sensory continua and com-
response process was not given due consideration in pare them using some generalized notion of sensory
the early magnitude estimation work. Responses as intensity. This is one of the underpinnings of the use
numbers can exhibit nonlinear transformations of sen- of a universal scale in the Spectrum descriptive proce-
sation (Banks and Coleman, 1981; Curtis et al., 1968) dure (Meilgaard et al., 2006). In that method, different
so the notion of a direct translation from sensation to attributes are rated on 15-point scales that can (in the-
ratings is certainly a dangerous oversimplification. ory) be meaningfully compared. In other words, a 12
Ratio-type instructions have been applied to other in sweetness is twice as intense a sensation as a 6
techniques as well as to magnitude estimation. A his- in saltiness. Such comparisons seem to makes sense
torically important psychophysical technique was that for tastes and flavors but may not cut across all other
of cross-modality matching, in which the sensation modalities. For example, it might seem less sensible to
levels or ratios would be matched in two sensory con- compare the rating given for the amount of chocolate
tinua such as loudness and brightness. One continuum chips in a cookie to the rating given for the cookie’s
would be adjusted by the experimenter and the other hardness—these seem like quite different experiences
by the subject. For example, one would try to make the to quantify.
162 7 Scaling
Cross-modality matches have been performed suc- or line scale? The idea of scaling word phrases takes
cessfully with children even down to age 4, using line shape in the labeled magnitude scales discussed next.
length compared to loudness (Teghtsoonian, 1980).
This may have some advantage for those with limited
verbal skills or trouble understanding the numerical
7.5.2 Category–Ratio (Labeled
concepts needed for category scales or magnitude esti-
mation. The use of line length points out that the line- Magnitude) Scales
marking technique might be considered one form of
a cross-modality matching scale. Some continua may A group of hybrid techniques for scaling has recently
seem simpler, easier, or more “natural” to be matched enjoyed some popularity in the study of taste and
(for example, hand grip force to perceived strength of smell, for hedonic measurement and other applica-
tooth pain). King (1986) matched the pitches of tones tions. One of the problems with magnitude estimation
to concentrations of benzaldehyde and Ward (1986) data is that it does not tell in any absolute sense
used duration as a matching continuum for loudness whether sensations are weak or strong, only giving the
and brightness. One of the advantages of the cross- ratios among them. This group of scales attempts to
modality matching procedure is that it is possible to provide ratio information, but combines it with com-
specify the intensity of a sensation in physical units, mon verbal descriptors along a line scale to provide
i.e., as a physical level on the other continuum. So a simple frame of reference. They are referred to as
sweetness, for example, could be represented in a category–ratio scales, or more recently, labeled mag-
decibel (sound pressure) equivalent. In one amusing nitude scales. They all involve a horizontal or vertical
variation on this idea, Lindvall and Svensson (1974) line with deliberately spaced labels and the panelists’
used hedonic matching to specify the unpleasantness task is to make a mark somewhere along the line to
of combustion toilet fumes to different levels of H2 S indicate the strength of their perception or strength of
gas that were sniffed from an olfactometer. Thus the their likes or dislikes. In general, these labeled line
lucky participant could dial up a concentration that scales give data that are consistent with those from
was perceived as being equally as offensive as the test magnitude estimation (Green et al., 1993). An unusual
samples. characteristic of these scales is the verbal high end-
If line marking can be considered a kind of cross- anchor phrase, which often refers to the “strongest
modality match, then why not the numbers them- imaginable.”
selves? It should be possible to cross-reference one The technique is based on early work by Borg and
continuum to another simply through instructions to colleagues, primarily in the realm of perceived physi-
use a common response scale. Stevens and Marks cal exertion (Borg, 1982, 1990; see Green et al., 1993).
(1980) developed the technique of magnitude matching In developing this scale, Borg assumed that the seman-
to do just that (see also Marks et al., 1992). Subjects tic descriptors could be placed on a ratio scale and
were instructed to judge loudness and brightness on that they defined the level of perceptual intensity and
a common scale of intensity so that if a sound had that all individuals experienced the same perceptual
the same sensory impact as the brightness of a light, range. Borg suggested that for perceived exertion, the
then the stimuli would be given the same number. This maximal sensation is roughly equivalent across people
type of cross-referencing should facilitate comparisons for this sensory “modality” (Marks et al., 1983). For
among people. For example, if it can be assumed that example, it is conceivable that riding a bicycle to the
two people have the same response to salt taste or point of physical exhaustion produces a similar sensory
to loudness of tones, then differences in some other experience for most people. So the scale came to have
continuum like bitter taste, hot chili pepper intensity, the highest label referring to the strongest sensation
or a potential olfactory loss can be cross-referenced imaginable.
through salty taste or through loudness of tones they This led to the development of the labeled mag-
have rated in magnitude matching (e.g., Gent and nitude scale (LMS) shown in Fig. 7.4. It is truly a
Bartoshuk, 1983). Furthermore, if numbers can pro- hybrid method since the response is a vertical line-
vide a cross-referencing continuum, then why not scale marking task but verbal anchors are spaced according
the word phrases used as anchor points on a category to calibration using ratio-scaling instructions (Green
2007; Lawless et al., 2010a, b, c). A growing number

Strongest imaginable (95.5)
of similar scales have been developed for various appli-
cations including oral pleasantness/unpleasantness (the
“OPUS” scale, Guest et al., 2007), perceived satiety
(the “SLIM” scale, Cardello et al., 2005), clothing fab-
ric comfort (the “CALM” scale, Cardello et al., 2003),
and odor dissimilarity (Kurtz et al., 2000). All of these
scales depend upon a ratio scaling task to determine
Very strong (50.12) LMS the spacing of the verbal descriptors and almost all use
a Borg-type high end-anchor phrase. Others will surely
be developed.
Strong (33.1) Instructions to participants have differed in the use
of these scales. In the first application of the LMS,
Green et al. (1993) instructed subjects to first choose
Moderate (16.2) the most appropriate verbal descriptor, and then to
“fine tune” their judgment by placing a mark on the
Weak (5.8)
line between that descriptor and the next most appro-
Barely detectable (1.4)
priate one. In current practice less emphasis may be
Fig. 7.4 The labeled magnitude scale (LMS) of Green et al. placed on the consideration of the verbal labels and
(1993). instructions may be given to simply make a mark “any-
where” on the line. A common observation with the
et al., 1993). In setting up the scale, Green and col- hedonic versions of the scale is that some panelists will
leagues had subjects provide magnitude estimates of mark at or very near a verbal descriptor, seeming to use
different verbal descriptors after giving magnitude it as a category scale (Cardello et al., 2008; Lawless
estimates of familiar oral sensations (e.g., the bitter- et al., 2010a). The proportion of people displaying this
ness of celery, the burn of cinnamon gum). These behavior may depend upon the physical length of the
results were generally in line with previous scaling of line (and not the instructions or examples that may be
verbal descriptors, so-called semantic scaling. Other shown) (Lawless et al., 2010b).
researchers have developed scales with only direct Results may depend in part on the nature of the high
scaling of the verbal descriptors and have not always end-anchor example and the frame of reference of the
included a list of everyday or common experiences subject in terms of the sensory modality they are think-
(Cardello et al., 2003; Gracely et al., 1978a, b). ing about. Green et al. (1996) studied the application
After the introduction of the LMS, a number of of the LMS to taste and odor, using descriptors for
researchers tried to extend this approach into the realm the upper bound as “strongest imaginable” taste, smell,
of hedonics (measuring food acceptability). A widely sweetness, etc. Steeper functions (a smaller response
used scale is the labeled affective magnitude (LAM) range) were obtained when mentioning individual taste
scale developed by Schutz and Cardello (2001). They qualities. This appears to be due to the omission of
used direct ratio scaling of the verbal descriptors of painful experiences (e.g., the “burn of hot peppers”)
the 9-point hedonic scale and included Borg’s type from the frame of reference when sensations were
of high (and low) end anchor (“greatest imaginable scaled relative to only taste. The steepening of the
like/dislike”). The scale is shown in Fig. 7.5. This functions for the truncated frame of reference is con-
shows some advantages in differentiating well-liked sistent with the idea that subjects expanded their range
items (El Dine and Olabi, 2009; Greene et al., 2006; of numbers as seen in other scaling experiments (e.g.,
Schutz and Cardello, 2001), although that finding Lawless and Malone, 1986b). The fact that subjects
is not universal (Lawless et al., 2010a). The LAM appear to adjust their perceptual range depending on
scale or similar versions of it have been applied in instructions or frame of reference suggests that the
a variety of studies with different foods (Chung and scales have relative and not absolute properties, like
Vickers, 2007a, b; El Dine and Olabi, 2009; Forde and most other scaling methods. Cardello et al. (2008)
Delahunty, 2004; Hein et al., 2008; Keskitalo et al., showed that the hedonic version of the scale (the LAM
164 7 Scaling
Fig. 7.5 Affective labeled Greatest imaginable like (+100) Most imaginable (+100)
magnitude scales, including
the LAM scale (Cardello and
Schutz, 2004) and the OPUS
scale (Guest et al., 2007). Like Extremely (+74.2) extremely (+75.2)
Like Very Much (+56.1) very (+56.1)
PLEASANT
Like Moderately (+36.2) moderately (+33.9)
slightly (+19.7)
Like Slightly (+11.2)
weakly (+15.0)
barely detectable (+6.5)

Neither like nor dislike (0) (0)
barely detectable (+6.5)
Dislike slightly (–10.6)
weakly (+15.0)
slightly (+19.7)
Dislike moderately (–31.9) moderately (+33.9)
UNPLEASANT
Dislike Very much (–55.5) very (+56.1)
Dislike extremely (–75.5) extremely (+75.2)
Greatest imaginable dislike (–100) Most imaginable (–100)
LAM OPUS
scale) will also show such range effects. A compressed 2004a, b, 2006) have argued that the labeled magnitude
range of responses is obtained when the frame of scales should anchor their endpoints to “sensations
reference is greatest imaginable like (dislike) for an of any kind” as such a reference experience would
“experience of any kind” rather than something more allow different individuals to use the scale in similar
delimited like “foods and beverages.” Apparently the ways/and thus facilitate inter-individual comparisons.
compression is not very detrimental to the ability of the Scales with this kind of high end anchor have been
LAM scale to differentiate products (Cardello et al., termed “generalized” labeled magnitude scales (or
2008, but see also Lawless et al., 2010a). gLMS). However, the sensory evaluation practitioner
Is this a suitable method for cross-subject com- should be aware of the compression effects that can
parisons? To the extent that Borg’s assumptions of occur with this kind of scale, which could potentially
common perceptual range and the similarity of the lead to lessened differentiation among products.
high end-anchor experience among people are true,
the method might provide one approach to valid
comparisons of the ratings among different respon- 7.5.3 Adjustable Rating Techniques:
dents. This would facilitate comparisons of clinical Relative Scaling
groups or patients with sensory disorders or genetically
different individuals such as anosmics, PTC/PROP A few methods have been tried that allow consumers
taster groups. Bartoshuk and colleagues (1999, 2003, or panelists to change their ratings. An example is
the “rank-rating” technique developed by O’Mahony, intense (or least liked) at the bottom. All other samples
Kim, and colleagues (Kim and O’Mahony, 1998; Lee are distributed along the full scale. This can provide
et al., 2001; O’Mahony et al., 2004; Park et al., 2004). good differentiation of the products, but obviously any
In this method, the consumer has a visual scale in front absolute information about what is weak or strong,
of him or her on the table and after tasting a sample is liked or disliked, is lost (as is the case with magnitude
instructed to physically place the sample on the scale. estimation). Because all the ratings are truly relative,
Subsequent samples are also tasted and placed and the contextual effects such as contrast might be expected to
important feature is that the consumer can change the be larger with such a technique, but this is not known.
position of any previous item based on their perception Another relative scaling method is when panelists
of the new sample(s). This procedure has relatively lit- purposely rate each sample relative to a reference item,
tle to do with rank ordering per se (in fact ties can be which is usually marked on the center of the response
allowed). Cordinnier and Delwiche (2008) chose the scale. Relative-to-reference scaling was studied by
more descriptive name of “positional relative rating” Larson-Powers and Pangborn (1978) and compared to
for this technique. traditional scaling for the descriptive profiling of bev-
O’Mahony and Kim examined the efficiency of this erages and gelatins. Significant differences were found
technique, mostly in simple salt solutions, on the basis with the relative scale (“anchored” in their terms)
of people’s ordering of salt solutions in increasing con- in 22.8% of all possible comparisons as compared
centrations (Kim and O’Mahony, 1998; Park et al., with 19.5% of comparisons using the unanchored
2004). The important data were “reversals” in which scale. However, panelists were given the relative scale
a higher concentration was rated lower than a lower first, and more practice using that scale. In an exten-
one, and vice versa. A good scale would minimize sive study of both trained and untrained respondents,
the amount of reversals. Given this criterion, the rank- Stoer and Lawless (1993) found a similar advantage,
rating has fewer errors than non-adjustable ratings. At with 33% of all possible comparisons were signifi-
this point it is not clear whether this apparent advan- cant for the relative scaling versus 27% for traditional
tage arises because people are allowed to re-taste, or scaling. However, this was not a statistically signif-
that they are allowed to re-position previously tasted icant increase based on meta-analytic comparisons
items. Both factors may be important. There is some (Rosenthal, 1987). The relative-to-reference scale was
evidence that adjustable ratings can produce statisti- also discussed by Land and Shepard (1984), who noted
cally significant ratings with fewer subjects, but the that it facilitates comparisons across occasions that
procedure can take up to twice as long as normal rat- would be otherwise difficult to make. They also warn
ings. A limitation of this technique is that only one that the choice of standard may have an effect on
attribute can be evaluated at a time. If a second or the scaling functions that result. The task is certainly
third attribute is needed, the procedure starts over. This easy for subjects and was touted by Land and Shepard
may be acceptable for consumer like/dislike ratings as showing “good reproducibility” (see also Mahoney
(O’Mahony et al., 2004), but would not be suitable for et al., 1957). Whether the method offers any consis-
a descriptive analysis task. The option to change pre- tent advantage over traditional scaling is questionable.
vious ratings is an interesting notion and is allowed by It may be useful in situations where comparison to a
some of the sensory data-collection software packages reference is a natural feature or explicit objective of
at this time. Whether the option is advantageous should the experiment at hand, for example in quality control
be the subject of further study. Re-tasting (Koo et al., or shelf life studies where an identified control sample
2002; Lee et al., 2001) is a potentially important fea- is used as a baseline for comparison.
ture of this method, and should be evaluated separately,
and in consideration of the extra adaptation, fatigue, or
carryover that could occur with some products.
A completely relative rating procedure is the 7.5.4 Ranking
method of Gay and Mead (Gay and Mead, 1992; Mead
and Gay, 1995). In this task, a panelist inspects the Another alternative to traditional scaling is the use
entire set of samples and places the most intense (or of ranking procedures. Ranking is simply ordering
most liked) at the top end of the scale and the least the products from weakest to strongest on the stated
166 7 Scaling
attribute or from least liked to most liked for consumer ask how many standard deviations apart are the mean
acceptance testing. Ranking has the advantages of sim- values for two products. This is a different approach
plicity in instructions to subjects, simplicity in data to measurement than simply asking how many scale
handling, and minimal assumptions about level of mea- units separate the means on the response scale. On
surement since the data are treated as ordinal. Although a 9-point scale, one product may receive mean rat-
ranking tests are most often applied to hedonic data, ing of seven, and another nine, making the difference
they are also applicable to questions of sensory inten- two scale units. If the pooled standard deviation is two
sity. When asked to rank items for the intensity of a units, however, they would only be one unit apart on
specific attribute, e.g., the sourness of taste of several a variability-based scale. As one example, Conner and
fruit juices, for example, the ranking test is merely Booth (1988) used both the slope and the variance of
an extension of a paired comparison procedure into functions from just-right scales to derive a “tolerance
more than two products. Due to its simplicity, rank- discrimination ratio.” This ratio represents a measure
ing may be an appropriate choice in situations where of the degree of difference along a physical contin-
participants would have difficulty understanding scal- uum (such as concentration of sugar in lime drink) that
ing instructions. In working with non-literates, young observers find to make a meaningful change in their
children, across cultural boundaries or in linguistically ratings of difference-from-ideal (or just-right). This is
challenging situations, ranking is worth considering analogous to finding the size of a just-noticeable differ-
(Coetzee and Taylor, 1996). This is especially true if ence, but translated into the realm of hedonic scaling.
decisions are likely to be very simple, e.g., whether Their insight was that it is not only the slope of the
or not two juice samples differ in perceived sour- line that is important in determining this tolerance or
ness. Ranking may allow differentiation of products liking-discrimination function, but also the variance
that are all similar in acceptability. With medications, around that function.
for example, all formulas may be to some degree Variability-based scales are the basis for scal-
unpalatable. It might be useful then to use ranking in ing in Thurstone’s models for comparative judgment
choosing alternative flavorings in order to find the least (Thurstone, 1927) and its extension into determin-
offensive. ing distances between category boundaries (Edwards,
Analysis of ranked data is straightforward. Simple 1952). Since the scale values can be found from choice
rank-sum statistics can be found in the tables pub- experiments as well as rating experiments, the tech-
lished by Basker (1988) and Newell and MacFarlane nique is quite flexible. How this type of scaling can
(1987, see also Table J). Another very sensitive test be applied to rating data is discussed below in the
of differences in ranked data is the Friedman test, derivation of the 9-point hedonic scale words. When
also known as the “analysis of variance on ranks.” the scale values are derived from a choice method like
These are discussed in Appendix B. The tests are rapid, the triangle test or paired comparison method, this is
straightforward, and easy to perform. It is also possible sometimes called “indirect scaling” (Baird and Noma,
to convert other data to rankings. This is a conservative 1978; Jones, 1974). The basic operation in Thurstonian
approach if the interval nature of the data is in ques- scaling of choice data is to convert the proportion cor-
tion or when violations of statistical assumptions such rect in a choice experiment (or simply the proportions
as the normality of the data are suspect. For exam- in a two-tailed test like paired preference) to Z-scores.
ple, Pokorńy et al. (1986) used ranking analysis of The exact derivation depends upon the type of test
line-marking data to compare the profiles of different (e.g., triangle versus 3-AFC) and the cognitive strat-
raspberry beverages sweetened with aspartame. egy used by the subject. Tables for Thurstonian scale
values from various tests such as the triangle test were
given by Frijters et al. (1980) and Bi (2006) and some
tables are given in the Appendix. Mathematical details
7.5.5 Indirect Scales of Thurstonian scaling are discussed in Chapter 5.
Deriving measures of sensory differences in such
A conservative approach to scaling is to use the vari- indirect ways presents several problems in applied sen-
ance in the data as units of measurement, rather than sory evaluation so the method has not been widely
the numbers taken at face value. For example, we could used. The first problem is one of economy in data
7.6 Comparing Methods: What is a Good Scale? 167
collection. Each difference score is derived from a a consumer is asked to choose the best liked and least
separate discrimination experiment such as a paired liked samples from a set of three or more items (Jaeger
comparison test. Thus many subjects must be tested et al., 2008). With three products, it can be considered a
to get a good estimate of the proportion of choice, form of a ranking task. When applied to sensory inten-
and this yields just one scale value. In direct scaling, sity, this is sometimes known as maximum-difference
each participant gives at least one data point for each or “max-diff.” Best–worst scaling is also discussed in
item tasted. Direct scaling allows for easy comparisons Section 13.7. Simple difference scores may be calcu-
among multiple products, while the discrimination test lated based on the number of times an item is called
must be done on one pair at a time. Thus the methods best versus worst and these scores are supposed to have
of indirect scaling are not cost-efficient. interval properties. If a multinomial logistic regression
A second problem can occur if the products are too is performed on the data, they are theorized to have true
clearly different on the attribute in question, because ratio properties (Finn and Louviere, 1992). A practi-
then the proportion correct will approach 100%. At cal problem with the method, however, is that so many
that point the scale value is undefined as they are products must be tasted and compared, rendering it
some unknown number of standard deviations apart. difficult to perform with foods (Jaeger and Cardello,
So the method only works when there are small dif- 2009).
ferences and some confusability of the items. In a The sensory professional should bear in mind that
study of many products, however, it is sometimes pos- in spite of their theoretical sophistication, the indirect
sible to compare only adjacent or similar items, e.g., methods are based on variability as the main deter-
products that differ in small degrees of some ingredi- minant of degree of difference. Thus any influence,
ent or process variable. This approach was taken by which increases variability, will tend to decrease the
Yamaguchi (1967) in examining the synergistic taste measured differences among products. In the well-
combination of monosodium glutamate and disodium controlled psychophysical experiment under constant
5 inosinate. Many different levels of the two ingredi- standard conditions across sessions and days, this may
ents were tasted, but because the differences between not be important—the primary variability lies in the
some levels were quite apparent, an incomplete design resolving power of the participant (and secondarily
was used in which only three adjacent levels were in the sample products). But in drawing conclusions
compared. across different days, batches, panels, factories, and
Other applications have also used this notion of such, one has a less pure situation to consider. Whether
variability as a yardstick for sensory difference or sen- one considers the Thurstonian-type indirect measures
sation intensity. The original approach of Fechner in comparable across different conditions depends upon
constructing a psychophysical function was to accu- the control of extraneous variation.
mulate the difference thresholds or just-noticeable dif-
ferences (JNDs) in order to construct the log function
of psychophysical sensory intensity (Boring, 1942;
7.6 Comparing Methods: What is a Good
Jones, 1974). McBride (1983a, b) examined whether
JND-based scales might give similar results to cate- Scale?
gory scales for taste intensity. Both types of scaling
yielded similar results, perhaps not surprising since A large number of empirical studies have been
both tend to conform to log functions. In a study of conducted comparing the results using different
children’s preferences for different odors Engen (1974) scaling methods (e.g., Birnbaum, 1982; Giovanni and
used a paired preference paradigm, which was well Pangborn, 1983; Hein et al., 2008; Jaeger and Cardello,
suited to the abilities of young children to respond in a 2009; Lawless and Malone, 1986a, b; Lawless et al.,
judgment task. He then converted the paired preference 2010a; Marks et al., 1983; Moskowitz and Sidel, 1971;
proportions to Thurstonian scale values via Z-scores Pearce et al., 1986; Piggot and Harper, 1975; Shand
and was able to show that the hedonic range of children et al., 1985; Vickers, 1983; Villanueva and Da Silva,
was smaller than that of adults. 2009; Villanueva et al., 2005). Because scaling data are
Another example of choice data that can be con- often used to identify differences between products,
verted to scale values is best–worst scaling, in which the ability to detect differences is one important
168 7 Scaling
practical criterion for how useful a scaling method was some small tendency for magnitude estimation
can be (Moskowitz and Sidel, 1971). A related crite- to be marginally more variable in the hands of con-
rion is the degree of error variance or similar measures sumers as opposed to college students (Lawless and
such as size of standard deviations or coefficients of Malone, 1986b). Statistical differentiation increased
variation. Obviously, a scaling method with low inter- over replicates, as would be expected as people came
individual variability will result in more sensitive tests, to understand the range of items to be judged (see Hein
more significant differences, and lower risk of Type II et al., 2008, for another example of improvement over
error (missing a true difference). A related issue is the replication in hedonic scaling). Similar findings for
reliability of the procedure. Similar results should be magnitude estimation and category scales in terms of
obtained upon repeated experimentation. product differentiation were found by Moskowitz and
Other practical considerations are important as well. Sidel (1971), Pearce et al. (1986), Shand et al. (1985),
The task should be user friendly and easy to understand and Vickers (1983) although the forms of the mathe-
for all participants. Ideally, the method should be appli- matical relations to underlying physical variables was
cable to a wide range of products and questions, so that often different (Piggot and Harper, 1975). In other
the respondent is not confused by changes in response words, as found by Stevens and Galanter (1957) there
type over a long ballot or questionnaire. If panelists is often a curvilinear relationship between the data
are familiar with one scale type and are using it effec- from the two methods. However, this has not been
tively, there may be some liability in trying to introduce universally observed and sometimes simple linear rela-
a new or unfamiliar method. Some methods, like cate- tionships have been found (Vickers, 1983). Similar
gory scales, line scales, and magnitude estimation, can results for category scales and line scales were found
be applied to both intensity and hedonic (like–dislike) by Mattes and Lawless (1985).
responses. The amount of time required to code, tab- Taken together, these empirical studies paint a pic-
ulate and process the information may be a concern, ture of much more parity among methods than one
depending upon computer-assisted data collection and might suppose given the number of arguments over the
other resources available to the experimenters. validity of scaling methods in the psychological lit-
As in any method, validity or accuracy are also erature. With reasonable spacing of the products and
issues. Validity can only be judged by reference to some familiarization with the range to be expected,
some external criterion. For hedonic scaling, one might respondents will distribute their judgments across the
want the method to correspond to other behaviors such available scale range and use the scale appropriately
as choice or consumption (Lawless et al., 2010a). A to differentiate the products. A reasonable summary of
related criterion is the ability of the scale to identify or the literature comparing scale types is that they work
uncover consumer segments with different preferences about equally well to differentiate products, given a
(Villanueva and Da Silva, 2009). few sensible precautions.
Given these practical considerations, we may then
ask how the different scaling methods fare. Most
published studies have found about equal sensitivity
for the different scaling methods, provided that the 7.7 Issues
methods are applied in a reasonable manner. For exam-
ple, Lawless and Malone (1986a, b) performed an 7.7.1 “Do People Make Relative
extensive series of studies (over 20,000 judgments)
with consumers in central location tests using differ-
Judgments” Should They See Their
ent sensory continua including olfaction, tactile, and Previous Ratings?
visual modalities. They compared line scales, magni-
tude estimation, and category scales. Using the degree Baten’s (1946) report of an advantage for the line
of statistical differentiation among products as the cri- scale is illustrative of how observant many researchers
terion for utility of the methods, the scales performed were in the older literature. He noted that a category
about equally well. A similar conclusion was reached scale with labeled alternatives might help some judges,
by Shand et al. (1985) for trained panelists. There but could hinder others by limiting the alternative
7.7 Issues 169
responses (i.e., judgments that might fall in-between 7.7.3 Is Magnitude Estimation a Ratio
categories). The line scale offers a continuously graded Scale or Simply a Scale with Ratio
choice of alternatives, limited only by the measurement
Instructions?
abilities in data tabulation. Baten also noted that the
line scale seemed to facilitate a relative comparison
among the products. This was probably due to his In magnitude estimation, subjects are instructed to use
placement of the scales one above the other on the bal- numbers to reflect the relative proportions between
lot, so judges could see both markings at the same time. the intensities experienced from different stimuli. A
In order to minimize such contextual effects it is now beverage that is twice as sweet as another should
more common to remove the prior ratings for products be given a response number that is twice as large.
to achieve a more independent judgment of the prod- S. S. Stevens felt that these numbers were accurate
ucts. However, whether that is ever achieved in practice reflections of the experience, and so the scale had
is open to question—humans are naturally compara- ratio properties. This assumed a linear relationship
tive when asked to evaluate items, as discussed in between the subjective stimulus intensities (sensations
Chapter 9. Furthermore, there may be potential for or percepts) and the numerical responses. However,
increased discrimination in methods like the rela- there is a wealth of information showing that the pro-
tive positioning technique. The naturally comparative cess of numerical assignment is prone to a series of
nature of human judgment may be something we could contextual and number usage biases that strongly ques-
benefit from rather than trying to fight this tendency by tion whether this process is linear (Poulton, 1989).
over-calibration. So Stevens’ original view of accepting the numeri-
cal reporting as having face validity seems misplaced.
Although it would be advantageous to achieve a level
of measurement that did allow conclusions about pro-
portions and ratios (“this is liked twice as much as
7.7.2 Should Category Rating Scales Be
that”), this seems not fully justified at this time. It is
Assigned Integer Numbers in Data important to differentiate between a method that has
Tabulation? Are They Interval ratio-type instructions, and one that yields a true ratio
Scales? scale of sensation magnitude, where the numbers actu-
ally reflect proportions between intensities of sensory
There is also a strong suspicion that many numeri- experiences.
cal scaling methods may produce only ordinal data,
because the spacing between alternatives is not subjec-
tively equal. A good example is the common market-
ing research scale of “excellent—very good—good— 7.7.4 What is a “Valid” Scale?
fair—poor.” The subjective spacing between these
adjectives is quite uneven. The difference between two An ongoing issue in psychophysics is what kind of
products rated good and very good is a much smaller scale is a true reflection of the subject’s actual sen-
difference than that between products rated fair and sations? From this perspective, a scale is valid when
poor. However, in analysis we are often tempted to the numbers generated reflect a linear translation of
assigned numbers one through five to these categories subjective intensity (the private experience). It is well
and take means and perform statistics as if the assigned established that category scales and magnitude esti-
numbers reflected equal spacing. This is a pretense at mates, when given to the same stimuli, will form a
best. A reasonable analysis of the 5-point excellent to curve when plotted against one another (Stevens and
poor scale is simply to count the number of respon- Galanter, 1957). Because this is not a linear relation-
dents in each category and to compare frequencies. ship, one method or the other must result from a
Sensory scientists should not assume that any scale has non-linear translation of the subjective intensities of
interval properties in spite of how easy it is to tabulate the stimuli. Therefore, by this criterion, at least one
data as an integer series. scale must be “invalid.”
170 7 Scaling
Anderson (1974, 1977) proposed a functional mea- Others have found support for the validity of mag-
surement theory to address this issue. In a typical nitude estimation in studies of binaural loudness sum-
experiment, he would ask subjects to do some kind mation (Marks, 1978). This argument continues and
of combination task, like judging the total combined is difficult to resolve. A review of the matter was
intensity of two separately presented items (or the aver- published by Gescheider (1988). For the purposes of
age lightness of two gray swatches). He would set up sensory evaluation, the issue is not terribly important
a factorial design in which every level of one stimu- for two reasons. First, any scale that produces statisti-
lus was combined with every level of the other (i.e., a cally significant differentiation of products is a useful
complete block). When plotting the response, a family scale. Second, the physical ranges over which category
of lines would be seen when the first stimulus contin- scaling and magnitude estimation produce different
uum formed the X-axis and the second formed a family results is usually quite large in any psychophysical
of lines. Anderson argued that only when the response study. In most product tests, the differences are much
combination rule was additive, and the response output more subtle and generally do not span such a wide
function was linear, would a parallel plot be obtained dynamic range. The issue dissolves from any practical
(i.e., there would be no significant interaction term in perspective.
ANOVA). This argument is illustrated in Fig. 7.6. In
his studies using simple line and category scales, par-
allelism was obtained in a number of studies, and thus
he reasoned that magnitude estimation was invalid by 7.8 Conclusions
this criterion. If magnitude estimation is invalid, then
its derivatives such as the LMS and LAM scales are Much sound and fury has been generated over the years
similarly suspect. in the psychophysical literature concerning what meth-
ods yield valid scales. For the sensory practitioner,
these issues are less relevant because the scale val-
ues do not generally have any absolute meaning. They
S1 p1 are only convenient indices of the relative intensities
P R or appeal of different products. The degree of differ-
S2 p2 ence may be a useful piece of information, but often
Response
Psychophysical Integration output we are simply interested in which product is stronger
process process process
or weaker in some attribute, and whether the differ-
ence is both statistically significant and practically
meaningful.
Scaling provides a quick and useful way to get
Response for Combination Judgment
(High)
intensity or liking information. In the case of descrip-
tive analysis, scaling allows collection of quantitative
(Med.) Levels of Stimulus data on multiple attributes. The degree of variability or
or Variable 2
noise in the system is, to a large part, determined by
whether the panelists have a common frame of refer-
ence. Thus reference standards for both the attribute
(Low)
terms and for the intensity anchors are useful. Of
course, with consumer evaluations or a psychophysical
study such calibration is not possible and usually not
desired. The variability of consumer responses should
offer a note of caution in the interpretation of consumer
scaling data.
(Low) (Med.) (High)
Students and sensory practitioners should exam-
Levels of Stimulus or Variable 1
ine their scaling methods with a critical eye. Not
Fig. 7.6 The functional measurement scheme of Anderson every task that assigns numbers will have useful scale
(1974). properties like equal intervals. Bad examples abound
Appendix 1 171
in the commodity grading (quality scoring) literature for categorical judgment as a means of measuring the
(Pangborn and Dunkley, 1964). For example, different psychological-scale values for the words. This model
numbers may be assigned to different levels of oxida- is at its most simple form when the items to be scaled
tion, but that is scoring a physical condition based on show normal distributions of equal variance.
inferences from sensory experience. It is not a report of Which leads us to the numerical method. Jones
the intensity of some experience itself. It is not track- and Thurstone modified a procedure used earlier by
ing changes along a single perceptual continuum in the Edwards (1952). A description of the process and
psychophysical sense. Scoring is not scaling. results can be found in the paper “Development of
a scale for measuring soldiers’ food preferences”
All hedonic scales seem to measure what they are
intended to measure rather effectively, as long as no gross by Jones et al. (1955). Fifty-one words and phrases
mistakes are made (Peryam, 1989, p. 23). formed the candidate list based on a pilot study with
900 soldiers chosen to be a representative sample of
enlisted personnel. Each phrase was presented on a
form with a rating scale from –4 to +4 with a check off
Appendix 1: Derivation format. In other words, each person read each phrase
and assigned in an integer value from –4 to +4 (includ-
of Thurstonian-Scale Values
ing zero as an option). This method would seem to
for the 9-Point Scale presume that these integers were themselves an inter-
val scale of psychological magnitude, an assumption
The choice of adjective words for the 9-point hedonic that to our knowledge has never been questioned.
scale is a good example of how carefully a scale can Of course, the mean scale values could now
be constructed. The long-standing track record of this be assigned on a simple and direct basis, but the
tool demonstrates its utility and wide applicability in Thurstonian methods do not use the raw numbers as
consumer testing. However, few sensory practitioners the scale, but transform them to use standard devia-
actually know how the adjectives were found and what tions as the units of measurement. So the scale needs
criteria were brought to bear in selecting these descrip- to be converted to Z-score values. The exact steps are
tors (slightly, moderately, very much, and extremely as follows:
like/dislike) from a larger pool of possible words. The
goal of this section is to provide a shorthand descrip- 1. Accumulate frequency counts for all the tested
tion of the criteria and mathematical method used to words across the –4 to + 4 scale. Think of these
select the words for this scale. categories as little “buckets” into which judgments
One concern was the degree to which the term have been tossed.
had consensual meaning in the population. The most 2. Find the marginal proportions each value from –4 to
serious concern was when a candidate word had an +4 (summed across all test items). Add up the pro-
ambiguous or double meaning across the population. portions from lowest to highest to get a cumulative
For example, the word “average” suggests an interme- proportion for each bucket.
diate response to some people, but in the original study 3. Convert these proportions to z-scores in order to
by Jones and Thurstone (1955) there were a group of re-scale the boundaries for the original −4 to +4
people who equated it with “like moderately” perhaps cutoffs. Let us call these the “category z-values”
since an average product in those days was one that for each of the “buckets.” The top bucket will
people would like. These days, one can think of nega- have a value of 100%, so it will have no z-score
tive connotations to the word “average” as in “he was (undefined/infinite).
only an average student.” Other ambiguous or bimodal 4. Next examine each individual item. Sum its individ-
terms were “like not so much” and “like not so well.” ual proportions across the categories, from where
Ideally, a term should have low variability in meaning, it is first used until 100% of the responses are
i.e., a low standard deviation, no bimodality, and lit- accumulated.
tle skew. Part of this concern with the normality of the 5. Convert the proportions for the item to Z-scores.
distribution of psychological reactions to a word was Alternatively, you can plot these proportions on
the fact that the developers used Thurstone’s model “cumulative probability paper,” a graphing format
172 7 Scaling
that marks the ordinate in equal standard deviations proportions for each phrase and the original categories.
units according to the cumulative normal distribu- The dashed vertical lines dropped from the intersection
tion. Either of these methods will tend to make at the zero z-score (50% point) show the approximate
the cumulative S-shaped curve for the item into a mean values interpolated on the X-axis (i.e., about –1.1
straight line. The X-axis value for each point is the for “do not care for it” and about +2.1 for “preferred.”).
“category z-value” for that bucket. Note that “preferred” and “don’t care for it” have a lin-
6. Fit a line to the data and interpolate the 50% point ear fit and steep slope, suggesting a normal distribution
on the X-axis (the re-scaled category boundary esti- and low standard deviation. In contrast, “highly unfa-
mates). These interpolated values for the median for vorable” has a lower slope and some curvilinearity,
each item now form the new scale values for the indicative of higher variability, skew, and/or pockets of
items. disagreement about the connotation of this term.
The actual scale values for the original adjectives
An example of this interpolation is shown in are shown in Table 7.2, as found with a soldier popu-
Fig. 7.7. Three of the phrases used in the original scal- lation circa 1950 (Jones et al., 1955). You may note
ing study of Jones and Thurstone (1955) are pictured, that the words are not equally spaced, and that the
three that were not actually chosen but for which we “slightly” values are closer to the neutral point than
have approximate proportions and z-scores from their some of the other intervals, and the extreme points are
figures. The small vertical arrows on the X-axis show a little farther out. This bears a good deal of similarity
the scale values for the original categories of −4 to to the intervals found with the LAM scale as shown in
+3 (+4 has cumulative proportion of 100% and thus the column where the LAM values are re-scaled to the
the z-score is infinite). Table 7.1 gives the values and same range as the 9-point Thurstonian Values.
Appendix 2: Construction of Labeled

Magnitude Scales
“ ” There are two primary methods for constructing

labeled magnitude scales and they are very similar.
Both require magnitude estimates from the participants
“ ‘ ” to scale the word phrases used on the lines. In one case,
just the word phrases are scaled, and in the second
method, the word phrases are scaled among a list of
common everyday experiences or sensations that most
“ ” people are familiar with. The values obtained by the
simple scaling of just the words will depend upon the
words that are chosen, and extremely high examples
(e.g., greatest imaginable liking for any experience)
will tend to compress the values of the interior phrases
(Cardello et al., 2008). Whether this kind of context
effect will occur for the more general method of scal-
ing amongst common experiences is not known. But
the use of a broad frame of reference could be a
stabilizing factor.
Fig. 7.7 An illustration of the method used to establish spacings Here is an example of the instructions given to sub-
and scale values for the 9-point hedonic scale using Thurstonian jects in construction of a labeled affective magnitude
theory. Arrows on the X-axis show the scale points for the scale. Note that for hedonics, which are a bipolar con-
z-scores based on the complete distribution of the original –4
tinuum with a neutral point, it is necessary to collect
to +4 ratings. The Y-axis shows the actual z-scores based on the
proportion of respondents using that category for each specific a tone or valence (plus or minus) value as well as the
term. Re-plotted from data provided in Jones et al. (1955). overall “intensity” rating.
Appendix 2 173
Table 7.1 Examples of scaled phrases used in Fig. 7.7

Original “Preferred” “Do not care for it” “Highly unfavorable”
category Proportion Z-score Proportion Z-score Proportion Z-score Proportion Z-score
4 1.000 (undef.) 0.80 0.84
3 0.999 3.0 0.50 0.00 0.96 1.75
2 0.983 2.1 0.20 −0.84 0.93 1.48
1 0.882 1.2 0.07 −1.48 0.92 1.41
0 0.616 0.3 0.03 −1.88 0.96 1.75 0.90 1.28
−1 0.383 −0.3 0.83 0.95 0.86 1.08
−2 0.185 −0.9 0.55 0.13 0.84 0.99
−3 0.068 −1.5 0.30 −0.52 0.82 0.92
−4 0.008 −2.4 0.14 −1.08 0.46 −0.10
Table 7.2 Actual 9-point Scale value

scale phrase values and
Descriptor (9-point) Interval LAM value LAM rescaled Interval
comparison to the LAM
values Like extremely 4.16 74.2 4.20
Like very much 2.91 1.26 56.1 3.18 1.02
Like moderately 1.12 1.79 36.2 2.05 1.13
Like slightly 0.69 0.43 11.2 0.63 1.52
Neither like nor dislike 0.00 0.69 0.0 0.00 0.63
Dislike slightly −0.59 0.59 −10.6 −0.60 0.60
Dislike moderately −1.20 0.61 −31.9 −1.81 1.21
Dislike very much −2.49 1.29 −55.5 −3.14 1.33
Dislike extremely −4.32 1.83 −75.5 −4.28 1.14
Next to each word label a response area appeared word/phrase your numerical judgment should be made
similar to this: proportionally and in comparison to the first number. That
is, if you assigned the number 800 to index the strength
Phrase: Tone: + − 0 How much: of the liking/disliking denoted by the first word/phrase
Like extremely __________ _______ and the strength of liking/disliking denoted by the second
word/phrase were twice as great, you would assign the
Words or phrases are presented in random order. number 1,600. If it were three times as great you would
assign the number 2,400, etc. Similarly, if the second
After reading a word they must decide whether the
word/phrase denoted only 1/10 the magnitude of liking
word is positive, negative or neutral and place the cor- as the first, you would assign it the number 80 and so
responding symbol on the first line. If the hedonic tone forth. If any word/phrase is judged to be “neutral” (zero
was not a neutral one (zero value), they are instructed (0) on the first line) it should also be given a zero for its
magnitude rating.
to give a numerical estimate using modulus-free mag-
nitude estimation. The following is a sample of the In the cased of Cardello et al. (2008), positive and
instructions taken from Cardello et al. (2008): negative word labels were analyzed separately. Raw
After having determined whether the phrase is positive
magnitude estimates were equalized for scale range
or negative or neutral and writing the appropriate sym- using the procedure of Lane et al. (1961). All positive
bol (+, −, 0) on the first line, you will then assess the and negative magnitude estimates for a given subject
strength or magnitude of the liking or disliking reflected were multiplied by an individual scaling factor. This
by the phrase. You will do this by placing a number
on the second blank line (under “How Much”). For the
factor was equal to the ratio of the grand geometric
first phrase that you rate, you can write any number you mean (of the absolute value of all nonzero ratings)
want on the line. We suggest you do not use a small across all subjects divided by the geometric mean for
number for this word/phrase. The reason for this is that that subject. The geometric mean magnitude estimates
subsequent words/phrases may reflect much lower lev-
els of liking or disliking. Aside from this restriction you
for each phrase were then calculated based on this
can use any numbers you want. For each subsequent range-equated data. These means became the distance
174 7 Scaling
from the zero point for placement of the phrases along Basker, D. 1988. Critical values of differences among rank
the scale, usually accompanied by a short cross-hatch sums for multiple comparisons. Food Technology, 42(2), 79,
80–84.
mark at that point.
Baten, W. D. 1946. Organoleptic tests pertaining to apples and
pears. Food Research, 11, 84–94.
Bendig, A. W. and Hughes, J. B. 1953. Effect of number
References of verbal anchoring and number of rating scale categories
upon transmitted information. Journal of Experimental
Psychology, 46(2), 87–90.
AACC (American Association of Cereal Chemists). 1986. Bi, J. 2006. Sensory Discrimination Tests and Measurement.
Approved Methods of the AACC, Eighth Edition. Method Blackwell, Ames, IA.
90–10. Baking quality of cake flour, rev. Oct. 1982. The Birch, L. L., Zimmerman, S. I. and Hind, H. 1980. The influence
American Association of Cereal Chemists, St. Paul, MN, of social-affective context on the formation of children’s food
pp. 1–4. preferences. Child Development, 51, 865–861.
Anderson, N. H. 1974. Algebraic models in perception. In: Birch, L. L., Birch, D., Marlin, D. W. and Kramer, L. 1982.
E. C. Carterette and M. P. Friedman (eds.), Handbook Effects of instrumental consumption on children’s food pref-
of Perception. Psychophysical Judgment and Measurement, erences. Appetite, 3, 125–143.
Vol. 2. Academic, New York, pp. 215–298. Birnbaum, M. H. 1982. Problems with so-called “direct” scal-
Anderson, N. H. 1977. Note on functional measurement and data ing. In: J. T. Kuznicki, R. A. Johnson and A. F. Rutkiewic
analysis. Perception and Psychophysics, 21, 201–215. (eds.), Selected Sensory Methods: Problems and Approaches
ASTM. 2008a. Standard test method for unipolar magnitude to Hedonics. American Society for Testing and Materials,
estimation of sensory attributes. Designation E 1697-05. Philadelphia, pp. 34–48.
In: Annual Book of ASTM Standards, Vol. 15.08, End Borg, G. 1982. A category scale with ratio properties for
Use Products. American Society for Testing and Materials, intermodal and interindividual comparisons. In: H.-G.
Conshohocken, PA, pp. 122–131. Geissler and P. Pextod (Eds.), Psychophysical Judgment
ASTM. 2008b. Standard test method for sensory evaluation of and the Process of Perception. VEB Deutscher Verlag der
red pepper heat. Designation E 1083-00. In: Annual Book of Wissenschaften, Berlin, pp. 25–34.
ASTM Standards, Vol. 15.08, End Use Products. American Borg, G. 1990. Psychophysical scaling with applications in
Society for Testing and Materials, Conshohocken, PA, physical work and the perception of exertion. Scandinavian
pp. 49–53. Journal of Work and Environmental Health, 16, 55–58.
Aust, L. B., Gacula, M. C., Beard, S. A. and Washam, R. W., II. Boring, E. G. 1942. Sensation and Perception in the History
1985. Degree of difference test method in sensory evaluation of Experimental Psychology. Appleton-Century-Crofts,
of heterogeneous product types. Journal of Food Science, 50, New York.
511–513. Brandt, M. A., Skinner, E. Z. and Coleman, J. A. 1963. The tex-
Baird, J. C. and Noma, E. 1978. Fundamentals of Scaling and ture profile method. Journal of Food Science, 28, 404–409.
Psychophysics. Wiley, New York. Butler, G., Poste, L. M., Wolynetz, M. S., Agar, V. E. and
Banks, W. P. and Coleman, M. J. 1981. Two subjective scales of Larmond, E. 1987. Alternative analyses of magnitude esti-
number. Perception and Psychophysics, 29, 95–105. mation data. Journal of Sensory Studies, 2, 243–257.
Bartoshuk, L. M., Snyder, D. J. and Duffy, V. B. 2006. Hedonic Cardello, A. V. and Schutz, H. G. 2004. Research note.
gLMS: Valid comparisons for food liking/disliking across Numerical scale-point locations for constructing the LAM
obesity, age, sex and PROP status. Paper presented at (Labeled affective magnitude) scale. Journal of Sensory
the 2006 Annual Meeting, Association for Chemoreception Studies, 19, 341–346.
Sciences. Cardello, A. V., Lawless, H. T. and Schutz, H. G. 2008. Effects
Bartoshuk, L. M., Duffy, V. B., Fast, K., Green, B. G., Prutkin, J. of extreme anchors and interior label spacing on labeled
and Snyder, D. J. 2003. Labeled scales (e.g. category, Likert, magnitude scales. Food Quality and Preference, 21, 323–334.
VAS) and invalid across-group comparisons: What we have Cardello, A. V., Winterhaler, C. and Schutz, H. G. 2003.
learned from genetic variation in taste. Food Quality and Predicting the handle and comfort of military clothing fabrics
Preference, 14, 125–138. from sensory and instrumental data: Development and appli-
Bartoshuk, L. M., Duffy, V. B., Green, B. G., Hoffman, H. J., cation of new psychophysical methods. Textile Research
Ko, C.-W., Lucchina, L. A., Marks, L. E., Snyder, D. J. and Journal, 73, 221–237.
Weiffenbach, J. M. 2004a. Valid across-group comparisons Cardello, A. V., Schutz, H. G., Lesher, L. L. and Merrill, E.
with labeled scales: the gLMS versus magnitude matching. 2005. Development and testing of a labeled magnitude scale
Physiology and Behavior, 82, 109–114. of perceived satiety. Appetite, 44, 1–13.
Bartoshuk, L. M., Duffy, V. B., Chapo, A. K., Fast, K., Yiee, Caul, J. F. 1957. The profile method of flavor analysis. Advances
J. H., Hoffman, H. J., Ko, C.-W. and Snyder, D. J. 2004b. in Food Research, 7, 1–40.
From psychophysics to the clinic: Missteps and advances. Chambers, E. C. and Wolf, M. B. 1996. Sensory Testing
Food Quality and Preference, 14, 617–632. Methods. ASTM Manual Series, MNL 26. ASTM
Bartoshuk, L. M., Duffy, V. B., Fast, K., Green, B. Kveton, J., International, West Conshohocken, PA.
Lucchina, L. A., Prutkin, J. M., Snyder, D. J. and Tie, K. Chen, A. W., Resurreccion, A. V. A. and Paguio, L. P. 1996. Age
1999. Sensory variability, food preferences and BMI in non- appropriate hedonic scales to measure the food preferences
medium and supertasters of PROP. Appetite, 33, 228–229. of young children. Journal of Sensory Studies, 11, 141–163.
References 175
Chung, S.-J. and Vickers, 2007a. Long-term acceptability and Gent, J. F. and Bartoshuk, L. M. 1983. Sweetness of
choice of teas differing in sweetness. Food Quality and sucrose, neohesperidin dihydrochalcone and sacchar in is
Preference 18, 963–974. related to genetic ability to taste the bitter substance 6-n-
Chung, S.-J. and Vickers, 2007b. Influence of sweetness on the propylthiouracil. Chemical Senses, 7, 265–272.
sensory-specific satiety and long-term acceptability of tea. Gescheider, G. A. 1988. Psychophysical scaling. Annual Review
Food Quality and Preference, 18, 256–267. of Psychology, 39, 169–200.
Coetzee, H. and Taylor, J. R. N. 1996. The use and adaptation of Giovanni, M. E. and Pangborn, R. M. 1983. Measurement of
the paired comparison method in the sensory evaluation of taste intensity and degree of liking of beverages by graphic
hamburger-type patties by illiterate/semi-literate consumers. scaling and magnitude estimation. Journal of Food Science,
Food Quality and Preference, 7, 81–85. 48, 1175–1182.
Collins, A. A. and Gescheider, G. A. 1989. The measurement of Gracely, R. H., McGrath, P. and Dubner, R. 1978a. Ratio
loudness in individual children and adults by absolute magni- scales of sensory and affective verbal-pain descriptors. Pain,
tude estimation and cross modality matching. Journal of the 5, 5–18.
Acoustical Society of America, 85, 2012–2021. Gracely, R. H., McGrath, P. and Dubner, R. 1978b. Validity and
Conner, M. T. and Booth, D. A. 1988. Preferred sweetness of a sensitivity of ratio scales of sensory and affective verbal-pain
lime drink and preference for sweet over non-sweet foods. descriptors: Manipulation of affect by Diazepam. Pain, 5,
Related to sex and reported age and body weight. Appetite, 19–29.
10, 25–35. Green, B. G., Shaffer, G. S. and Gilmore, M. M. 1993.
Cordinnier, S. M. and Delwiche, J. F. 2008. An alternative Derivation and evaluation of a semantic scale of oral sen-
method for assessing liking: Positional relative rating ver- sation magnitude with apparent ratio properties. Chemical
sus the 9-point hedonic scale. Journal of Sensory Studies, 23, Senses, 18, 683–702.
284–292. Green, B. G., Dalton, P., Cowart, B., Shaffer, G., Rankin, K.
Cox, E. P. 1980. The optimal number of response alternatives and Higgins, J. 1996. Evaluating the “Labeled Magnitude
for a scale: A review. Journal of Marketing Research, 18, Scale” for measuring sensations of taste and smell. Chemical
407–422. Senses, 21, 323–334.
Curtis, D. W., Attneave, F. and Harrington, T. L. 1968. A test of Greene, J. L., Bratka, K. J., Drake, M. A. and Sanders, T. H.
a two-stage model of magnitude estimation. Perception and 2006. Effective of category and line scales to characterize
Psychophysics, 3, 25–31. consumer perception of fruity fermented flavors in peanuts.
Edwards, A. L. 1952. The scaling of stimuli by the method Journal of Sensory Studies, 21, 146–154.
of successive intervals. Journal of Applied Psychology, 36, Guest, S., Essick, G., Patel, A., Prajpati, R. and McGlone, F.
118–122. 2007. Labeled magnitude scales for oral sensations of wet-
Ekman, G. 1964. Is the power law a special case of Fechner’s ness, dryness, pleasantness and unpleasantness. Food Quality
law? Perceptual and Motor Skills, 19, 730. and Preference, 18, 342–352.
Einstein, M. A. 1976. Use of linear rating scales for the evalu- Hein, K. A., Jaeger, S. R., Carr, B. T. and Delahunty, C. M.
ation of beer flavor by consumers. Journal of Food Science, 2008. Comparison of five common acceptance and prefer-
41, 383–385. ence methods. Food Quality and Preference, 19, 651–661.
El Dine, A. N. and Olabi, A. 2009. Effect of reference foods in Huskisson, E. C. 1983. Visual analogue scales. In: R. Melzack
repeated acceptability tests: Testing familiar and novel foods (Ed.), Pain Measurement and Assessment. Raven, New York,
using 2 acceptability scales. Journal of Food Science, 74, pp. 34–37.
S97–S105. Jaeger, S. R.; Jørgensen, A. S., AAslying, M. D. and Bredie,
Engen, T. 1974. Method and theory in the study of odor prefer- W. L. P. 2008. Best-worst scaling: An introduction and
ences. In: A. Turk, J. W. Johnson and D. G. Moulton (Eds.), initial comparison with monadic rating for preference elic-
Human Responses to Environmental Odors. Academic, itation with food products. Food Quality and Preference, 19,
New York. 579–588.
Finn, A. and Louviere, J. J. 1992. Determining the appropriate Jaeger, S. R. and Cardello, A. V. 2009. Direct and indirect hedo-
response to evidence of public concern: The case of food nic scaling methods: A comparison of the labeled affective
safety. Journal of Public Policy and Marketing, 11, 12–25. magnitude (LAM) scale and best-worst scaling. Food Quality
Forde, C. G. and Delahunty, C. M. 2004. Understanding the role and Preference, 20, 249–258.
cross-modal sensory interactions play in food acceptability in Jones, F. N. 1974. History of psychophysics and judgment.
younger and older consumers. Food Quality and Preference, In: E. C. Carterette and M. P. Friedman (Eds.), Handbook
15, 715–727. of Perception. Psychophysical Judgment and Measurement,
Frijters, J. E. R., Kooistra, A. and Vereijken, P. F. G. 1980. Vol. 2. Academic, New York, pp. 11–22.
Tables of d’ for the triangular method and the 3-AFC sig- Jones, L. V. and Thurstone, L. L. 1955. The psychophysics of
nal detection procedure. Perception and Psychophysics, 27, semantics: An experimental investigation. Journal of Applied
176–178. Psychology, 39, 31–36.
Gaito, J. 1980. Measurement scales and statistics: Resurgence Jones, L. V., Peryam, D. R. and Thurstone, L. L. 1955.
of an old misconception. Psychological Bulletin, 87, Development of a scale for measuring soldier’s food pref-
564–587. erences. Food Research, 20, 512–520.
Gay, C., and Mead, R. 1992 A statistical appraisal of the prob- Keskitalo, K. Knaapila, A., Kallela, M., Palotie, A., Wessman,
lem of sensory measurement. Journal of Sensory Studies, 7, M., Sammalisto, S., Peltonen, L., Tuorila, H. and Perola,
205–228. M. 2007. Sweet taste preference are partly genetically
176 7 Scaling
determined: Identification of a trait locus on Chromosome Lindvall, T. and Svensson, L. T. 1974. Equal unpleasantness
161–3 . American Journal of Clinical Nutrition, 86, 55–63. matching of malodourous substances in the community.
Kim, K.-O. and O’Mahony, M. 1998. A new approach to cat- Journal of Applied Psychology, 59, 264–269.
egory scales of intensity I: Traditional versus rank-rating. Mahoney, C. H., Stier, H. L. and Crosby, E. A. 1957. Evaluating
Journal of Sensory Studies, 13, 241–249. flavor differences in canned foods. II. Fundamentals of the
King, B. M. 1986. Odor intensity measured by an audio method. simplified procedure. Food Technology 11, Supplemental
Journal of Food Science, 51, 1340–1344. Symposium Proceedings, 37–42.
Koo, T.-Y., Kim, K.-O., and O’Mahony, M. 2002. Effects Marks, L. E. 1978. Binaural summation of the loudness of pure
of forgetting on performance on various intensity scal- tones. Journal of the Acoustical Society of America, 64,
ing protocols: Magnitude estimation and labeled magni- 107–113.
tude scale (Green scale). Journal of Sensory Studies, 17, Marks, L. E., Borg, G. and Ljunggren, G. 1983. Individual dif-
177–192. ferences in perceived exertion assessed by two new methods.
Kroll, B. J. 1990. Evaluating rating scales for sensory testing Perception and Psychophysic, 34, 280–288.
with children. Food Technology, 44(11), 78–80, 82, 84, 86. Marks, L. E., Borg, G. and Westerlund, J. 1992. Differences
Kurtz, D. B., White, T. L. and Hayes, M. 2000. The labeled in taste perception assessed by magnitude matching and by
dissimilarity scale: A metric of perceptual dissimilarity. category-ratio scaling. Chemical Senses, 17, 493–506.
Perception and Psychophysics, 62, 152–161. Mattes, R. D. and Lawless, H. T. 1985. An adjustment error in
Land, D. G. and Shepard, R. 1984. Scaling and ranking methods. optimization of taste intensity. Appetite, 6, 103–114.
In: J. R. Piggott (ed.), Sensory Analysis of Foods. Elsevier McBride, R. L. 1983a. A JND-scale/category scale convergence
Applied Science, London, pp. 141–177. in taste. Perception and Psychophysics, 34, 77–83.
Lane, H. L., Catania, A. C. and Stevens, S. S. 1961. Voice McBride, R. L. 1983b. Taste intensity and the case of expo-
level: Autophonic scale, perceived loudness and effect of nents greater than 1. Australian Journal of Psychology, 35,
side tone. Journal of the Acoustical Society of America, 33, 175–184.
160–167. McBurney, D. H. and Shick, T. R. 1971. Taste and water taste
Larson-Powers, N. and Pangborn, R. M. 1978. Descriptive anal- for 26 compounds in man. Perception and Psychophysics, 10,
ysis of the sensory properties of beverages and gelatins 249–252.
containing sucrose or synthetic sweeteners. Journal of Food McBurney, D. H. and Bartoshuk, L. M. 1973. Interactions
Science, 43, 47–51. between stimuli with different taste qualities. Physiology and
Lawless, H. T. 1977. The pleasantness of mixtures in taste and Behavior, 10, 1101–1106.
olfaction. Sensory Processes, 1, 227–237. McBurney, D. H., Smith, D. V. and Shick, T. R. 1972. Gustatory
Lawless, H. T. 1989. Logarithmic transformation of magnitude cross-adaptation: Sourness and bitterness. Perception and
estimation data and comparisons of scaling methods. Journal Psychophysics, 11, 228–232.
of Sensory Studies, 4, 75–86. Mead, R. and Gay, C. 1995. Sequential design of sensory trials.
Lawless, H. T. and Clark, C. C. 1992. Psychological biases in Food Quality and Preference, 6, 271–280.
time intensity scaling. Food Technology, 46, 81, 84–86, 90. Mecredy, J. M. Sonnemann, J. C. and Lehmann, S. J. 1974.
Lawless, H. T. and Malone, J. G. 1986a. The discriminative Sensory profiling of beer by a modified QDA method. Food
efficiency of common scaling methods. Journal of Sensory Technology, 28, 36–41.
Studies, 1, 85–96. Meilgaard, M., Civille, G. V. and Carr, B. T. 2006.
Lawless, H. T. and Malone, G. J. 1986b. A comparison of scaling Sensory Evaluation Techniques, Fourth Edition. CRC,
methods: Sensitivity, replicates and relative measurement. Boca Raton, FL.
Journal of Sensory Studies, 1, 155–174. Moore, L. J. and Shoemaker, C. F. 1981. Sensory textural prop-
Lawless, H. T. and Skinner, E. Z. 1979. The duration erties of stabilized ice cream. Journal of Food Science, 46,
and perceived intensity of sucrose taste. Perception and 399–402.
Psychophysics, 25, 249–258. Moskowitz, H. R. 1971. The sweetness and pleasant-
Lawless, H. T., Popper, R. and Kroll, B. J. 2010a. Comparison ness of sugars. American Journal of Psychology, 84,
of the labeled affective magnitude (LAM) scale, an 11-point 387–405.
category scale and the traditional nine-point hedonic scale. Moskowitz, H. R. and Sidel, J. L. 1971. Magnitude and hedo-
Food Quality and Preference, 21, 4–12. nic scales of food acceptability. Journal of Food Science, 36,
Lawless, H. T., Sinopoli, D. and Chapman, K. W. 2010b. A com- 677–680.
parison of the labeled affective magnitude scale and the nine Muñoz, A. M. and Civille, G. V. 1998. Universal, product and
point hedonic scale and examination of categorical behavior. attribute specific scaling and the development of common
Journal of Sensory Studies, 25, S1, 54–66. lexicons in descriptive analysis. Journal of Sensory Studies,
Lawless, H. T., Cardello, A. V., Chapman, K. W., Lesher, L. L., 13, 57–75.
Given, Z. and Schutz, H. G. 2010c. A comparison of the Newell, G. J. and MacFarlane, J. D. 1987. Expanded tables for
effectiveness of hedonic scales and end-anchor compression multiple comparison procedures in the analysis of ranked
effects. Journal of Sensory Studies, 28, S1, 18–34. data. Journal of Food Science, 52, 1721–1725.
Lee, H.-J., Kim, K.-O., and O’Mahony, M. 2001. Effects of for- Olabi, A. and Lawless, H. T. 2008. Persistence of context
getting on various protocols for category and line scales of effects with training and reference standards. Journal of Food
intensity. Journal of Sensory Studies, 327–342. Science, 73, S185–S189.
Likert, R. 1932. Technique for the measurement of attitudes. O’Mahony, M., Park, H., Park, J. Y. and Kim, K.-O. 2004.
Archives of Psychology, 140, 1–55. Comparison of the statistical analysis of hedonic data using
References 177
analysis of variance and multiple comparisons versus and Stevens, S. S. 1956. The direct estimation of sensory
R-index analysis of the ranked data. Journal of Sensory magnitudes—loudness. American Journal of Psychology, 69,
Studies, 19, 519–529. 1–25.
Pangborn, R. M. and Dunkley, W. L. 1964. Laboratory proce- Stevens, S. S. 1957. On the psychophysical law. Psychological
dures for evaluating the sensory properties of milk. Dairy Review, 64, 153–181.
Science Abstracts, 26–55–62. Stevens, S. S. 1969. On predicting exponents for cross-modality
Parducci, A. 1965. Category judgment: A range-frequency matches. Perception and Psychophysics, 6, 251–256.
model. Psychological Review, 72, 407–418. Stevens, S. S. and Galanter, E. H. 1957. Ratio scales and cat-
Park, J.-Y., Jeon, S.-Y., O’Mahony, M. and Kim, K.-O. 2004. egory scales for a dozen perceptual continua. Journal of
Induction of scaling errors. Journal of Sensory Studies, 19, Experimental Psychology, 54, 377–411.
261–271. Stoer, N. L. and Lawless, H. T. 1993. Comparison of single
Pearce, J. H., Korth, B. and Warren, C. B. 1986. Evaluation product scaling and relative-to-reference scaling in sensory
of three scaling methods for hedonics. Journal of Sensory evaluation of dairy products. Journal of Sensory Studies, 8,
Studies, 1, 27–46. 257–270.
Peryam. D. 1989. Reflections. In: Sensory Evaluation. In Stone, H., Sidel, J., Oliver, S., Woolsey, A. and Singleton,
Celebration of our Beginnings. American Society for Testing R. C. 1974. Sensory Evaluation by quantitative descriptive
and Materials, Philadelphia, pp. 21–30. analysis. Food Technology, 28, 24–29, 32, 34.
Peryam, D. R. and Girardot, N. F. 1952. Advanced taste-test Teghtsoonian, M. 1980. Children’s scales of length and
method. Food Engineering, 24, 58–61, 194. loudness: A developmental application of cross-modal
Piggot, J. R. and Harper, R. 1975. Ratio scales and category matching. Journal of Experimental Child Psychology, 30,
scales for odour intensity. Chemical Senses and Flavour, 1, 290–307.
307–316. Thurstone, L. L. 1927. A law of comparative judgment.
Pokorńy, J., Davídek, J., Prnka, V. and Davídková, E. 1986. Psychological Review, 34, 273–286.
Nonparametric evaluation of graphical sensory profiles for Townsend, J. T. and Ashby, F. G. 1980. Measurement scales and
the analysis of carbonated beverages. Die Nahrung, 30, statistics: The misconception misconceived. Psychological
131–139. Bulletin, 96, 394–401.
Poulton, E. C. 1989. Bias in Quantifying Judgments. Lawrence Vickers, Z. M. 1983. Magnitude estimation vs. category scal-
Erlbaum, Hillsdale, NJ. ing of the hedonic quality of food sounds. Journal of Food
Richardson, L. F. and Ross, J. S. 1930. Loudness and telephone Science, 48, 1183–1186.
current. Journal of General Psychology, 3, 288–306. Villanueva, N. D. M. and Da Silva, M. A. A. P. 2009.
Rosenthal, R. 1987. Judgment Studies: Design, Analysis and Performance of the nine-point hedonic, hybrid and self-
Meta-Analysis. University Press, Cambridge. adjusting scales in the generation of internal preference
Shand, P. J., Hawrysh, Z. J., Hardin, R. T. and Jeremiah, L. E. maps. Food Quality and Preference, 20, 1–12.
1985. Descriptive sensory analysis of beef steaks by cate- Villanueva, N. D. M., Petenate, A. J., and Da Silva, M. A. A.
gory scaling, line scaling and magnitude estimation. Journal P. 2005. Comparative performance of the hybrid hedonic
of Food Science, 50, 495–499. scale as compared to the traditional hedonic, self-adjusting
Schutz, H. G. and Cardello, A. V. 2001.. A labeled affective and ranking scales. Food Quality and Preference, 16,
magnitude (LAM) scale for assessing food liking/disliking. 691–703.
Journal of Sensory Studies, 16, 117–159. Ward, L. M. 1986. Mixed-modality psychophysical scaling:
Siegel, S. 1956. Nonparametric Statistics for the Behavioral Double cross-modality matching for “difficult” continua.
Sciences. McGraw-Hill, New York. Perception and Psychophysics, 39, 407–417.
Sriwatanakul, K., Kelvie, W., Lasagna, L., Calimlim, J. F., Wels, Weiss, D. J. 1972. Averaging: an empirical validity criterion
O. F. and Mehta, G. 1983. Studies with different types for magnitude estimation. Perception and Psychophysics, 12,
of visual analog scales for measurement of pain. Clinical 385–388.
Pharmacology and Therapeutics, 34, 234–239. Winakor, G., Kim, C. J. and Wolins, L. 1980. Fabric hand:
Stevens, J. C. and Marks, L. M. 1980. Cross-modality matching Tactile sensory assessment. Textile Research Journal, 50,
functions generated by magnitude estimation. Perception and 601–610.
Psychophysics, 27, 379–389. Yamaguchi, S. 1967. The synergistic effect of monosodium glu-
Stevens, S. S. 1951. Mathematics, measurement and psy- tamate and disodium 5 inosinate. Journal of Food Science
chophysics. In: S. S. Stevens (ed.), Handbook of 32, 473–477.
Experimental Psychology. Wiley, New York, pp. 1–49.
012345678
95
634 52

!"#!"$! !%!&'(!)%"#!"*!&'+&

!&%"+!*'(%&,-! !&#(!(!%)'&.(.!&#%&! )%"/%&
0!''!"!1("!/!&)%""%"!!&%"+"$(!!"/,)%''%0!#$+
%"'"!*!0%)!2"#!"*!&'+!&1(!3!4'*%"5"%2'!,-!
!&#!"$!!!6("!5"%2'!30!''"%"!"+#!"*!/!%#(
7(&*!8!"*!9&'+&#!:!"(//!%#,-!!&'!#!"!#!"
"%(.!;$+;!'%&%)%&!&(&#$''%;"&!#.!&!"#!"*!
&'+!,-!!&.'.&##(%/!%)!(#!%/"&.%&*!&%&'
#!"*!&'+!&1(!,)%''%0!#$+&&;#!#(%&%)!*";
%&%& ! !/!%)#!"*!&'+()"!!%!"%2'&.&#<
"%2'&.,
=>?@AABCD?EFAF?AGA?ADBHEB@ID@G?AJB@BHGD@G?AJB@G>FJEFEB@GAJAKADG?LJEAKCDM
NO!&"P!
0Q454
78746Q4Q
RSTR UVWXYZWV ,,,,,,,,,,,,,,,, [[\ 8!"*!!&%"+&'+!"! !/% %;
RST[ ]^W_`^ZaZb^Vcd^ ,,,,,,,, [[e !# %%'& !"!&'%) !!&%"+!&,
RSTf gVhYh^VX`^ZaZb^VcdZ ,,,,, [[e !!!&1(!''%0 !!&%"+!& %%$&
RSTi `^ZaZb^VcdZjqp^kVZlY^,,,,,,, [fR
mn,o,m 4'*%"5"%2'! ,,,,,,,,,,, [fR %/'!!!&%"+#!"%&%)"%#(3%#!&;
mn,o,r 7(&*!8!qp"*!9&'+qp ,, [fi )+(&#!"'+&.&."!#!&&#"%!*"$'!3&#%"
mn,o,s !6("!5"%2'! qp,,,,,,,,,, [f\ %#!!"/&!0!&%"+"$(!"!/%"&
mn,o,o :!&%"+:!"(/ ,,,,,,,,, [fe %!&!,9.!&!"#!"*!&'+0%('#
RSTt u^V^Z`^ZaZb^VcdZ ,,,,,,,, [iS
mn,v,m O%0 %8%8!"*!9&'+&"!!
w+:! ,,,,,,,,,,,,, [iS
((''+*!$!0!!&&#mr&!' 0%('#
mn,v,r :(#!x%/"&.8))!"!& *!$!!& "&!#30 !(!%)"!)!"!&!&;
x%&*!&%&'8!"*!9&'+ #"#3%(&#!"&#&#."!!%& !/!&&.%)
!&1(!,,,,,,,,,,,,, [iy !"$(!(!#,!+0%('#((''+(!1(&;
RSTy zZZWVWVk^jk^{^ ,,,,,,,,,, [i\ *!'!)%"&!&+0''%0 !# %
mn,|,m }&.9 "$(!x%&4"!1(!&! $!''+&'+!#,!!&!'0%('#&%
~&!#%)9 "$(!~&!&! ,,,, [i\ $!!#)%"!"!#%&"!%&!% !"%#(,
mn,|,r 8!*%&)"%/p!)!"!&!P!%# ,, [ie O%0!*!"30!0''!!& !"3!"!"!
mn,|,s ~&!&+"%&8!"*!
P!%# ,,,,,,,,,,,,,, [i !*!"'#))!"!& #!"*!&'+/!%#&#3&
mn,|,o x%/$&%&%)8!"*!9&'+ .!&!"'3!!"!<! *!"+#))!"!& !&%"+'%%;
&#/!;p!'!#~&!&+P!%# , [i !&#"%!,}(''+3#!"*!!&1(!
mn,|,v 4"!!x%!5"%2'&. ,,,,,,,, [i "%#(!%$!*!#!"%&%)"%#(& !"/
mn,|,| 4'5"%2'&. ,,,,,,,,,,, [t[
^_^^V^ ,,,,,,,,,,,,,,,,,,,, [tf %) !!"!*!#!&%"+"$(!,8!!&#&.%&
O,,0'!3O,O!+/&&3D@GBC?K?AJB@BHBBI34%%#:!&!!6:!"!3
8~mn,mnn;m;oom;|o;vmn3:"&.!":!&!(&!P!#3xrnmn
001 23 456789
95 696
!" C!@! A# &!!$?
# $ "%&!'$$!!$!& $!AA !!$
!!&( !&!$! '$!B
@!@A !@? AB
(U?$@!!$@!&!@A
)*+,-./.012/.3456758/9:;<=./. ?!!"$!?"'$!@!@!B
(VE#&'! AW%!#XA
>&!!$?!@!$$?A$!?B # YZ[\]^' !!$!@!@"_!B
!'!!$! A? '!?''&' $(`
!" A!@$ !#! A A !?$@' !?!
?AA!# @&!$( ! A!A@$"
$ ?#"?A#!
C! A # #B #(X &#? #$!A#&B
D AA@(>&!!$?!! $?! # &?
E!$?'?##D #!$! @!a!(CA
$ &!
AAA#?(C! !! A# !@!$!

?
(C
!$A$AB$A@!$$?A%@' A#!$!"$Y@&$!@!@ B
'$$!!! &#(>& Â!$!##!(
!A$?&$B b A#! EB
# #!'$!' (CAA$ $#!?A#
!@ !!"$? A?B AAA# !#!!(
(F!$?!!&! b !A#"?!&$&@"
"&!$!"$'?! A!"$# !"!!@!$a!YGca!b&$$
#"(>&" Z[[d^(V#" A!&'!
E&A!?BB!?!$?$!" A#!#!?E#!?#$!B
#!$A$'"$ @#!% !$?A$! !$@!
##$!(G &#! !# @!@ A$YF!`DG!?
" ?H#!$$!( Z[[Z^(V@$E!#$#!??A
>&!!$?$&" Y!$$?!$$!?B
'#"!!$$&# ^"!$?!$$'!$
!$$"!!&?$! " @!$a!"!$!'"B
!"$( !!(C @!$a!$!@'!e$?
Y!!#^!$
)*+IJ;:KL;K/;:M2/.3456758/9:;<=.5. YP#!! $"E #$$A!!
b$Z[df^(
F!#!!''!!B
C!? A$!@!@!#$?&?!? &!$'#"!A$ A!$!#!@B
$!@!@$E!$$!@!@!$!@!@( A$A#!"?E@!$ !
N&?!?$!@!@!$?&!! #!?!!!A!"$(P '&A
#!?&!?!$!$"@!@@!!$ "!!&?$!!!'YAE!#$
@(OE!$$!@!@$!@!@A '^!@$!!#!?"!!(b
!?!$!@!@#!?"&?!? A#!#&''B
&!(P '&A'$#!$?$EB !@??(QE!#$R$# !$(
!$$!@!@&!(Q# A$E!$ YZ[[[^' e@' !@%A!!$
$!@!@" '#( &@e !@%!#$!A
R$!@!@!$$?!A !!'# #!!# @
!#!!$$?&?$? !!$&@!@$A!B
(CA$?S%!@T!!' !A!!" A!!
!$( !"(F!#!#$$A
0123 4567857956
9 965
!"#$%&##''#("#)# I"###%$#*/#%)(3(#%9#"#
)#'(*#+,-.!#+/01123%("#'4 #(!%9#(')(=#*7##)"##&#
####5#("#)(#$(!( #%9#"#%%('#%)(%*%/F%&%#
"&#(#()!%(#)'(*%(+6%%*)(4 M=#-2>NO3+C"##%9#"#%%%4
((#""##('#'###(#( #9!F%&%#M=#("## %#
##%!#%*%#(#''#"##(! %(#('))(5%*#%*)(#%C9#21+2+
%#%"(7##)"%*%+ I#=% (%##"('"##"#%%%+
6')#%#(##5#(!#%)(8 C"#####%)("(%%*%#*($
#%)%("#!*9#%#+:%#)4 "#*)#D"##'(#-"#!"(%%#)##%&#
#%'#;#!#%<#"%="#"#!#7#( %''###*($"#*)#+C"-%'=###&4
#&#)((%9#'(="%""#!"&#( %$9#!P%#*)# "#*)##
#(#)+:*($#+/2>>?3;(#(#(' "##5*#"#('#"#A#((%#%!B
"#%%#)#%@A6=("#=#(= =((9##'#%)(+Q"#("#"-%'
#%#(="#%=(#%)%( "##((('"#9#!P%#*)#%''#-#
*#+BC"#$( %'( )#%(#"#*# ()(#%$(%%('(#5*)#-"#A#((
(#)(9#9#((**%#)#%#!=%" %#%!B=(9#))()%##%)(+
(#("#D%("#=("#%%$)(###
A'*#('#'###B'("#)#$()/.! R9S#T#U="
VWXYW Z#%9#"#%%""(9##*#*4
#"((%$#*'(#%)%&#!%%#/%
#+-01123+C"-*()%(%*)4 (#('%*)
%(-#%)%&#!%#;%#)#%#)#%E Z%%*%#(#3 .(#%*)(
(#)%#%#'!"(#%#%E4 [(4#
$$#+C"# $$##9!(*#(#%9# \##((*##)#J#P#%(
#(!"#%%%*(=!((%*)#%# \##(%*#()"!%
(4)#%E((="##(!)#%%(*#4 H%*# #*#
##"##!%$(#)%=! ]#%# $
#%9#
"=% )(&%#*#%$'+
F(#)'(*%(#E%%(9#%# G*9%$(
F(
# *#%$
'((=+6"#G%#H#*(I## \#'####!((9%
(
(%#(#&#!!(#)('((#&#!%*4 F(**%#
%-9##=##$""%#((%# \##(#%! M#%*)(
#% 9#=%"#%%*%+6("#=(-%'
"%!""##&#('(7###9#-"#
"%=% 9#(""##&##$##+6'"#"% C"# "(# #* "( 9# (*)##! (4
)#%%*%*%$"#(("#"#"%=( #=%"("##*D#5*)#('#4
9###'(&%%(J(("#)(9#*+F((% !%="#"#)#%##&%$#7
"=#4#(#)'(*(%%&% "#!#7#(#9(""#)##%&###4
)(##=%#!#((%#%E$$#'( #"#($"#('"#*#/\''#)#$#
%#%)%(+K(=#&#-=%"("##(!%9# #+-2>Ô3%#"#!9("%%#"#*#(4
"L&("%%(#+6(#=##! #)%*#+6=(9#*"9##(#%#"
#%9#"#L&((''((%)#%##*+I#4 #%"#"##*A($"#B("##*A###B
!!"%$ %7#A"#'#"!97#9#*# "(9##%"##&%(('"#*#*4
$((B(A"#($"!)#9+BC"###" )#+:%%(!"##*"(9#("($(+
=%"((=%"(# 9#/'(#5*)#- Q"($( #%)(#((##=%"#"
"#.# ,((7('F((39'(#-((- ("#+[(4("($( #%)((&#)D'(#5*4
#5#"##%(A.# ,((7B"="# )#-7%$)# ((#"#A#'%%#%!B
=#=((##"("##(#)=###( (']%((%=%#((#A"#!%#%!B
)#%#!#E#/)#'#9!=%"#'###3 =(9#7%$"#*((#(4("($(#*+
"#%#%E $$##(#%9#"##(! 6%&#!('%$-#4*(%&%$-*#!'4
#%((%#=%""#)(%#+ %$("#)#%="#"#!#7#((#
012 32 456789
95 696
!"#"" ""=%+"2"'"
"$"%"&""'& 2"'"'O)&B"%&$&L"
()$#" 5$&L""$"=%+"
&'"%""**'+"'# 2""""$''$*%$$
,&-.//01"$+%% $5*;%&"''"<%&,"
''2"34+5$2 =%-.//B1'"%$'"J'"
'+"'2'"%"** '""'$$%%&$"%
"6#"&""%'' *
* $"'"$'$&"2 :'"$%"'%"
2&)"%" "*+=%""""'
*+$ 7#""$"%2" ''&#*#'#*+&'*"
2'""$''&#& $" C&%$$$""
&%**$'8 +""%"&&"5
9"*+$''"+"% %%32""#&
2"''2"':" "'#$"&$ ()$#%*
""""*'&$$";' '""+*'"$'$""
<-="+">2#./?@A>2 *'"*#+"'"&#*&*#
="#.//B1C)&"$"* '4+-(PQPI#ORRN#*
D&'$"**'%2 )'++"21E&"++"'"2"
#"'' "3&%"&"$"*
"&*% !* *)"E#
%-' '#*"%1 '""'%""*'""
*%-E"' -,%#.//01#$"%"
"3%#*";''%<" +'$*'""
F"G"1H)$"$"$ "*5*#2"*%"'#"
'2$'#2" '%""*'$"%""&#$""#
$22& '#'3*'"$2*
++& 5$%2"'$'"$"3'&
5*'"$"+&* "&'"%9'+$
"$'"'$ 322J)"*""'"
$5*#""+&$*"*'"$ $%2""'"$:&'+"
"'"$"+&"'% '"&S5$"'"LS5SC"$"%
''$'''$I' "2"""&%&$"A
$'C&$J"*' &$"L"'3''$
$"%&,"=%-.//B1 '2"'*+&"$&#
';%&"''"<-'"'1 ;"$<&"5""'&
"*$K'*& ***"*'"$"+&"$
"$"#&;%%&+"<C"' !"%'"$'%2"
&"'"+%$ $'"""%""&%&$" 9"
*""'''$'I'"* *"&"&"*$'"
'$'"'"'' J#%%"5&
E#"'"$'% %%$&$"')
&"'-"*""321*$' $'"$&
()$#&'"2") %"*'"$F"*'
$J""'$"'"$-!L'L"3 ""*&%""'
#./@B1C"$"% C"$"J"**$"*$&
'"'*4+'$ "'*'*'"$&
*"$' ()$#,& %" ,2+#"*J'""%""$&"'
M%-./?N1;%$$$<'"% *'$+$
0123 456789
95 69657956 0
!""#$%"!&$"!'!() $"'"-!!(#""!""-!'*!
*)+ 3!'!#!!'"'#)-!'"+0!""#%
,-"'"-!""%'(-$$%!-! !# '".->&%'>*'&#
(%'"%'./&+0*'"1 '%--&)+
$""%'.%'"'.%""#
!#$!.!'!"%')') NOPQRSTUVWXYWZS[\]^_TWT`SUa\WbcST
'"-!!('!'+0!""#% !#
$ '.%"''!!)*!'2
%-!#!-.'3!"!& !%+ 0#*!&"-!**! (!* $/
,#-"''"""-!'*!-'"-!2 -"'!"!"#'"-!!()"!"
(*2#'%"4!&%')!& -!H%"+d(!*"-.#%'!?$!+
--"#!""'-"5"!6 8ABeC<1f*"8ABgg<1h!"!8ABBA<1i)$
'(!-"#%""#!#$!.!'#$ +8ABBj<1J%)+89::A<1k'k!'
"%')+7!.%+89::;<"""'-!-%!) 89::E<1'J!&'+89::e<+?''!!)1
"%'.(!''!'=!&"")'"-!"1# J%lM'm!(!8ABBg<-)3!'"$
3$15!")6"%'.'='"5*!- #!"!- '!##-"*! "-
$>"!"6.%"5*!-'%-""%' !!&'"-%"&$&'!##
*!!".!+6,"%"""""#2 -!H%"+
-"''"*%'!-"%!!)#"
'=!!"''=!!""%'.3-"%."!2 nopqpnrstuvwxwvysz|{
%"#'='$"+?3$*%'.5,!"
!-#$!"()%&6'"%."!%!&'=2
!!#5%&6*%'")5,!"!-#$ 0!"!&! !-!K(f=8Kf<!"
!"()'!#=-%-*+6 H%!!('"-!!("+,$'-2
@"8ABBC<*>!&*!*!""*' !H%*'$>'?%L+}!'m+1
"$ #- "''"1#3$1 m$.!'&1J?+,!"-!H%*"'('!
)-!"%!#-!'!* ABE:"')ABC:"?%L+}!.)}~
!-#.!'%.."#%..)"$1* k/"$1k)m!-""1'm%+Kf*"
'!"!-!(' &-&#!""&' ="%"''"-!.-$3(")"$"$2
'+D-$!-"*"%--""#% " "%!&##-#$"'!%$&%$(
#-"''"1#3$13 (>' -!+D()"Kf*"-!%)='+
&"")''"-!#$ .%9EF #!"2 ,"("!#Kf!">*"f=?!.%
!""'%-$'"-!#$EAF# ?)"!"8m!-""'k/~"$1ABC:m%1ABC;1
!""*!"'!(!''.*-)1!("1 ABe;i1ABCgJ!&'+19::eJ">*!M1
.--"$>1'''%-+G#!2 ABggJ%)+19::Af*"1ABggk/~"~$1
!!"--%"!()#H%)*!"!& ABCE<+
-$!--$%'"+ K(=!&!"-""%"-!H%+,
,%"#$%!#-"''"#"!2 (-.%)%"''"-!.'%-''2
&---"!&'%"#-- %-(%!!"#!"-!('.)-!&&2
80"!!'DIJ)1ABBA<+?''!!)1 '2 $$& $$."+,Kf-"!'"
"*!.'"")#-."#-!!2 ( ('!'!(!'% '-.(
!&+K3$1!""""'# -$"##'")"$+,='"-!."
!#.!$'"$)!-%''"-!""%-" ( ('("'"!$"
$'1-)1-%&'"1?$1'L!" !"!)#"'"-!"'$!%'
"!"+? #"'"-!"#%')2 8( !$""!<+,-!H%(!'".%2
!&.M')'--! "'%-"+0 !#-!('("1!!"!!"1!
"%')!""$)"'2 '#-!1!#""1'!(2
%-$!'"$#-'.'1!1' !"'!+ !$""!8$!%'<+0#!""!'
,3!-' '*! !M " !)!".%!!"'%-!.+
010 23 456789
95 696
!"##!$#% !6(#2F!(1#'$&*1$/
!$&'!$"!!() /##$'!/1#'H&*1/##
!#'*+,&!!"#! /!!"!11#'H
!*-%!.%/0#!123"!!1% &*1/##/!!"!1
#'''#&!!"! &''1 1#'2
!1$,&!$1!"'$"23#' ''*$FE#'11#'6"!
(#!!/!"#!/ *!!&',&!!!"
"!!!*24"(#!#'&/ #!2G!/&$1#'1*6#''
+#! 25" 6'#!+'23"!''!/' !
+!"!#!'! 1#'>
#2!"##!#"%
1#!&#!!"! 5 ?(#'!
#!24!1#'!!"# =8 I*'!/
#'&+"1!"""! 9 J!/
(#!"#! 2 - C 1
31#'&!#'1 . G
"!1*/!'6&!!123$
"#'/**#+''+'' 3#''&!#!+'"!1
/!'6&!#'#2 #!!"#'2DFE!
7''*$!"#&,&! #!!"&!$6!%
!/!"!''!/'8' !66*!%&'!!"
6'*6(#9/ #!9:#! #!6*#'#''23+'
' !"!/$;-<$!#'1 #!#!6*!"*1%
*16!'=> 6!'246'$!16!!"
1'!*16!'!166*
5 ?(#'! #'!#!''*1"'#$/
@ A!# #&6'836'9@2-=!#$
=8 3!'!)!6' K3
62L
K6$L/! '*$/
9 B' # '#!!"FE'8M!
- C!
. B! B)N!1$9OP@=241%'%
!'*''
3!/,&!!#& !*!"6&'2
'!!6'#!+'23" 3!/&!K1L
+!!/!,&!1#!'"! "!11#'!6*!8"!1'"!
#'"/''!/23#'"% =!#2Q'*16!'6
91"#!/''!/2 !6#!$1#!6'!'*
31#'!"6'6'% !6/*6*!&!'%
!",&!2D!##!!6& '#!23"!$FE'+
!"!&'''*!"#!!#% '&#&2
#!!'#'<!#!! Q!!!"1''$6/
#!2E!#!!"FE1&* 9:#!$8C!0!/$9ORR=$F'&!E!+'
"+'"!!&#'!&!! /1E!+'464'*8E44=2
#!"!1 !#!"1#'2G!/&$ S&"!1E441*6''*'*
#'! !"1/ 6'!#!6'!&FE%*#!%
(#!!FE1!#!% #!23 !"1''''!/
!*231#'+!&'' 1#'!*1!! %
1#!!"6'6'!"#!2D '!"'#!2E44
$1#'!!6 !!$) 1!&FE8G!'2$9OR.=2
0123 456789
95 69657956
!"#$%&'()*+',(-.'-./.0'./&*()#1#2(0%0(3
4&'.*/5678($%(.9*'1#2(0%0(4&')(0*/0:'6%#**9'.;9*+<7=>
$#9-*#9-*+#*#*0#9-'5?C%#-'&%0(5/,'.0'./&*.
#55'5%+(.%+#*' 0#%95&67C0(%'0*0#9-9-N9.,09*9,#&;+'-/.9-N*+'.'
?&#2(0 @**09A/*'. *',+-9\/'../,,'..)/&&67
B-*'-.9*6# P0/'?C0'.9.*.$(.*#**'$%*.)(0$#*+'$#*9,#&
C0(*'9- DE ,+#0#,*'09]#*9(-()*+'5#*#7^./#&&6#.'09'.().6$3
F'#*95'-*9*6 G A(&.$/.*A'9-*'0%0'*'5/.9-N9-*/9*9(-#-5'"%'09'-,'
H'0/$6 G (-*+'%#0*()*+'0'.'#0,+'07C@@U(-*+'(*+'0+#-5U
I%#/.'J ,#-A'#-#&6]'5/.9-N%#0#$'*09,*',+-9\/'../,+#.
F'*#&&9,K#0($#*9,#-5)''&L GM #-#&6.9.()2#09#-,'#-5./9*#A&'$'#-..'%#0#*9(-
K8#009'.*+0(/N+L
C(/&*06
GE
GM
%0(,'5/0'.78/00'-*&6U*+'?C*',+-9\/'9./.'5'"*'-3
O0(*+6 GE .92'&69-*+''2#&/#*9(-();#*'0U%0(A#A&6A',#/.'
I&#NJ ;#*'0/*9&9*9'././#&&6(-&6+#2'*+0''*()(/0%'(%&'*(
P/0:'6 G *0(/A&'.+((**#.*'#-5(5(0,($%&#9-*.K@__@UGRRV`
Q0N#-$'#* GE O#0*'&.'*#&7UGRYaUGRYb`c$d03c]A':#-5[9'*09,+U
F'*#&&9,K#0($#*9,#-5)''&L G D<<YL7
O9**'0 LK #
@)*'0*#.*' B-*'-.9*6
F'*#&&9,)''& DEefghijihklmnopqrsht
C(/&*0A6 GE
Q*+'0 ?&#2(0C0(4&'u/5N'..+(/&5A'.,0''-'5)(0&(-N3
P/0:'6 LKM
*'0$#2#9&#A9&9*67B**#:'.*9$'U'))(0*U#-5$(-'6*(
Q0N#-$'#* LKM
@5#%*'5)0($8+#$A'0.'*#&7KGRRDL *0#9-#%#-'&#-5*+'%#-'&9.*..+(/&5$#:'#,($3
#H,#&'SLKT*+0'.+(&5UGT.&9N+*UDT$(5'0#*'UVT.*0(-N$9*$'-**(A'#2#9&#A&')(06'#0.U9)%(..9A&'7B*9.
-(*/-/./#&*(4-5?C%#-'&9.*.;+(+#2'.'02'5
AWQ*+'0X,+#0#,*'09.*9,.9-*+'#)*'0*#.*';'0'-(*)(/-5A6*+'
'-*90'%#-'& (-*+'.#$'%#-'&)(0$(0'*+#-G<6'#0.7C(*'-*9#&
%#-'&9.*..+(/&5+#2'#:''-9-*'0'.*9-*+'%0(5/,*
H6#09')#-5,(;(0:'0.KGRYZL,($%#0'51#2(0%0(4&' ,#*'N(06#-59*9.+'&%)/&9)*+'6+#2'.($'A#,:3
0'./&*.5'092'5*+0(/N+,(-.'-./.;9*+1#2(0%0(4&' N0(/-5:-(;&'5N'(-*+'%0(5/,**6%'7P+'.'%#-'&9.*.
0'./&*.5'092'5A6,#&,/&#*9-N$'#-.,(0'.7P+'$'#- .+(/&5A'.,0''-'5*(+#2'-(0$#&(5(0#-5*#.*'
.,(0'0'./&*.+#5#.$#&&'0,(')4,9'-*()2#09#*9(- %'0,'%*9(-.7C#-'&9.*.#0'.,0''-'5)(0-(0$#&#,/3
*+#-*+',(-.'-./.0'./&*.#-5*+'%09-,9%#&,($%(3 9*6/.9-N.(&/*9(-.#-5%/0'59&/*'5(5(0#-*.K.''
-'-*#-#&6.9.KC8@L()*+'$'#-.,(0'5#*##,,(/-*'5 8+#%*'0DL7P+'6.+(/&5A'2'06#0*9,/&#*'#-5.9-,'0'
)(0#+9N+'0%0(%(0*9(-()*+'2#09#-,'*+#-*+'C8@ ;9*+#-#%%0(%09#*'%'0.(-#&9*6K-(**9$95(0(2'0&6
()*+',(-.'-./..,(0'.7O#.'5(-*+'.',09*'09#*+' #NN0'..92'L7
#/*+(0.,(-,&/5'5*+#**+'/.'()$'#-.,(0'.N#2' P+'%#-'&&'#5'09.#-#,*92'%#0*9,9%#-*9-A(*+
./%'09(00'./&*.*(*+#*(),(-.'-./..,(0'.7['.%9*' *+'&#-N/#N'5'2'&(%$'-*#-5'2#&/#*9(-%+#.'.()
*+'.'0'./&*.U.($'%0#,*9*9(-'0..*9&&/.'A(*+*+'?C *+'.*/567P+'%#-'&&'#5'0$/.*$(5'0#*'*+'9-*'03
#-5C@@#.#,(-.'-./.*',+-9\/'7 #,*9(-.A'*;''-%#-'&9.*.U&'#59-N*+''-*90'N0(/%
C0(%(-'-*.()?C.*#*'*+#**+'5#*##0'#,,/0#*' *(;#05.($'/-#-9$9*6()(%9-9(-7B*9.,&'#0*+#**+'
#-50'%0(5/,9A&'9)*+'%#-'&9.*.#0';'&&*0#9-'57P+' :'6'&'$'-*9-#?C%#-'&9.*+'%#-'&&'#5'07P+9.
-','..9*6)(02(,#A/&#06.*#-5#059]#*9(-#$(-N%#-3 %'0.(-,((059-#*'.*+'.#$%&'%0(5/,*9(-U590',*.*+'
'&9.*.,#--(*A'(2'0'.*9$#*'57['*0#,*(0.()*+'.' %#-'&'2#&/#*9(-.U#-54-#&&62'0A#&9]'.*+',(-.'-./.
%0(,'5/0'.,($%&#9-*+#**+'5'092'5,(-.'-./.$#6 ,(-,&/.9(-.()*+''-*90'%#-'&7P+'%#-'&&'#5'0;9&&
#,*/#&&6A'*+'(%9-9(-()*+'$(.*5($9-#-*%'03 ()*'-0'./A$9*.#$%&'./-*9&0'%0(5/,9A&'0'./&*.#0'
.(-#&9*69-*+'N0(/%(0*+'%#-'&$'$A'0%'03 (A*#9-'57P+'0')(0'U*+'%#-'&&'#5'0.+(/&5A''.%'3
,'92'5*(+#2'*+'N0'#*'.*#/*+(09*6U()*'-*+'%#-'& ,9#&&6#0*9,/&#*'#-5:-(;&'5N'#A&'#A(/**+'%0(5/,*
&'#5'07@52(,#*'.()*+'*',+-9\/'.,(/-*'0*+#*;9*+ *6%'7P+9.%'0.(-;9&&#&.(A'0'.%(-.9A&')(0,($3
%0(%'0*0#9-9-N*+'%#-'&&'#5'0;9&&%0'2'-**+9.)0($ $/-9,#*9(-;9*+*+'%#-'&#-5%0'%#0#*9(-().#$%&'.
(,,/009-N7@559*9(-#&&6U,+#$%9(-.()*+'$'*+(5 #-50')'0'-,'.*#-5#05.7P+'%#-'&&'#5'0.+(/&5#&.(
012 34 56789

67
7
!""# PLLIG+S"""*')$')'"$
$%$&'('#!)$ ""$X*('!*)$'*+
*"'($'"*+ e$')X*'('!*H
''('#"+U$&$"'
!*"!"*W*")"H
,-./.012345653BA567898:;<6=5678 ($''('#'!$($'
>43?@:6: +Y!*('('##
'('#""*')$!*$($*"+
N)'))') H
C*!D'!EFCDEG&"!H *!*'+D*')$'#)$
""*')$IJKL ''#($'!" "'fghi('"')"*
'#"&$$M!'N'H "*)#'*$'('"'"
FO"O"PLLQRO+IJKQG+S "'"*#'X*'"$+U$
'MN"NEE$"')'" "'"'$'"*
$'*)$ *"*"'' !*+
!'"*'**'"' j`MNCDE##'!"T
*""'$('"'*+O $ *#'+M'T#(M!'
+FIJKQG $$')'$$ N'!* '*$&*"
T""$T"!'"* '!#( '*"&$"'"
$(*"$$#'"*$" )+U$CDE$$$$
(*$ ''($ ) *"(($"'#($ '*#H
!")!'$)$'!'&'+U$'" +Z&!' *"#"$ '*
&"'E"'V*"FIJKLG( `"&$*)#'('# "(('$
(*#*'#$)W*")#+ '*`"&$)+D")$*
E&$MNCDE$#"!"$$H $CDE#'!&"((' '*
X*$T!'!&"FYIJJIR #`"&$*)"$$'`"
Z#+IJJ[R\)'"+PLL]R\*'' &$)&$$&'#!"('$
+PLLIRN&'IJ^^RO"O"PLLQR '!"$ '*#+
O+IJ^LR_`"a#IJKKG+ U$*'"*!*''('#"
D*')CDE')ILbIPW*")' $W*")"!"***&$"H
T"#!'($'"* "$+O"'"''*$#H
( *' ('#+U$ $( ")$)$)T')"')
')(#""$*'($*" &#'*"('$!*$+
"#'MN)'('# E]+)'$ $'"&$&'")H
$"'"(('#)$'"*+U$ '"$*"FM)+IL+IG+
$'*)$ *"!"'"c" U$'*)" c"
!*'"'$'"(('#) *)(!'"#*!'H
$#+U$""$'('H $X*+S'('W*")'
"'""%'!'"$$*" $'W*")#*T##
*"$'$"'!'#+E *'('H &$' $`$
"'"'*"*ILd($#R ($"!"*"($&$+
**!'"'*"F\*''+
vwxyz{yz ET#($ klmnÿpqrslmÿ klmnÿpqrslmÿ

u
CDE)'$+U$
#'`#"$ tt u t t
!'"*#'!*
#*')('#$h|}~
|gf}~|hg|+
0123 456789
95 69657956
89:;<=> ?@AABCBDEFGBGHA ÿ!ÿ"ÿ#!$ !ÿ%ÿ#!$
EIBJ@DBGKLJBMNOLDBJ@GEGEILE &'(ÿ) &'(ÿ* &'(ÿ) &'(ÿ*
LCBKLJ@MCLEBPQRSTUVWREHBLKI
HEIBCXYJJCLE@DZG[BCBOJHEEBP +#( , , #
HDEIBGL\BJ@DBGKLJBAHC - ./
@JJFGECLE@HDOFCOHGBGX ,000000000000, ,000000000000,
(#123ÿ344#1ÿ5!6ÿ(!
,0000000000000000000000000,
,00000000000000000000000000,
#123ÿ344#1ÿ7!5ÿ(!
]BOJ@KLE@HDGLJGHLJJH[HDB^[LNLDLJNG@GHA_LC@LDKB PBEBC\@DLE@HDGXÌFGeEIBLMGHJFEBGKLJB_LJFBGLCBDHE
HA@DP@_@PFLJOLDBJ@GEGLKCHGGOCHPFKEGXÌ@GLJJH[GEIB @\OHCELDEXuE@GEIBCBJLE@_BP@AABCBDKBGL\HDZOCHPFKEG
GBDGHCNGOBK@LJ@GEEHPBEBC\@DB[IBEIBCEIBOLDBJ@GEG EILEOCH_@PB_LJFLMJB@DAHC\LE@HDXnHCBcL\OJBevFPZB
KLDP@GKC@\@DLEBHCDBBP\HCBECL@D@DZXÌBDF\MBCHA YGKHCBGEIBKC@GODBGGHAOHELEHKI@OGL\OJBrLGLD
CBOBLEaFPZ\BDEG@GGH\B[ILEOCHPFKEPBOBDPBDELDP seMFEvFPZBwGKHCBGEIBGL\BGL\OJBLGLxyEI@G
GIHFJPMBPBK@PBPMBAHCBEIBGEFPN@G@D@E@LEBPXbEFP@BG PHBGDHE\BLDEILEEIBE[HaFPZBGLCBDHE\BLGFC@DZ
[IBCBCBOBLEBPB_LJFLE@HDGLCBDHEOBCAHC\BPGIHFJP EIBGL\BLEEC@MFEB@DEIBGL\B[LNeMFE\LN\BLDEILE
MB_@B[BP[@EIBcECB\BKLFE@HDX EIBNLCBFG@DZP@AABCBDEOLCEGHAEIBGKLJBin@ZXrzX{lX
d?Y\LNMBFGBPEHKH\OJBEBJNPBGKC@MBEIBGBD^ ÌBCBJLE@_BCBGOHDGBGHAEIBGBE[HaFPZBGHDLGBKHDP
GHCNGBDGLE@HDGLGGHK@LEBP[@EILOCHPFKEACH\@D@E@LJ P@AABCBDEGL\OJBiGLNtLDP|eCBGOBKE@_BJNl[HFJP@DP@^
_@GFLJLGGBGG\BDEEHLAEBCELGEBeHCOLDBJ@GEG\LNMB KLEBEILEEIBE[HaFPZBGLCBKLJ@MCLEBP[@EICBGOBKEEH
@DGECFKEBPEHAHKFGHDLDLCCH[CLDZBHALEEC@MFEBGGFKI EIBQRSTUVWRP@AABCBDKBGMBE[BBDEIBGL\OJBGXvFP@K@HFG
LGEBcEFCBPBGKC@OEHCGXfH[B_BCeJ@\@E@DZEIBCLDZB KIH@KBGHAGELE@GE@KLJOCHKBPFCBGGFKILGPBOBDPBDE
HALEEC@MFEBGB_LJFLEBP\LNJBLPEHEIBgPF\O@DZh UÊBGEGLDPY}~YLJJH[EIBCBGBLCKIBCEHCB\H_BEIB
BAABKEiGBBjILOEBCklXÌ@GBAABKE@GBGOBK@LJJN@\OHC^ BAABKEHAFG@DZP@AABCBDEOLCEGHAEIBGKLJBX
ELDE[IBDLKHDGO@KFHFGGBDGHCNLEEC@MFEBEILE_LC@BG d?Y ECL@D@DZHAEBDELpBGJBGGE@\BEILDEILE
LKCHGGEIBGL\OJBG[LGH\@EEBPACH\EIBMLJJHEXmIBD CBoF@CBPMNnXjHDGBDGFGMNOBCGHDLJ@ENPH\@DL^
EI@GHKKFCGOLDBJ@GEG[@JJeOCHMLMJNGFM^KHDGK@HFGJNe E@HDeLOHEBDE@LJOCHMJB\ [@EIne@GFDJ@pBJNEH
BcOCBGGEIB@CACFGECLE@HDMN\HPFJLE@DZEIBGKHCBGHD HKKFCG@DKB@DP@_@PFLJaFPZ\BDEGLCBFGBP@DEIB
GH\BHAEIBGKLJBGFGBP@DEIBGEFPNXnHCEI@GCBL^ PLELLDLJNG@GXuDLPP@E@HDed?YPLELLCBCBLP@JNLDL^
GHDeEIBGBDGHCNGK@BDE@GEGIHFJPMBBcECB\BJNKLCBAFJ JNBPMNMHEIFD@_LC@LEBLDP\FJE@_LC@LEBGELE@GE@KLJ
LMHFECBGEC@KE@DZEIBENOBLDPDF\MBCHAPBGKC@OEHCG EBKID@oFBGXbELE@GE@KLJOCHKBPFCBGGFKILG\FJE@_LC@^
FGBP@DLPBGKC@OE@_BLDLJNG@GGEFPNXbH\BE@\BGeG@\^ LEBLDLJNG@GHA_LC@LDKBeOC@DK@OLJKH\OHDBDELDLJNG@Ge
OJNLPP@DZLGKLJBJLMBJBPgHEIBChKLDOCB_BDEEI@G ALKEHCLDLJNG@GeKJFGEBCLDLJNG@GIL_BAHFDPLOOJ@KL^
BAABKELDP@AEIBOLDBJ@GEGLCBLJJH[BPEHPBGKC@MBEIB E@HD@DEIBLDLJNG@GHAPLELZBDBCLEBPMNd?YÊNOB
gHEIBChKILCLKEBC@GE@K_LJFLMJB@DAHC\LE@HD\LNLJGH OCHKBPFCBGiLCEBDGLDPLCEBDGe{zzryBFJJBDBE
MBHMEL@DBPX BELJXe{zzy@ZZHEEerkstlXCLOI@KLJOCBGBDEL^
?FC@DZECL@D@DZeHDBHAEIBKILJJBDZBGALKBPMNEIB E@HDGHAEIBPLELHAEBD@D_HJ_BEIBFGBHAgKHM[BMh
OLDBJJBLPBC@GIH[EHIBJOaFPZBGGHCEHFEEIB@DP@_@P^ ZCLOIGiOHJLCKHHCP@DLEBZCLOIGLXpXLXCLPLCOJHEGe
FLJ@DEBDG@ENKILCLKEBC@GE@KGHALOCHPFKEACH\H_BCLJJ n@ZXrzX|lX
@\OCBGG@HDGHAoFLJ@ENHCJ@p@DZij@_@JJBLDPqL[JBGGe ÌBCB@GGH\BLCZF\BDELMHFEEIBLGGF\OE@HDHA
rkstlXYJJPBGKC@OE@_BB_LJFLE@HDGGIHFJPHDJNMBMLGBP DHC\LJP@GEC@MFE@HDHAEIBPLELGBELDPIBDKBEIBFGB
HDOBCKB@_BP@DEBDG@E@BGLDPGIHFJPMBACBBHAIBPHD@K HAOLCL\BEC@KGELE@GE@KGGFKILGLDLJNG@GHA_LC@LDKBe
CBGOHDGBGX UÊBGEGXY AB[ LFEIHCGABBJEILEDHDÔLCL\BEC@KGEL^
?BGO@EBEIBBcEBDG@_BECL@D@DZB\OJHNBP@DEI@G E@GE@KLJECBLE\BDEHAEIBPLEL@GCBoF@CBPi~LIHDNe
\BEIHPe\HGECBGBLCKIBCGLGGF\BEILEaFPZBG[@JJ rksty]LDPLJJerkskleMFEEI@GLOOBLCGEHMBL\@DHC@EN
FGB P@AABCBDEOLCEGHAEIB GKLJB EH \LpB EIB@C HO@D@HDX
012 34 56789

67
7
!"#$"%&%"''(' !")"#*+$',),- $1++"'.))',%5))1 '$,-(,)+,).,$'++

!.+,+()/01()'#'"'" '"2!(+$',),-!3 "5)&'(#'")1$)' ",)/45%3'6,"+(,##',6%.
.+,+"#45-,6"%!$&,++#5$),""#$"5)'."#"',6, "')1)1?4*-!5#"')1)))',%5)(,##',8,+1'B+
78'$"'9&:!(;(+5%3'6,"78'$<4,)=',+> ?4* 5!),!$" ',+")+)7C'')!/>DEEF>5+(&,)1
4$''>?",'@9&:!(<&)'>='$"))>A5';/8"' ',++,";/
01+"#()!.+,+5+,6G*.$)5!!. G* 1+%)+,-!.5+(>%5)"#))1

%$"+,('(""#)1 '"%!+"#)1)$1,H5/ ',)+'")(+,6($)!.+(+$',%(%.
01)($.)"5+)1+$!++%+"!5)+5'+ 4)"(4,(!7DEEV;/01'!),-+, !,$,)."#)1
"#))',%5)')1')1+)""!)"+'!),-(,#3 )$1,H5!!"&+,))"%()(,.(,##')
#'$+%)&+ !+,+-'.$""/A)5',6 &.+/W"&-'>.()),",-!,()+)15+"#
)")1 "))"$1, !>($,+,".%( )1G*)"(+$',%)1 '"$(5'/
)1)+ !++$"',6!++)1I")1$',++++$! (-)6+$,)(%.(-"$)+"#G* ,$!5(
'")$$)%!#"'+!/+&+&>J5(6=B+ )1,(+)1))1 !,+)+ '#"',(()X5(63
$',+++,)+,)."#I&+-'.(,##')#'"J5(6 )+()1)'+5!)+'")$"++5+(',-(/
B+I/=.)+,">&$+)1),#)1),' ((,),"!!.>)1()'+,!.!.L(+)),+),$!!.
!5+()15 '("#)1+$!>"+ !+ (6'1,$!!.''+)(/C!!656(-!"3
&"5!(%$"+,('(5$$)%!%.)1,+$',)',"/K# ),+#'"# !!(',Y5$+(,+>,63
")1' !>!.L,6)1++ !+>5++"!. '!>%+("$"+5'!656(+$',),"+/G*
)1!"&'("#)1+$!>"+ !,+$$)%!/ +5##'+#'" )1+(,+(-)6+8C>+,$,
01G*()5+))1'#"'%-,&(+MNOPQRSN %")1$++)1 !+5+)%)',(#"')1+$,2$
-!5+(")+%+"!5)+/G*+)5(,++1"5!()1'3 '"(5$)$)6"'./Z.[4#""($" ,+,),
#"'%(+,6()",$!5("')1"+ ! +') !+#"')1,'. '"(5$)$)6"',+/01,+
(T"'%$1'U"'+)('( '"(5$)+"#)+ ,+-'.+,-(.!,,))15+"#)1,+)$13
"++,%!/ ,H5%.+!!'2'+/[!,U8C>)1G*'+5!)+
0123 456789
95 69657956
!"#$""%!!&'!'"& Z[\]\^_àbcdèdfgh̀ji
"()*+!,%'#-#."!-
#"%!"'!%.##"/-&''!0 X%Xk-!R!l#* !.$""*!W0
,"&%"1%.2"+&%-$( !m!#Q-!"4%nopq *
3""##$,"" .+,%!-# !!#0 "-4&./-#$1"l.$+!#!$'"#"
"+ .#-,"'#"1 $-"&&! Ur! #(,nopstu"+"## v"W,nowxty-z{,
#!4( no|ptP{{"W,nops,nopp,nowxtP{{"W#(,
nopsV(X%4#&%Xk-!R!l#UXRV*
56787975:;<;=>?@A@BCDEF;GB@HIDJA>?>J>?K; 1 +" !$%"/-%*-# ##*% 0
& ##%k-!%!!""& '!-,
L;G=H?M>?K;NAJ<OG?G &!1 l!."%!-4%1'#1 "",-"4
P"1"#! QR 2-4,S-""+ T!"'"+ R!4l"#
!"4'!""'#(X%!!.%Xk-!
3#$"UST3V'#"%-#.!&!#40 %Q#+ %''"!.$+#'!&
!1 +"#."#"$(3*"%QR,"W"1,&&!, u"+"## !R v"
!l#(X%k-!'!l#* l.$
W UnowxV
1$!" '#%'#"%-#1 W
11"1. +"#.#&!$!,"&'".#( %!$ #$"&%k-!1'#k& &"
X%".1 1 41-''!"-*%% %!1 &"1%"#,41!"#,& 1"-!
!!"",%4!&%'!%!!
'#"!#&!1*"%"1'$"% "*%
1'#$1 $'-."#"1 * $&!1%"! 1 ""% %$ ''!&!1 l."%!-4%1'#
"U'(noV(
1 "2.(X%'#" !!&!!1 #
! '!'"-"4 -#'!-&!1 X%Xk-!R!l#- !"{!1"#4$
%4!$(X%'"#'#"%-#.+!$ !".%k-!#%!!""& $'!0
!"-# "!( -(P'"l%!!"" !!"..$.%%"!
Y#"WQR,%'##!" "+'!0 '%$"# !$ '(R!-0'"l!1
""'""%!%#4-4+#'1!% .1'#$ !%&!1 %!"{
+#-"'%&%-$(X%'##! !1"#4$!".%k-!& '"l'!0
&"#"!#$ #!"!% -(Tl"" !!& ''!&%!1
'#(X%"'!*"##.!'".#&!11-"0 !"%!-4%-.$%XR'#"(
"*"%%'#'!'!"&1'# } "4# "*"%%k-!#!1 !
!&!!( !"{UX.#nq(sV(
~J<;567 k1'#k-!'!l#%!#
P#+#- R!- P1'#"{ X1'!-! u1'""
n(q u!1% n-. qxu R%"##'%"!1%U!&V
(x 44*%" n-. }1 ) !0W,x1"
(x 31
n !
" %
qxu ##*'-!"{
-. %Uv
vWV
p(q #"+ n'" }1 P-&&,P'"%#"+*"%'"1
n
!1+Um$ QV
w(q Q!W&-!!. #" }1 r&Q!W,W&!x1""
."#"4* !U).!* "#
%!QV
o(x R- n'" }1 uW"#'-"+--1"
UR#!, ."r!V
nn(q 3#1 n'" }1 P%## #1UR#!, ."
r!V
n(x ) !$ n'" }1 v"&P+!U ."r!V
3'&!1y-z{Uno|pV
) !"l %&!!/-"!."1'##$%!-4%1'#'#.*1#!%
.3! 1'!*"%1#!"'!###-
012 34 56789

67
7
!"#$#%& EFGHGHIJKLMNOIPJQRNSTVU
"'()#"'*!#$&
%+""%#",-#"' WX*.."2!W8""'
%(%)*'"'%. 34Y:(%9#!-9<"$,0
)""%")%" #& (A)'0")0#%A
$',-" ("! !%)"'"-9<"$,
'"#"%/0112,34567, 0")0#%#"' 9#&
8" 9%#'/-(3:,67% -
"%#""##'"#"'(. "" '#*)A,-A
%",;"'""'/% '"
"0#%##"#&
""2'!!."*.# '("! #&#$*"()"
. "%"''''7'& ''"1)
("#"'()
'9""%/X*'Z)"
#"-<'& 34457,-!! '"'(##"'&
)"%%"'),=".*#"'& "#"
)%)" "%#. #"'.!"')" %%"
"%#""##',-.! ''1'*' %,?''")
'#%'' # #",-# "'.%&
"!#"'*)#$''$& ')[(" '"
('%"')'"") 0#%%"' "#"'*
# "
%.)"##'"%#*", (",-%"'(#"(&
>""-<% "'!9#%'")"%#
" " #(%,?#& '""%#'"%%#.'
%*%!## '*'''),-#""#)!!
"9'-<#"',<#"*! #A!"#".!#&
'*""%#"'( !')"& "!"."%%)'"
"',<"'"('"(. %'! .%#!"##)$'&
'.".''1'.),@)'! !"!1'#"#",-#"#""
#!#$9#"'" %"''#"'"0#%
0112/34567$""9& %"'%"\B?'#",
,-)(A)9#"'".' ?"'!"0")0#% \B? %
*)"""'#"'' ,B! !'()#%*')
'##$2 & %"2)"( '""% !!,
"!1!')!'A)!'!. ]*."0#%%"'*(#"*''
9#$""'!"), ()<"./34^^7_ ),/D::37'_!'
-"%)2*.2"')& '"."2/D::57,
!"(%)(.".,C%( <!" 0#%%"'%
,/D:::7."2!.(%"' %"9*\B?!'#'
9*!'"! ")' %"'* "\B?,?\B?
),
--9<"$(##'"%)#& + '!9#"'".'*)"#"'&
$#"'!"'!(2 <"$
#$#"'!"),?-9
.##'"##!""2%2""', "# #'#"*'9*"%"
=".*%)9#%. """!' !',-')!%&
)''-<'(& #''*
##9#"'()
"#(.#&
)" %"'""!) *9! #'")# #""#"'
A%"-<."''"'! "'',0%" -9<"$#
',C#"'! #"*''."'/
''1'%"'""!)"'!"%%'' 0#%7%)(' '9"()0")
"'(#*'
()"!#"#""A,
0123 456789
95 69657956
!"##$%&'(#)&$ ` abcdefgv ]8+&))!60%0)$&*!
*&!+&%,---./%0%&1*+ #6)07k&0!%0!)
2&!3---45!$0# ) 0 j0%0 +&) K0!)
&6!0%07!)!0!*6# k" <'= <w)0B"0&)
)!0&8*#&!660$*#07) Y'= tF 0C0:p7?
6#%0!)9:;&*0!&'3<==>?'@!!&&83 ['= x$!:G)08($?
%&0)%%&!"#060!0!'A0 I'= GG&G&
07)3%2 *)&0060"&&B H<'L o0!)/GC:K%%0!*
!0! $7"#6"#007)0 H L'= H>
A0$?
w )0B"0&)
%0+!!'@/$%&")&!7+&&!+&& k)0 <'= ='=L w 0!B"0&)
$&C!D00$:;)EF!G+&&3HIIJ?' Y'= p ) 0
& %%&):;?
K&)8&#0)$0&3 L'= t
)!60F0*m)
))&&8HLB%&3!MNOPQRST&C" :;);!?
RUVWTXOMQ:(7&H='YZ;)EF!G+&&3HII[?' J'= k"%C&:u&?
G+&&:@%0&HII>3%0& $$)? H='= q#0!&&%C&:u&?
##&0 0!#+\)B80 k& HL'= ='<=w 0!B"0&)
<'= ='<w!)$ #&0!B"0
&']#0"0!3,L9#"& &)
\)&8,L9#&8& L'= k&!! 0C0:K0$)$?
!#+\)&8,L9# ['= @$0 #:q06?
60)8&:(7&H='L?'G+&&:@%0&HII>3%0B J'= ;8:r&&$1?
&$$)?8#*&#7#+! I'L k&!% #%:A0Bx8?
6060*030$3!^+0&7)60 H L'= H'Lw!)$ #&0!B"0
&)
/)0&'_0$"#C%&6# o0 <'= o &!*0%60)m):q06?
\)B8 &$#00%)7&#!! Y'= G#&70:r0#8?
L'= ='=Jw 66B"0&)
['= t"!+
àbcdefgh i/$%&60$ 060$%&)!60 I'= G&08!
%0)$& H='= ='HLw 66B"0&)
k& HL'= ='<=w 66B"0&)
j0%0 +&) K0!) @! %!60$;&*0!&':<==>?
@&&#7+&0)60$=HL
@0*8 >'L (7*C!60H#
>'L l0%m):_&#1?
G0$&F! n'= o0"i!*GC:p7? )%%0'r"+03# %6 0B$!&
)*0 $#*$8$C#7+ &$07&60
Y'= k)*0GC:q0*0?
Y'= k&(GC:p7? & *#!3:"!)0?37)
['= o0!)/GC:K%%0!* $8707&60"!#0!0

A0$? 60)! #":k+3HI>IZk+!
i** L'= ;8:r&&$1? ;0C3HIJ=Z_0!3HIJ>?'
i**^+0 Hn'L r0!7&!**
s0* n'= s0*j0C:rBG? @&3#7&86#7&)&&0'
$%&/ s&7!x"&:<==J?6)! /)&#6*
>'L t)!60F0* #H LB%&+60/+0*'
@" ##(/)0K02&3&0#0!78
0:;);!?
['L A0#&8\)F!0*m) 06060%']##$&
I'L s0* 0:(*? "!%0607&8#02+06000$B
t! ['= G66:;/"&&r)? $!!'(#060%0 #0%0
HY'= i%0 66:;!*& !660#& ))$'(#060B
j1s0? %0)!%0&8 &70#%&
u&& ['= k)*0GC:q0*0? #$"8%r7)660 &70%r$0'
@! %!60$;&*0!&':<==>?
@&&6#7+&0)60$=HL (#%&0,)!9&C0)0)$'
012 32 456789
95 696
!" +$"*$#)$
#"!$%&!!'!$#"$( !'"!+"*#"'#$
")"*$")"!'% !78$"$$!%
&$#$#*$$! '"
$'""%
&!"#+"#$!," BCDEFGHGIJKLGMKIJNOJPGQHRSTMJM
!$#!$#*#!#"(
$$!-.$"#%/-.0#1' -. #,$'," "!1!+*
!"$+###$'""*"$( ##'#2'%/$$
2'%/"$$!-.!," $!$2+!'#$+#!
!$#$ !#"$" $!#6#7-.8#7,$'
!2'%&!"$!$#!+*$( ,"%8U$'$#)"$+(
%&!!"$"$# !"!!'##+$$ !
* #$#2!")# ##!$#$$!+$!+#'$!
*')!"$$""*' #%"#""!1'$'!(
"+$$"$!$'$$!+( #$!!$#$$$+
*%&!!$*+$#+$$3$ $#"%&*?V%A!$2!"$#"
2!$ #"+"!1$1( "#"+'W!2*
'"$$$ $"$# #"*##!5"$%##$'
"%
.#+$!$"#$"#2! $+ '+$$!"$+$
!#)"$#+$#"% !2*#

"# "+'!+*"##
"#,"$?V%X%Y%
&$," '+'
"$%4!+$*5$#$!(
# ##!+$"*'"!$$#"*
!$#"%&!'6#$! #'( Z[\]\Z^_à_b_bcdefghagicjklmndgd
"!"#$!$#"#!' gkopfccqldnrachd
5"#$!+*" ##$!!'(
$$'#'"!$$'$$'"$% /$*$'"$$'"$(
!!' $*5+'7#8$ $#"+'#'!'%
$$!$! &!!0##!0#
!"!2'%9$! $#"*'s"$"'#!+!0#+(
!+$"*$"*"!+#'% !%92#""!$!
/""#+##"$ $#%
#$!'$$"#"6")"$(
#+$#"#$
"$$%&!$"'
2"$$2$# BCDEDBDBtIRJHJHuOvGwRHGSJMOM
5"!###+$:#5;$'
")"$#"<*"$*$$*"$$ 2!+2!!-.#,$',"
3$%=$2+$,"#!+* !$#!2$!$#$0#%&!
2!$(3$###+$#"!! )$$+#!2!2#$
$*$'"*#%>+:?@@A $#"!")""$'%46#
$"$ "$;!#!*$"( $!#"$#"##
*$*$$**$*( ##$#"*#"$!$#"
*'$$*"'#"$#$ '*'"$$$"$%x$"(
"$% '22"!7"$%8&!"$#
.$!,""!1'3# !$#$$+#!2!2#
!$$!-.#%&!#+$$ $$#"2!!"$'22$#
0123 4565789
5978 8568 0
*'''7+&()7"())"#&$"4.7 !*$--*'+&+A+&2
!"#$%&$'(%#$)*)$+#%'"%+#!#, *$*-."#"
;&+'-#""0[7 !*$--*'+2*.*-"&+'(% +'($S
/01 +2#$ !+&3%&43%#,56"'"%+#!#, *$*-."#"7 +#)7
2 7&'8
90:"*4-#"7!+&'(%"&4 ("';#7%-#$"<+"*+%7+"'*$+2"06$&(+-*4&+*&+#"7 %&$"$"("2 7&'
=01 +2#$ ;77+%&$"$"("&+4*--&+*#$#$)#"2&"#"("(*--.("'#$+"*+%7"('#";#77 ?%!S
*!!+&!+#* #&$&A2 *+"*+%7"('#"0Z&+2 *; $'&
(" 7 4*--&2 7&'E2&"-.4%*(" *2(-#(' &A
>0:"*4-#"7 ?!+#2 $*-'"#)$*$'"*#"#%*-*$*-.""
"('#"#$7 *+*7*"%&$,#$%'("&$-.*-#2#'
*0@*#$ AA%"*$'#$+*%#&$"A&+*$*-.""&A,*+#*$%
40@(-#,*+#*%7$#B("8 $(24+&A'"%+#!&+"*+ +*'#-.*!!-#%*4- &2 *0
C0-%D*$'&!#&$*--.E"%+ $F!*$-#"" 6$%&$+*%;&+OA&+\A&&'*$'%&$"(2 +!+&'(%"
6A%7&&"#$)&'&%&$"$"("+*#$#$))&&G06A%7&&"#$)&
'&4*--&+*#$#$))&&H %&2!*$#"E; $'&(" 7 %&24#$*#&$2 7&'E
G0I&$"$"("+*#$#$) "#$% 7 %-#$%&2!*$#"&A$7*, %+*#$+2"
7*7.' 2#2!&+*$0[7";#--7$4 "())"'
*01(+#$)#$##*-+*#$#$)"""#&$"!+&,#' !*$-#"";#7*
;#' +*$) &A!+&'(%"#$7 "!%#J%%*)&+. 4.7 !*$--*'+E#A7 !*$-#""'&$&(" 72
40K*$-#"")$+*'"%+#!&+"D*$'#'*"A&++A+$% "!&$*$&("-.0
"*$'*+'"F
%01(+#$)"(4"B($+*#$#$)"""#&$"!*$--*'+!+&,#'"P.!#%*-"B($% &AW%&$"$"("+*#$#$)X"""#&$"
!&$#*-+A+$% "*$'*+'"*"; --*"!+&'(%" ;&(-'4 7 A&--&;#$)5
'0K*$-#""+*%7%&$"$"("#$+2"&A*+#4("E+A+$% 6$##*--.E7 !*$-#""*+ ?!&"'&7 $#++*$)
"*$'*+'"E*$'"%&+"7 "B($%#$) &A7 !+&'(%"0[7.*+*"O'& ,*-(*7 "$"&+.
H0L*--&+*#$#$) '#AA+$%"*2&$)7 "*2!-"*$'&;+#'&;$7
'"%+#!&+"7*'"%+#4 7" '#AA+$%"0[7#"&%%(+"
*01(+#$)#$##*-+*#$#$)"""#&$"!+&,#' !*$-#"";#7*
;#' +*$) &A!+&'(%"#$7 "!%#J%%*)&+.0 #$"#-$%0V7$*--!*$-#""%&2!-7#"!&+#&$&A
40K+&,#' !*$-#"";#7*;&+'-#"D"*2!-"%&+"7 F
*$'+A+$% "*$'*+'" 7 *""#)$2 $E7 !*$--*'+*"O" *%7!*$-#"&
-#"7 ;&+'"("'&'"%+#4 *%7"*2!-01(+#$)7#"
%01(+#$)"(4"B($+*#$#$)"""#&$"!*$--*'+!+&,#'"
+A+$% "*$'*+'"*"; --*"!+&'(%" !7*" &A7 +*#$#$)##" ?+2 -.#2!&+*$7*7
'0K*$-#""#$'#%*;7#%7*+#4("*$'+A+$% !*$--*'+2("4 %*(#&("$&&-*'&+&3(')
"*$'*+'"A+&27 ;&+'-#""7&(-'4 ("'#$7 "!%#J%
*$.'"%+#!&+A+&2*$.!*$-#"0Y&; ,+E7 !*$-
"('.0K*$-#""2*.*-"&#$'#%*"B($% &A*+#4("
&$"%&+"7 -*'+2*.*"OA&+%-*+#J%*#&$E#A$ ''0\"(*--.E7
M0N$% 7 +*#$#$)!7*" 7*"4 $%&2!-'E!*$-#"" !*$-#""72"-,";#--4)#$&2&, &;*+'#$##*-
!+A&+2*$% #"%7%O' %&$"$"(";7$7." 7 &*--#"&A'"%+#!&+"
-#%#'0
*0P"(4"&A"*2!-"*+!+&,#''#$'(!-#%*D+#!-#%*F
($'+*%(*-"#$)%&$'##&$"
401***+*$*-.Q'*$'*$.#""(";#7+!+&'(%#4#-#. (4"B($-.E7 !*$--*'+"7&(-'*2!&
!+&,#' !&$#*-+A+$% "*$'*+'"4*"'&$7
*$'<&+*+#4(("*) -*'&*''##&$*-+*#$#$)R"#$)
2*.&%%(+*)*#$*A++S+*#$#$)0 #$##*-%&$"$"("0[7" +A+$% "*$'*+'"*+%72S
T0I&$'(%"('. #%*-"E"!#%"E#$)+'#$"E&+!+&'(%"7*%*$4 ("'
/U0P$*-.Q *$'+!&+'** &7-!7 !*$-#""#'$#A.*$'+2 24+7 "$S
"&+.*+#4(A&($'#$7 "*2!-" ,*-(*'D]*#$.E
/TMGF06$)$+*-E7 !*$--*'+"7&(-'"+#, &("
*%(*-!7."#%*-&43%"*"7 +A+$% "*$'*+'"4(
-#"&A!&""#4- '"%+#!&+"*$'+A+$%"7*%&(-' #$"&2 %*""!+%#" ;+#$'"%+#!#&$2*.4 ("'
4 ("'&'"%+#4 7 !+&'(%"0V ;#--+A+& #$"*'D[*4-/U0HF0P7 $?"""#&$E7 !*$-#""
7#"2 7&'*"W4*--&+*#$#$)0X6$!+*%#%E4&7 *+ ?!&"'&7 "*2!-"*)*#$*$'*"O'&'%#'
7 %&$"$"("*$'7 4*--&2 7&'"7*, *$*!!-#S &$7 !&""#4-+A+$% "*$'*+'"06A+A+$% "*$S
%*#&$0Y&; ,+EO !#$2#$'7*(-2&$ *-0 '*+'"*+ $&A*"#4-E7 !*$-#""%*$*-"&4 *"O'
D/TTTFA&($'7*!*$-#""$''&!+A&+24+ &,+4*--.'J$ 7 "!%#J%'"%+#!&+0[7#"+J$S
;7$+*#$'4.7 W%&$"$"("XD+*#$'4.'&#$)F 2 $&A7 %&$"$"("-#"&A'"%+#!&+"E+A+$%
+*7+7*$W4*--&XD+*#$'4.4#$)&-'F2 7&'0 "*$'*+'"E*$''J$##&$"%&$#$("($#-7 !*$-#""
Z+B($-.E7&; ,+E*%&24#$*#&$2 7&'#"("'06$ *+"*#"J'7*7.7*, 7 4"!&""#4--#"*$'7*
7 %&24#$*#&$2 7&'!*$-#""'+#, "&2 '"%+#!S ,+.&$ ($'+"*$'" *%7+2%&2!--.0@(++*.
&+"&$7#+&;$7+&()7%&$"$"("*$'&7+"*+ *$'1 -*7($.D9UUUF7*'7#+!*$-#""'+2#$
010 23 456789
95 696
!"#$%#%$%"&%!'"('$(#$'$''"()'*$%*'+,'!"-./%%$%#%$%"&%!'"('$(0%$%,%( "'
(%&$!*%!,(1#*'"++'%%"&%#$(##%$%"!2%2$'/&+&'!"34(567789
:$''!!$;,!% !"
<=1' >?@A#;"("2!0"%+!0!/+2/!%$5'++0%(!;,$"'"(!/%";+0",!5%++=%
<&!&/ 68+#8B+,!"#CDEC,!%$"DE$=F!(-5$'$%;+%"(%(&!&/0/=%3F"("5G"2+'"(9
E,$;" 68+#8B+,!"#4'+=%$A(%+,H%!$'2/!E,$;"4/=%13I$'4'+=%$D<"-5
E'$(!0"5JK9
L, 68+#8B+,!"#E'&'$(<,%$OL$M,%$!L&'"L,3E'&'$($-5<'"C,'"5M,%$!L&9
:+"( 68+#6->8B+,!"#N&$&= M,$%:+"(%H!$'&!3N&$&=D-5P"&-5I,"!Q'++%15
NR9
J'/+,' 68+#6->8B+,!"#$2"'OL+N%H&'"J'/+,'3J'/+,'<-:-5L<'"C'S,"5N%H&9
N%(&"'+ 68+#>?B+,!"O#L%'&+ ,!/0'/3N%$$%++R0M/'$'&%,!&'+5P"&-5"&""'!5TI9
E,!!%$1 T"%%&%#F#%'*%$ E,!!%$L,&'"(13UV';&W(5P"&-54 "!"X<'+%5V9
$%%<(' 68+#>B+,!"#</'!'OL &$%%('3<O/L'!'E%*%$'2%P"&-5I'10'$(5:9
W$,!1 68+#Y?B38Z69+,!" #4%+&/AT$&/'$( '+%X2$'%X&/%$$1#$,![,&%&&=!'+#$\%"
&"&%"!$'!%'"(4%+&/AOL 6??B0/!%2$'%[,&%#$&"&%"!$'!%3",2'$'((%(93U4%+&/A5
"&$(5N:9
M$,"% T"%<,"0%%!OL %(,$,"%3<,"0%%!]$0%$5<!&=!"5:9
.;'&& T"%"&/#+'$2%\%E%%&/X",!<#!%$DN!%$&/%0"2!;'&&3UV'!"'+.;'&&5F,*++%5
JK9
G'$!/1 672#;+'&=($!#$N ,$
N,!1 ^_àbccde_OLfg_ebhibebjkabh_j_glm
V,!!1 >nYJ$"%$ '+!%(!'&/3/%++%('"(&,!"!%&%93J$"%$-5"&""'!5TI9
ocbpq`bll`rasl_a tqjkqhrlrqg
:'$%!! 68+#8B+,!"#R'$""X:'$%!!T$2"'+%3P++*'<'$""5<'$""5P!'+19
<0%%! M'"%+!0%$%"!$*(%(0!/'$%#%$%"&%5;,!0%$%2*%"'>'"(uB+,!"#,2'$0'!%$(,$"2
!$'""2!'"&/$!/%&'+%
W$,!1 68+#8B3 8Z69+,!"#4%+&/AT$&/'$(OL '+%X2$'%X&/%$$1#$,![,&%#$\%"&"&%"!$'!%'"(
4%+&/AOL 6??B0/!%2$'%[,&%#$&"&%"!$'!%3",2'$'((%(93U4%+&/A5"&$(5N:9
G'$!/1 6';%++<,'"1#$%/'&='2%(,/$v(&%(3'(%"5VC9
M+%'%"!%!/'!#$!#!/%%'!!$;,!%*%$1$%&%$%#%$%"&%!'"('$(0%$%&$%'!%(v!/!/%(%'+-E,!#$!/%'!!$;,!%
";+('(%w"!"'"("$%#%$%"&%!'"('$(2*%"v!/"!'"(%'+!,'!"
':+++,!"'(%,"2,++2'"(,X#$%%($"="20'!%$3,++2'"4'!%$F!(-5+,;'5NT9
;:++!/%$)'*$!'"('$(0%$%!/%'%'!/%#$'$'!'"('$(
!/%,!';+!1#%'&/!%"!'+$%#%$%"&%!'"('$(#$ ./%'"%+!'$%'=%(!(%&(%"!/%0$(
&/%(('$&/%%%;1/'*"2!/%&$%!/%'!!$;,!%" "%%(%(!'"&/$!/%&'+%,&/'""%!%H!$%%
'"'$$'!%"%&'+%- $+2/!!*%$1!$"2-4%'+#$%S,%"!+1'++0!/%
R,$"2!/%w"'+!$'""2%"!/%'"%+!&$%X '"%+!!(%!%$"%!/%%S,%"&% "0/&/!/%1
'!%!/%&$%/%%!-./%1'1;%'++0%(!(%&(% 0,+(+=%!%*'+,'!%!/%'!!$;,!%5#$%H'+%5
"!/%&'+%!,%5'+!/,2/ ",$+';$'!$%0% *,'+'!!$;,!%w$!3,"+%!/%%'$%%$#$%(%X
,,'++1,%%!/%$!/%,"!$,&!,$%(+"%&'+%3X '$'!%+1 "'&+$%*'+,'!"&/';%$,&/'!/%
+'$!W2-6?-69$!/%68X"!,"+';%+%(;H&'+% ]$%!'2XN'&E%!/C,(2%PP9!/%"'$'#++0%(;1
3W2-6?-x9#$!!,(%- !'!%5)'*$X;1X,!/5'"(,!/#%%+'"(+'!+15'#!%$
%H%&!$'!"$0'++0"25'#!%$X!'!%-W$%'"X
%+!/$(%$'1;%&/'"2%(v#$%H'+%!/%1'1
&/%!(!/%!'!%5)'*$;1,!/5'"(,!/
yz{{|}{~~ÿ}|{}~|ÿ #%%+!%$$$!'$'-T"&%'2'"5!/%'"%++%'(%$
'=%,$%!/'!!/%'"%+!'$%&#$!';+%0!/'++
{ !/%!%$5$%#%$%"&%5'"((%w"!",%(-:!!/
y|} "!!/%'"%++%'(%$0++!'$!!%*'+,'!%[,(2%
GH'+%#'68X"!,"+';%+%(;H&'+%- $%$(,&;+!1-
0123 4565789
5978 8568
!"#$%&#'""()#" RSTUTRTVWXYXZ[\]\]^_`]Xa\bYcXdZefgh\i\a\Yj
*#!+&,$##*"(-."/,"' WgZ\]^kZ`\]\]^
0#+#,"''"(#$,'#+!12,
'3+#4!,"#'+$$'"5#"( .55+$','""(,,"'
,512,#!'""16,""7 #+,,4!#",#$,!+*&("1
#5,#'#"#$,("5"/, l#*4'/"'/,:'*##','#"'
"+'(4,"*#'+8#'57 !+#+'5"=!+(#""1!&#$57
#',9$#','#+!12,*#'+#"7 #&!+$#','!+'4+#,
"*#'+/+:"#"/"+#$","+' ""'12,+$'#5,#"'
*#,4'$'""+'+4&#",#' "Q+E,"#'"*!+,(":7
,+'#'12,''"!5&'#$!&,+ "4#$,"'#"$$#+*,
*#'+80#"94&$#'4'#$$##+ "1."*7'"+",$$*#!+
"+'#"''#+!1 "#"70,!4 &"#(":"+$$'"5#"(=!+(1.$,''
(4","+#$,#"12," (":""7#+"'#"$$,
',"3+#"+,'#!(,#""!*,, "#'"*+'5"*,,=!+(,#!+
#$,*#'+/'$'""+'+/"++:"#" &$!',''"+",!#$*,,+'#'1.$
,#!+&!+",:!+12," =!+('"#''#+!&,","+#'!'"
'#*+#++#'+'5,'#!(,#""!1 #,'""(,1l#*4'/,'!!!"+7
2,'#3+#!",+'#'#", ,#"#'*#!&=,4'#&5*,#"
&#1 #'*#+'#'12,'#&5"!!&
."!&!"#","'0#+ '#4++!'"($*#"7#"7#"'""(#"1
#,5(""+3+###3,&7 @$$18FGGG9,#*+,'""('#('+
#,,'4#!'+12,,",4# ,"+'++4#"#+*,BG#$,'BD
++$,'!,#',,*"#! '&!+'+1."#5,+'*
*,,'#+!1;:"5"#$,#',/'$7 '(8G1CG#"BG5"$#'#0+Q+'#5
'""+'+/"++:"#"#""!!", "+K4#'9"+"#,'5'8mG1GM$#'('"7
"':+,,,&#&#' ('"+#!'912,'""+#$#!"+,
,/&!"/"+,4'#"!"+'"+ ,&(('""($$#!''+*,",,#7
,'5#51<#*,"+''+ "'$'""+'+*'!"5&(!#!1>&#*
#+'5"=!+(''#+!&1 $#'5#'"7+,+!#"#""'$#'5"
>#5#$,4&"#'0#"84#&!7 5#"#'"(1
'9',>2? !&#",#4''( ;"#5*#'3#",$$#$$+&37
"!5&'#$'#+!(#'8@4"+A#"/BCCDE &'#"#""'""(,&"!&,+8N"+
;!+(/FGGH9*I'3"+@48FGGJ9 1/FGGD/FGGL912,!,#'$#!"+,55+7
*,,#4','#"#$K4#'0#""+, (',#5!'Q+$+&3#"'$#'5"
"!5'#!'$'"#4&*#'+1$* ","#'&##,+!'"('""(+#'+!+
'"*#'+'@$$18FGGG9$#' '""(5*0""'$#'5"1
=!/I##18FGGC9$#''#+!/I'3 ?I#"18FGGB9#$#!"+,$+&3",
18FGGL9$#'#"+*,'#""*##!"7 $#'5#$'"#5#"""#/"
'/;4!18FGGM9$#'N'",,/A #$4'",#*"#,"$',+'4
"+@,5&'8FGGL9$#'('"/O'"31 "+!,'""('#"+5+,
8FGGD9$#'+55&"/"+;!7!51 "5#'#""1<#(!'72''#"18FGGP9
8FGGP9$#'+'("12,''#!&,+'#' '"++'4"#4',."'"#4!
#$("'+'4"!"('5"##(, !(12,''""('#""4#4+
'0'5#Q+1"05*#!+&<"+=" $+&3#"'$#'5",#""+"'+
18FGGL9*,#'"+('#!#$+!4('" '""(+!'#""'+,'."'""n
,.4#'@##!,#"(!(#+'& '$#'5"'4#,'$#'5"#$"
,"#'+$$'"5#"(5#$%$#!#!) '"+5#'#"4"#"1l#*4'/?',"#
8#!"++591 18FGGG9,+$#!"+,$+&3*#4
011 23 456789
95 696
,,, )"%"P)
!"# $%&% $$"" !(,V[#'
"'(""$")'
*)$"
+,")"&) + <=>?><>\]CHMD]M^_`^KCHaMb`HGF`^GHI
) ,++
,$ ,"$)"&-,%") 56V1'W'V'0cd :"
,")$,)+ ,)&"$)c$
,")"& ,"$"$&$ +""))""$%,"%"))&,9
,"'*%$"") ),""$O
)'.+)+&$ +)+'S+%
/)"'!0112#'*$)%$,"3+ $"& ,",9
4"5)"6,76 %",,"'*
)8$%)",,",9 ))""$+,")&)%
,$,"+" ) &,Q) &$''
- '5""$,"-" +e, ,"'fg- ,"
,"!+ ,"$&$ 9 ,"h67%$
,$#+ ""$ %" 6$5"$(,"%&,,9
,"!+ ,"$&$ ,$#' ,"V2[0!," )
:";-,",% i( &Xj.&0112#'k
%' e ,$,"fc,",9
R,Q&c )))"
,", '.
$"&,", 9
<=>?><>@ABCDECFGHIJCKLDMN ++"$,"
," ))"$
6$,) ,"9 $ "'
O%P))%)" *P)) ,",
&)"& ,"$)%") "+ ,"+
,)$'* ,",,% "&%'
)""&O'*Q))"%"9 *8$, $
)"" ,""),"&),&"$ !#%)","&"$Z
,"'6 ,"))""$% ""$ !&#%)",",)&"$Z!#%)"
""&),R ,"%") ," +,"+"Z!#,"
&- ,"%'S+%+ &"$Z!#,",)&"$'
&$TOO)""U)!V222#S!011W# 3) )""$%"&"R
Q&+)" , ,"':"
+ ,"% ""$)"9 T)"""'!011Y#* "'!011Y#
)"$!"" ,"%&) T/!011W#"
,"#'S+%&)9 ,'d+8!011W#,)&9
)"8%") "l%"),", ':)
)")%'7" "'!011Y#,)&" -9 ""
"" ,"+""&%"9 ,,)656!6X($3(#'6 l
+,"'X !+l9,8#",",9
,&",,,-,9 P)+"" $$"9
",")/P)&" 9 $P)!/mS)011[#'5""$
,"&"8)"&""+!((- :"(8+l9&,
V2WYZ:V2[W#'*))""$"$O %"&"+",noo+++',"8' o
&$"$%'S+%"$&$ !h""p9S)8"'011YZ* "'011Y#'X
0123 4565789
5978 8568 3
!"#$#$%%&#' )#**+*$)++,
(")*$!%"#$*+*$, .%$%*+'6%!
)+)-+#)%++%))$% %#"%F#$+#%-$,
!*+.+#+"%$*"# +*+!*"#+'6%#"
*+' "+%+++:+"*!)*+,
<-%+#)+!'4
%#"++!*%"++'($+*+5#
/6878 5059168235 "+:+<%%*+)++
:"+<+"+,.++"
4,+!++!$.++%*"#+% +)+P*)'($+*+5#"+
)+$$$+%*++"+%+#+ :+<%%*+)+"+:+<
%!.+#+%"."#+*+, #++"$$+%%+$+
5")+!++%*++!' *)'
6%+#)*%+*+%7", =+*$)++.%*+++%
)+!$+*8+##"%+.+- $+%+#++.#+!%%
9,.+#+")+*+!:;<.+#'=+ --%*:Q+"RSEBOOT<'(%+%
%-"*++!#""%+.)+)+ *+5$+%+#++$)"
&#+:>?@<.+#'A *$;,.+#! +F'A#+F-$+%+#%
>?@.+#+%!)#+, +"!8"-%)++F.*+$
#!.+#+%")+!+"*++! +%*"#+"%+F-%)+'@+%*+,
:9-'BC'D<' 5#)#+.+%*"+.
A%,+!++!$.++%)+$$ %#+F'6%#+*F)+
:*"#E*+E+"*+<+"+ +--%E+"%%+.+$
$$:*+!*"#E*+!*+E U+FV+*%%$+%+#
+"*"#!*+<$+!&#F!"+ :9-'BC'W<'
X3 878 5059168235
A **+)*++!:=YA<$+%
+#$+%*+"+%,
+:+-)<+)-%*+$%+
+#:ZPF%#EBOOD<'(%+%*+,
$+%*"#$%*8"+#
+#"+%.++:#)<%++!'
($%#+++-):+<+)-
%*+%%)+P!$%.++%#"
7*+"!%8")'(%"
$%*+#%*8+#)+!
%%=YA%#"")#")+'
[#+!E$,+"*+%+)#$.++
GHIJKLJM A7+)*$+;!>?@*$+*+ 7*+"%8")+-$)+#DC
:!)*+++)"<$++)+)+:++)< \ ^
C]:9-'BC'\<'
%+':_CCO<$#"%+$#+",
+".#)#%$:.>9<'=+N%7
")+!:;,.+#<++7*++! #)%.+#")#%E+--
:>?@<$.#)#%$'9++)%*++ $) +#BD+#_T]'6%!
%+7")+!:;,.+#<"**++, +++#+++:Y<$+%
!#'=+O%+*++!#E*+!$%
+##+%+")+!#%.#)#% +#$) %=YA #'ZPF%#:BOOD<
$+"+7%++)+)+' "8"Y+%+$%.++7*+"!
012 34 56789

67
7

!"#$%&''%($$#$)%*(%")($+,, +,!'0",)+)(!.()(0,)(!1!0"!!))0+-/)678)(!,

)(-))" &)(#$)+%)(.,%,%/%0!,1&0)(%#.+2. ,)(.+)0/%!0"!!))0+-/)698*
!))0+-/)*30")(#$)%+)+%4+5,))(!))(#!,$+%)%0"0
: ;,!"#$&)
<=;#!,$+%).,%,!,.
#$)%*>,)(20%)#$)678)(0
+%5+%!'0",)!",')(
#!,$+%)%+,)(+0/%!'&)(
%#.+2.)0"*>,)(%.,5
#$)698)(0+%"0
!'0",)!",')(
#!,$+%)%+,)0"%&)(+0/%
&)(%.,5)0"60#0+,)5
+)(#0"+%%+,&0"?
@+',)!$*ABCCD8*
)(20%)5+",%+,))(%/"&)(0"!+,+,'4!0+E RSTUTVWXYZ[\]^_àbc[def[gg\c\dX
!,.%*?!0'4!$/%&=/$5+,5+.!))(!))(0 ^_dh\dX[_dbif\]jc[aX[h\kdbil][]
!%!'0",)!",')(#!,$+%)%+,)(/%!'&! m\jnd[oY\]
%#.+25)0"%+,.)(4.)0%&0)()0"%/$5
F#+,)G+,)(%!"5+0.)+,*(%,%0H%.+,)+%)
"/%)-.!0&/$),)I/%)-$+,5$H.!$./$!)=%+,. p+%4+1)!$*6KLLB8!,5qH"!,,6KLLN!8&/,5)(!)
$!0'4!$/%&=!0#%%+-$(,)(0!0$!0','E $$E)0!+,5+,5#,5,)#!,$%6+,)5+&&0,)./,E
!)+4$!5+,'%,)(20%)5+",%+,!%$$!%$!0' )0+%Ar0!H!,5s0+)!+,A!,5+,)(%!"/,+40%+)H
#%+)+4,%*(/%#0+0).!$./$!)+,'=,%(/$5 %))+,'A0%#.)+4$H8'!440H."#!0!-$0%/$)%*;
!$!H%#$))(<=;&0!.(!))0+-/)*J$$!'$+)!$* %)/5H-H?),')!$*6BCCB8,)(4!$/!)+,&
6KLLM80#0)5=4!$/%0!,'+,'&0"!-/)C*N) 0!,'I/+.%-H)+,5#,5,)$H(+'($H)0!+,5#!,E
B*O&0!#!,$4!$/!)+,'>)!$+!,50HE./05%!/%!'%* $%6,#!,$/%5+,5+4+5/!$I/5'",)%!,5)()(0
=!0-,$$)!$*6BCCP8&/,5=4!$/%0!,'+,'&0" .0!)5!.,%,%/%4!$/!)+,8%(5)(!))(0%/$)%
C*NM)N*M&0!#!,$4!$/!)+,'Q#!,+%("!,5!0+, &0")(5+&&0,)#!,$%0."#!0!-$*J0!1)!$*
I/+.%* 6BCCP84!$/!)5)(5%.0+#)+4%,%0H!,!$H%%&
0123 45675879
9
8
!" *(!+!)**$! (
#$!"!%&' *1*1D(**$!2
()'*(++, + $+!$+
-!())!*$! *(2$1*2
*./01(!1(2 !/0(+ 1(
**11* !$1+(((!!!
1++!!/3 SH(%R*!6!/"SUU.L+
(4555(*11 %+!/"GHHV.SS(%+!/"
+!" ((()/55536(! GHHU.S+%N(!/"GHH./0*2
1+(5557899:;<=>?8@!() 11*1+ )**$!
!)()*(**+2 !1*1D(**$!2
!(!/AB!*(C6D( +(*!)((!VSG(
***!)()! )**$!/O!E$)
)1!/3*+1* 1+ (((!!!S%N(!/"GHH.
**$!1+B!"& G%R*!6!/"SUU.GV%+!/"
'",-!!*)! GHHV.V%+!/"GHHU./*
!()E"$!2 !6$!(
)()*$*+!(!%F6!/" (!*!* ((!6)**$
GHHI./JK*!/%GHHL.M$)! !(*!/B*!
!%'"*B!"N*" 6**M(+ 1(
O()!.$!(P(+!6* +*$!**(*/
1(!!!)M*!*+ R*!6!/%SUU.6!(
+)*"!()+112 M$+"N((
*(1(/Q$"R*P!/ M$#+1*"+
%GHHG.1(1+*1SGP( !/%GHHU"GHHV.(+#+(+ 1M$
!$!()*+! +1J()(O+#+(+
+!*61D()+ !+ 1#+1'+$!"
1(!/'+!!(*+) *$!/
*$!!(*(+ B!*!*(!$)(*!2
!6( #%>.$(!!1+*
2*+!(!%T(!/"SUUV./0(" *(!*$!(
*)"+ED(!
*$!(!!"*$! > G*+%SG.?%SH/S.
(!/0(+(!(*+2
)*$!1+11*("()
11)*D("11$*(!)2
"*/"!!!!**!((! *+ (+ 1*+++( 2
+!!)!)!/ *M*!!*
S(+ 1+( *M*!
!*S
WXYZ[\]^\_^àbà_cdecdfd G(+ 1+( *M*!
!*G
ghijigklmnopqqrmstquvmqwqmxn ? (+ 1(*
yruztun{mul|nlquw}x~pqqrmstqu 0>$!(*)1+ H%(*!2
|nqunlmqmul *!*.S%1*)+
!*./B ())1+
(!*D(((!! ( !>$!(!H/G(!
" ((! 1(* (11(!/+!/%GHHV.
!!/0*D($!$+!) 1(+>1H/IS%$)IS1+
012 34 56789

67
7
!"#$%& IJKLKMNOPQRSQTUVWTXYOVOWOUZO[OS\T]
'(!)*+,-%!'(!)*+./+0"123!
!4&!+5-**6$)!&'(!)*+76 H" ^>! )!& _)# &"! 5`!=
!'(!)*+-8+
9! " &: " 32! 32!%.6/a$)#&
#!);#5<)$ !& "&! = # "#"!"&>+H"
> +?%!"<)!) # 2!)))##" ")=
.@A 5!"! ! #!"& #"!+H"b&92+.*+@
#!! .@A !)"= !+H") # &##
$";+B#"= 5!c")! # )#&
;)!&!"&<)5>2 &&!)d#$+
#
H">=
#!#!$!)#"!= ! ! ")#+H"
&#&2+4!!#= #!" ")#)!#c &"
" !# !=!"=&! #2> "! " #!#&! !

&!)"!#C#5D#%-**/E #"!>592+.*+a$+
F2"%-**@$+4!!#; `!=3!32!5.6/a$#!#"
#!2# +0 "# "!2= ">!=)!&=)##&!> "=
G !#!2# !) "& <) #!###!)"!#>
"! "!# " +e!>%f!`5.66,$
#!&+
45-**6$%!!"" )#!!+"H"
"&"!!##=
" #!&#!);##> ") " "#";!
"#!>!#>+0 " "!d#>!!) # &!2&)c#!"
4&!+5-**6$)! " "!&&= &2))"# >!>#!&!!
"!&"! " "#! #!#!!b&
+B &!)&2))=
""!
);#&"!%!22;= #"!2!!#
&;& )!&! " &"">2!##"2 %#!& !
")=)=
&!!#%& !&!#+0 25`Gf#"&%.6a@$+H"
!"!"#";&#&!)= !!#!1!!+5-**-$"b>#"2
)#"#!>!#>+H" "))#!)!!%&%:
"!)"+ !2+H">!)!&)#
ghijkljm 5nop$qb&!)
^>!)!&_)#
#5rossot$D"#
!!)
>!=)!&=)#
+
0123 45675879
9
8
!"!# !L''(!$B
!$!%&&'() !!&(K!E|"
'B")P &""
*+,-,./01203415678471490:23;84<14=2 ( !&'!&(B
&O&"!B%
>21?9@ & (&)P!"
(! &&&'"&!
A'& &(&"BC&!DEFGHI(% && !#'!&&)AB &!
&!B!!'!"!# !# (!B"&!
!&("!!'( !& "

!&B!'}(K
&!#)
!'&)J(KB$!L &
L!B!#!#!!B%
#&!#!'(!)M!!! _àbaca~rhnogldnhifgfjrnrfgtinf
'N!!B!&!!!&(!L dsrtunv
!"(&!"B
!'O!# &! '(! A'&&"&!A'NP!!B(
#"&"B(! ! DM(GI)}B$&(!(%
''')AP!!BQ!R( &'!&"#!'(!
'&!!&(!! !(% (!!&!#!!%
K&!(&$! !(% B)S!!&!&!&
K&'!)A# &&&! !B&'!!!!!B!'
!"B& !#!!"& !#'(!&O!#!"!)
'O!#(!)A#!!(!&&
" &"& !'O&
!#)S!!!#$"&% _àbacaihrjgffifdrjoitimrfglei
('!$!&&!B !&&&(
'&#B)T !!'& A'& &&"BJ!& ODFI
!B"!& #"&" "#)AB !!&
&(& (&) (!'BO&!'!'&
!!&!'B"'!#!&
*+,-,UV9WX40714909Y:23;84<14=2 !#"#&!#'()P!
Z07[534370@\4W2]^2[712@ &&(!"B$B
'(B&''!
/01203415>21?9@3 ")A!&"(!%
&''!&'
_àbaca_defghijklgmnoponqlrsrtunv "#'()!&&!'!#
''!BB#!!
ARB!'wSKx&DRy$EFFzI "(&&B (&DL&(I)S! (%
{!!'"!!&( O!D(O!I
!& O&
!&'!O!
!B!&'N!!B'&#B)T&"& (K&'!
"BRy(!!& !&&B!#') ((!#
(&!!E|&!&'(!!
D&$$#!$(!&$}$(B$" !$
&B$!$B$$ $"$!&I *+,-,822V?94;289[40
')A&#(B(!&"B
'%&'!!!L$
!#(!B(K&'!! R!#EFG!B!&
B&(&%"(K'!!)A !& ('& &(!LO! !
012 32 456789
95 696
!"#$%&'!(" %&!=AQ%'!R=!%&='%&S=&'&A
)(&*%+#!%&%&%(,-" =$"&/='#&=(%*&/=%'!&'I/=
%".%'&/012345%/012647%/01184 BA=&"&!=&=&%'&!%%&='#&=(/
7%&%%"9)(/01224:/01214.%'&%" %=(/"=("&!=&A=!%&=&/=&>
5%/012;012;< => =&=%&%#==C%*%==)#=)*%#%="
?#&%&'#$====?#&"&#&&" )(=%=!%=#%!%&=B=&!&&)='%%
!*#&(@$*/='=""AA&%"%(%= %%(&&$%=$"&#&")(="*"#%
%&==$#=& !%&=&"=!*"%(#%&==
B=&=!%/*%)#%(="&)=C%> &=%/%*%#%=A=%%*&!A'"
*=&&%="%*$%(B&=%"A #%)%=&/!%&=#&"="&!=Q>
D=&*(=%=!%&=&=%=%&&#& &=#AAR=%)&%P##&=($%&%#&"/&
*%)#%(A=!"#=/?#&=%=%!%> $%&I"=!%&=%=='!=A=D!>
&=%=&&E$"&(%=&=A"&!=* '=$%=&'%=)(=&"&!=:=#AA
='&%&=&%%$"=*%#%==!"#= &=%'A%&#"&&=/==$$
"AA=$%(& ('%(=#/=%&=/&' A#&#&"=*/&%==
('%(*%#%=&%!//&&/=&='> A!"#=&/'&==%)(=%($%I==
#=%==&=&='+%&&%=&%="% &=T&B=&%&=$%&!&&)===
&%#&='&A=!%&=F&$"*& & =#"(&&(%%=&=)*%#%="
"*"#%(%="='&"()#"&=" @$*/=&%'!%"*"#%#&"=
)(=&!!%&=@$*/="*"#%'#&= "&!=Q''F&IR$&'&!&&=
#&=='&&&==($*%#%==!"> %%=K'=%"=%&&('%=#
#=&+%!%&==#&&&E$#?#&= -&$%*&/%%"*"#%!%&=
A='&=*%#%==!"#=&-&$=GH-%" *%#%=&=!"#=&#&=$"&(%=
=:!=#''="&/*%#%=&%!A'" "&!=&:'!%&=&'%(#&*(A$"&!>
"*"#%)=&/#"&=%"%""=& =&%"=&'%(#&'%("&!=&-""=%(
&"#?#A%=#A=&=&=%=&> =='&%=")(=!%&=&'%(="
=%=%='=A=&&A'=!%&=& & A/&=%"%"#*%%=%"'#=*%%=&=%>
"%=%%'%='%=%('%!#%="=#=#& =&=%=?#&I%%(&&A*%%/!!%
A%!"#I$%&=J%K"#&=& '!=%%(&&/'#=>%&&%=)
-%(&&,J$/01LM4J$%"HNI&=#&/O88;4 #&" "%=%A'&=#"&%%%(K"#&
P&I*=%/01104:/0121< #&=& %K"#&=&%%(&& =?#%$&
%%(&&#&#%(!*"&%&&#&!=#A= =&%/C=/%"=%='#=!"%=%'%=>
"%=%A' %"*"#%!%&==$>=> &,A%!%&=A%!%=<=
"'&%&!%B=&!&&)=%*%#&=& )=%%&&#&&!%,J$/01LM< =%>
&#=$='=%="'&&)#==&% =*=?#&%"#&=&%%(&&A
#&#%(==!=%) =@'(=(#&=&/%I%'A
'&="&==%"*%=%A=&=?#&= H%'%&=&(!'/'%Q=&==R$%&
%*"%A!%=%+D!'=&'%()'> %))$*="=%*&=&=%(%=&#&
!="A%&=%"%=&&D!&@$*/=&% ,9%*=K/01LM<BA=*&=""==&)"
&%==')#"%=%"AA=)%= $#"=&==='#=AA=&='%I
A%"*"#%!%&= %)=(A=!%> ='==)"@&#&=&)=#&%!%=%="/
&==#&$"&=%=%*#?#'%==%= #&=&$%&%)=!'&A=&#&=&F
"*"#%'%(%$A%''!=%%(&& !&&&&&%=&&#
A=&%'!C%*'!=&P==%"/ #&=&%%(&&%&%&&A>=&
="&(%=%=#A=*%)#%&'%('%I ="%=%'%=&A'="*"#%!%&=&=%
=!=%=A=&#&A"*"#%C%*=& &&&#&&!% '&='!=%=%&!=A
"A#='!&&)D%'!/%!%&='%( =#&=&%%(&&&=%==%$&=%%(&==
#&=$"Q%'!R="&)=C%*= "='=='&#&")("*"#%!%&=&=%=
A%!"#= &%&A"=#&&$%= %!!%=)'%&#=&%'&&(%==)#=&%&
0123 45675879
9
8 0
!"#"" R@"%'""2 "%/",2$$$%
""$%&"'(""')" <8@9 "''(+'%'""$.
*)"$&"'" $2/ ()($//"
%"+,-""',"). "+2"&+ 1"2
"*)" %",+( " (%"/%+".
%$"&"'"#&"+'"," <:"8/R"+2"/''"
%","'%/$$ R"82/:"+@/ "
%+,$$"%",+'" ("'/("$$+%'".
01"%('2$"&"'"&"''" "$$*("$"''"*
"%2%+"% /# ("('" "'"("'
&"'%'3%,243",243,24 / $"'++
3'$"45 ", ,*""" ""',6+,Q"'6
%",6""+"""" -: "',KL/" "'<GHMH@"-
/$&"''" 7"&* "" <GHH=@%("'$%&.
""""',$%" "$/ "'08-"08-/"+"
%/""'$"'*"%" (,%*,$"(.
(+'%/*'$08- ,%$"+" "&
"',"''/ ,&(. " "&""&&"'08-
&'(""+'$%"5 /2 F,%" <GHHI+@";"" "'<=>>?@"'
'/$$"%"%('+ $" "E&("'"("'/%1
, "+'$$"% ,1(2(&',2
(""$ &,%("(&'. &"'&"+'"',*'
(9 "2+,"''/"3.""4 .)"Q-:';"""'<=>>?@".
&"'" !"&"("'%", "'',&"+'",&+,("'&.
$,""$"(""& + "'',$08-"""($"
"%""'"(("* "%("' "(&""',*,
&'%%",(&" //", $'"%(&"$("'
$$"( $(&""',F/&2/$'"
:$/1"%('$08-";"" "' "%/""+/",""(&
<=>>?@$$$2AB%,"*((<=>>?@ ("'
&"'""%2":(""'<=>>C@$ 08-,$%,/
" "%/"B("'"'$%.
D%""&$" "E&("' !" +(" $
"&$)',",%<F,%" 2 "''"%":'<GHMI@(+'
GHHI+JKL/" "'2GHMH@: % )%(" $08-/(&.
08-(" +" (, ,(9""&"'%("
%%",%(&% " ( 08-%&"'(&!0
<A'',2GHNNJO,'"P2GHC>JD/""'2 1"%('2" "$2S"'<GHMH@
GHMG@"'+%('%5 %(' %("08-(&""','
%2(,%"/",$'. 5%"+$08-!
%'('(&%$%("' ""$(("'%"(($(("
"$%(""%($+5 -("'%"("$!'," %"B
%("'"(/+ """"'"+""
""($<"@<Q"2GHHI@* / "08-"''/" $(("'%"(
""%""+"((""'" ""&,%'""+,""'
"""" +$% """. %"((!2"%'%"'"'
& 1"*/+ "" $!2(("'%(""',
""+"',""/""(". $(&""2"(("'%(""',
'"B+//+<:" $%"+""',<L'%"F,%" 2
010 23 456789
95 696
!"#$%&!' )#.*U1/+*7Y+$1)6/0.#(#
(&)*)+,)-+)+*$)&.)&/0&+*# 4>::[')#$/+)7+)&*&$$)/3&$#,7
1#$1$#1+)1)/0/+*#1&)$$#1/1+2 )$%&/.+!#*+#)(&
)$/),/+*$33#1)/%#/*+1))0$&/&# 1+71/*&)*)$1/$$%
*)1#$$.1&$-+)45$,#*678 ,/+&*$+/3&)*&#)+)//.$*
67!,9#)/.$1&:$,,* 6/0.#00/+*)#$/+)7-+)$/0&&#&)
67!,'(&##$*$1$/)&;<=$) +,#)#)+)$,0$&#/*+1)&#/#
1,/3#.$%#%*$33#1)/%#/*+1) 3$%+$%
,+$))))+11))3+/#)+,*$33#1)(&) $$#/3#1&/$1#/P$%&*#.2
)+,*$33#1)#/#)$7*#$*+)$% +*,7%#$^*=#/1#+))7)$)0&#1&
#$**)1#$$.4<#$)/. >::: $)1/##)/*)/*#$Q*1/)2
$2?.>::!' )+)1/P%+#$/$)#/*+1*V3&))/#7)1$$)
0)/.+&$*$.$*+$)_)#3/#1
&*#7^*,7/2077)$)/3.#$1
@ABCBCDEFGHIJKLEMNO 4$331`#/*+1'3/#1&$*$.$*+$)
a$$*&&#$%*#/#2
#$1*)&/+*$*7,7^*,7/##$1
;)&#/P$%0)$.*$>:::,7$33# ))+1&)&;#$*)&/0.#&;#$*
(&1&$-+1/,$)$*$.$*+$)./1,+2 7)$)*/)/&*#$1$/)*&+)&
#7%#$/&#/+%&3#1&/$1#/P$%3//0* $*$.$*+$))_*#+)+77^*,72
,7#$%/3&)$+/+)7#)*0&/ 7)$)/3.#$1V)&/+*,/*&*/$%&$)$)
#/*+1)3/#1&#$,+=#//)/3&$)1&2 1/##7/&/#$7))+$//37)$)/3
$-+$)$)&&$))1&/)3/#&;)& .#$1,+S$#/+*$33#4>::>'3&
#/P&./,))/#7.+$/Q#)*R/# &#)+)0#*-+/.+$)##/2
#/*+1Q#)4S$#/+*$33#>::>T)/ *+1$,$$7*#$,+*$)1#$$,$$7$1T)/
>::U'V&P#))))$/&$))#1$. 4>::U'*$*/70/#$1$/)&71/+*+)
&$##/*+1)*#)*/$*$.$*+7 &#_)1/##$/1/3P1$/*#$
%#))/#7*)1#$/#)&*$33#$/% 1&$*$.$*+$)_)##/*+1$,$$7
&#/*+1)(&7#)/$)#+1*/./$*&*/$1 W11/#*$%/S$#/+*$33#4>::>'&
#)W&)))$/&$))#)&/0//* *.%)/3X)&#/P$%#4'$))*,1+)
#$,+$)*#)*/+*4***R/#)+,2 &#$)/#$$%/3&$))*)/#/*+1)
#1'&$#/0$*$.$*+$))$3&70*/W #.+*)$+/+)7&+)&##$1$/)
&Q)))$/&$))#&0&/#/*+1 /7#-+$#&#)))$/)*4,'&&1&$-+
)3/#1&/3&$#$*$.$*+#$,+$))($)# /0)3/#*$.#)$7$$)/$/3.$0)$1
+)+7/0*W)+,)-+)))$/)&#$% 1&$)+))&$)R&#/0$#$^*./1,+2
#/1))$)#*V$)#3#,/*/)&# #76/0.#&$)#-+$#)&&$))$&##
#$1$/)4S#+*$33#>::>',+)/ Q#)$))/#7*)1#$$//##Q#)$#)
+&/#)&.*/X)&#/P$%0$&/7*+$1 /3/0*%*Q#$10$&&#/*+112
#$%)4T)/>::U' %/#7W11/#*$%/&,/.+&/#)*$)*.%)/3
W#/*+1)#)#.*/&$)))$+2 &1&$-+#&4'#/*+1)+),.$2
/+)7&$)))#)/,0&&+,#)/3 ,)$+/+)7*&+)$1/+*#/,,7/,
))##$.7/0;/#QS#+* +)*3/#)&32$3)$%4,'&$))&./,
$33#4>:::')#.*U)#0,##77/%+#)/ Q#)41'$##$//3))/#71&#1#$)$1)$)
&$#S$#/+*$33#4>::>')#.*! *$3P1+)$1&#)#$*$/)71#$1/1&2
Y)S#+*$33#4>::!')#.*$&#U $)4&$)$))$$#/3#1&/$1#/P$%'*4*'&
)#0,##77/%+#)/#8#$1/2X./#*3#)&1&)) #$//%71/,+)*,7*$33#
T)/4>::U')#.*>*#7/#)+)%) </#$)/),0X)&#/P$%*%#$1
Z1&#4>::[')#.*>:Y$)\))/+* *)1#$$.7)$)&.,+,$)&*S$#/+*
4>::')#.*80&,#*)*]#/)4>::' $33#4>::>'3/+*&&0/#/1*+#)
0121314516 789

.6JBK6. %6.K (6QBCDO/

!"#
2 $
% @

$%
$ &
% .6P 2"@J6/D6*UIH*
6N6(6Y6F6A%B6'6

' ()*++*, N$6P
#%%!
- 1RB ( 6J-K*++/0
% #&
$ $%%
#
- E S@ % 0 T%
$
- .%6' 6CO6GU+IG/G
.)*++/, V6M6*F 6N@6K%6JB6K6P@
++*(%!%

$
$
%
B #%_- J

.0 6 T K
%6*G63CDI3HH
@ 6("W` NJ@CDUH&-

#%01
J0 (6*O63+3I3+D
2 ' ( N6(W (a6M.CDG+0$

R
)*++3, %45556 " # @
$
#0 %636

#- N
H+OIHCC
$
6 %

6W 6'.16061 16'*++O@

"6 %(
R@ %-
!%@
06 &% JP
2F6C36C+3ICCG
#$2

N
1
6W6.6J6M6J6'6N
6'*++DN
$%

$%

%
%
R@ !%#
$
5557 0 T%
E
. %E
6 RC+C+CUb !*++D

N#6M6QV! 6M N#6Q*++/(%
+O ++C
&
8 %(
(#

0 T% E6CO6
H*DIH3C
9:;:<:=>:? N6J0CDG/
$%@ $
0 F6/6CI3+
N6J0CDU/
$%@ $
@ 6ABCD/+0
%- 0 F6/6CI3+
E%F$ 6//6CGHIC/+ N#6WQ16. 6JF (6W@CDD*
@
6JE6K 6KL F61M*++/ 0$" 6 "%

'
$%%
%

$ #%%
-
#%#-
@%

$%J0 (6G/6
N@6GDG6*HOI*3/ G*CIG*H
@6P6KQ$6B6K' 6' .6QCDD/ N$6P1 M%6.CDDU@(K M&1#%
(%% $RN
'
$@%@(%
" S0 T% E6O6CI/ K6E

@ 6P 26@@CDOUV N$6P1 M"6QBCD/GK
-

!%%QRJFE %&

) ,6(E 0 FW$ !J&( 6U6CDIHC
@
(6M 6XY6
*HHI*G3 N$6P1 M 6BCDOU

-
@22@)@22"@,K6CDDH #

J(%( 6
0$E@%R( E C6*C/I*HU
@22@6'$6N7 N6K@626Y6W 6.J Y6KN*+++
@6K@6E#6FK F6WFCDUG '$
$#%

E
(%W$0 @ 6A J0 T%6*H6/HIOU
Z"6AZ N62P N&6PKCDG/W&
'
.[6E6E\VW 6@#6K6K]6Y6.$ 2%6A Z"
F6M6(6K066E6M%6( N$6W6E6@ E6JF*+++
M$%6E*++/@% '
#%
-
$ ^"&Q %
RN
$
-
'%J6C/6CCHDICC3/ W
0 F %6*CC6
.6JBK6.6P@ (6QBCDOU UGI/C
0$
%R '61 (6J-K*++*@

@22"@J6/O6G+IGG C3V #%$
!"
012 34 56789

67
7
!"#$%&'()**+, 'C&'""!!*$D(
-.+/-01$ D*!EPW?()DF.11.-/0.$
2&%$)''%34$.555$6!'(!7#' RX$X$GN-,$?!:C!*"C:@
8&*9!!:**;'!'!<!=3 2C"""*'=!!:C&
>::&!$?@A$B$7!&)$CDE!$F *8&$%&')!:)&!.G-,ZGN-$
G0?):"!&'H'*D!IG1 R&7$R&T$KO$GNN-$D''*'"
J&"B*"K!*H;B!!$ !*<"'*""&!
JBKHL6A""$..M/..+$ =&>'!"!$()**D*:
2)$B!J$7"B$GNN+$B* ?1-M/N5$
":!*!!:>&!'**;' R= %$ B$ GN,M$ R; "*&!! :!!$
?:3*&!&!$4)**1..M/0M$ 7!:*91500/M5$
2KC2$GNN+$D:* C"#$( R=%$B$2Q9!&!R$L$.551$7*&!!7>!$
D*:M5MM/+5$ H<'6C!:7!!P=V9$
2'J$LO$PCJ$)9C!J$ R*4$.55,$B!"*O:!!7**
.55M$O""'*"*&"K$ )*J$B"XIBKBP=V9$
%&')!:)&!.5.G,/..,$ XK$T$GNM-$(C!&:""*!4*B*9
2Q9!&!R$GNNM$O!!!!"*!!&!$(S&: B":?*$?@O$2$T?*$E$F(C
7'*+,/GG$ K!*(O**"*$KP=V9PV
2:T$4$O9A$B>!J$.55N$O ""$..1/.15$
<*'!!::!!'<&""*'" X!%$J$A2$O$)P$($GN-0$D!!3
"&*!$%&')!:)&!.1M-G/+55$ 9@B'*!&"'!!!!:
294$O$BCR$U$.550$(C<*!$ !*"!$LC[G5.G5-/GG+$
B"!CKC=!'()**()': X>)$GNN1$T!"9&'
G00/15$ C$(7&*2!-00010N151011
294$O$V!4$2$R7$2$2&:B$4$ 1+/1N$
)J$4$D&>&K$7$2!D$4$.55M$ XA$O$.55M$7*"'UC(&:O!
B"!'''*!>=<*!'!*"3 K[JC&'2!*"CO:!!7
C:!!'B*! C?P= 6!O*!!7&*U!&!O*!!O>&)C$4)
W6)!'O*$? D!!6C!:'B'2C!$
2:%&GM1,0/1-0$ X!!2$4$B>!J$?U$GNN0$2'
294$O$%!U$)$K&!!D$XK$R !!: C>&!'&*!$%&'
7$2$.55,$B"!'<*!'!*"C:!! )!:)&!-G.G/G0.$
'=:!:"!P=W6$)$O$ X:X$GNN1$O*"!'!*"C:!!'
%&')!:)&!..100/1M.$ C>:=":"!$%&'
J!4$O$GNNG$2!*"C *8&! )!:)&!N.G/0.$
:>;$?@X$D$T=!!L$7$AE!$F X:X$GNN1>$O *"!''**"#
)!:)**D:O""*!(!$4* &!!*'C!"!$%&'
299P=V9PV""$0G,/00-$ )!:)&!N11M/1M0$
J%$X:X$GNNN$7*"&"!'"3 X:X$X2$T$B''4$O$GNN0$4!&
"!'*>>C!*>: ' C>:!!:!*"C*8&!$?@B$3D$X
'3**"#(S&:7'*G5 B$X$4:E!$F(C4!&$7*!'
.GN/.., ?!&'(D*!!L!*):"!&P=
(:B$%$B!&%$B$)**7$T!!*C?$.55+$ H!TOB"+""$GG0/G0G%&GN/.5GNN0$
6!''>*9*>&*' X:X$P>O$B$GN-,$2!*"C:!!
="!$(S&:7'*G,.++/.,+$ '**B>!&C=!B'$
(:B$%$B!&%$B$T!!*C?$.55,$(>*9 O*%&'J:U*&&0-1G/11$
*>@O'!*"C"!$( X!2$ P\!D$GNN1$O "**8&'
S&:7'*-0.G/0.-$ !!!!''*!!'9!$%&'
(Y!$4$L$%YQD$.55,$6!*!!$ B*!--G/N0$
?2:%&G,G.N-/G0GG$ X2$B&*%$GN-1$K''>*9*:!;
R!P$GNN1$D":""*$?@X$%$X$ !&&!!*8&!&;'
4*(2$4$X$D!E!$F4!&'( 3#*!$%&'J<"7!:*:
7'*!$L*9O**7'!!R!= G5-0/N1$
""$MG/,+$ ?!K$H]4:4$GNNG$6!'&"!!
R>&O$2*B$J''B$R&)$ #!!:**!*!'!*"C:!!@
6"?$B2$2&>!2$.55,$2# O!"*!'**"'$%&'()**M+
!!: !*"!@D=!= &!>! -0-/-1.$
:$ ( S&: 7'* G- %!2$D9?$K;X$KX$.55N$)3
.+M/.,1$ *&C*&:Q&*'>""!@O"3
R>%$4$X:X$GNNM$B"!''&!3 ':C'*'!!:"'*$J&"(
!:!!C!!*"C:!! K!*D*:..NMG/+G$
0121314516 788
9

!" F

$
$% 9$#!9)#%"!
%

!
#$ $
$% %$$&$
' $ 99

E
)$
$ %$$))
$N

(#$)

!)#$$
$)
%$$ 2
#@# %
$9 :9
2
&')*
!
#(9 + ,

$ @#
!$%><.>-
-.-/ 6
19
15%!
NM 9T 2-//73
9
$0+12
393 4

5678$
9)
!%
)
$$ $#$
"
)

!)) 9

$!
)

#
$
" $ T9
9U2O% 0

$%))

$ )
$$
!
)
;-.7
9 #
5

2 :9
;<.- 6 9B@#!F
$,3,$@@#!F
$I @

=
%:>>?#
0$%#9% 0
$ @ $$ B-//-1# 9
$$
)

#9#%
+ +
A+A
$
$%)
$$ 9

$)
%$$
=!
%
5B ?
))
4-//<8
$ 9
!" !
#$9 5
$%5
$;-.

$$)

) %)$$ 9
"#
)* 9 6 @19
$B5 5) E5-//I
$ '"
9

$) $%"C"))%#@#
!

$ !
)
$$2O%
5
$
$-7>-;.>< 0

-/->.</<
=D"3
=3
$EF
9
9=GH6D 9"6DE1 12
393/:$$
$$!
#
$
$%))

$
0))I-//;?
$ 9)
$$
%#@:0 + M

$77;.<
!
# )*
%$$9 @#
!!
$ 1#$ @B$6 12
393-///I

---7.--
F $
$% 9K:)
! %$%9
=$
5J>?#
$

$ $ $
$% 5
$%5
$>.<>
%$$J
5

?
# 9%<->.-;- 1$#M9 =F%50775
$%!
$
!

=$
5J1
19 2
9$ B47>?$
$
N
F%
#
)* 99 5
$%
$K?#
')*
!
#:!
J
J$ 5
$<<.7/
:$$ 9 ;;<.< 1
$3 1
$1-//1
: %$$
=E,;>J#L$ J#
M! O%K: J

%@##
$
8=
1
#9%@$ +0
+
A)-// 1 = 6
9B-//>:$$
$$ 9#
)
!

= $%42,
1:@
B,
663
N $
$%)
K0
$! 9 99
=J 5
$?3-//?#

)!

N" @#
!!
$>.
E

9
F
$%F
$&
!!
(9 1$$
3/7 K94
&
(2!O $
5
$%5
$-.>< 2
#I 6
4 @4$ 1:
6FE?0 5#!1=7>:

$#
"
1EE#
M 9"Q@!)$

$ 9 $2?
# 9%<>;.< ! $!
$$!)
)
$
!
$
6 9 507<?#
)) )$
$$$$
$)
%$$ #
!9 5
$%
$
$%)* 9 K::J!$ M=: 5
$-<>.-7
&
$(5
$%O% 2$ 4

9
$K,
* 1@$
%605% 1 :#
$50
1
$
!
@ I$3@##
$
8= ,
$)
%$$
R%
N)
$

! 9
))7.> ))
$F%@# %!9 5
$%
6$ "0
$+1 0 9F M1;7,
$)
5
$.;
%$$ #
$
$%))

$ F

9
$ 9
1,
I3 "4
55#
# I1 ,
# %
9$$
$% #
$

$9 2 @1-//,

)!

9)
$$ $
$%
5

<;.> F%#

$

$
F

"
6$$
+,

96 %4 1# @-//7 9
$$ 9
$)
%$$9 5
$%5
$
4
)
N
K@
$F

$!
->.>
$
$%#
E $ 9'$#)*
9 1I 9:@95 ?#!$ ,137?#
@

5

7<<.< ))
"#
)* 9!
#$
$"
6
$$3? @9 @9<5
! $ $ " 9
#
$
$%#
$$ #
9
9
% K==#+5E 3P9&
$( 5
$%5
$<-;.-7
P ?$
Q0
9$#
#
1I 9:3
I:
!
69 6

5%!)$! P ?$
#
-;#9)
$
0-//-0*
%
$ 9 $
$%)
$K1
$ 9
5%!)$! ?$
5!
5) 9
"B
9?% )
)
!
2 O% 0

<
))-77.-- 7./
6
#I9 +F
:@7@!)$ F

$$ 1?9
1@=
#
33 5F
9
197

R
F%!
"
$%)

@#
! @!)$ $
$%
!
#$$$
$$"
5
$
$<<.<> 9
%F
'9 25

>
6
9 @#!F
$,-//;:
N '
$)" -;7.-7<

%$$ 9

9 5
$%5
$-- 1
91@
@B @4?-//5
$%
-.< 1
9:= 5
3-:$$
$$ 9#
R% '"
6
2A1
B164 R
9 1

)*
%$$:!
J
J$:$$
91-//<:
))#
N!
#

$#)$ 9 77.
012 34 56789

67
7
!"#$$#%&'(()*#+# ,#++4G#JL+0#$J#/+3 &'((=
#$,+-#-./0#%.. 12.+%2//,+-3. T*## 1+/;#$87Z##$2#*!7.-%
4%5#/67879 +689 ?@+/$+/;0....+. $U#%#$,+1+/
+:#$2;3#6+&"<=><2.+%?+#@+@ $R<(<(<CSV1 $p#'((=(>(('
+. 1.#-A$//+#3+# 1 $2//BC ,+."L<=>KJ.!#$0#%.. 1T?@+3.
D==EB('B(= #+/J66+879 +689
.67AF"<=>>0@@$2.+%0#%.. 1 $. ,!!"<=>C2#./#,+/$+. $".#+/;
G .H#$HH&"&5 /#"#: T.*+0@@$2//: $Mg
IA0<=>CJ*@3#$#@@/# 1?+ ,!!"#$4#.,<=='0 /3@#+. 1
+1+/./#.+#12.+%2$.<KKE>D 1+/;/@+N!#$;+@++%!+$3;$;
IA0#$&*LG<==>M*+.#@+$/#$ Q#*+@+N! 1/$++# 12.+%2$.)
#+-.@/N/./#!#$;$*@3 1/33 <DDE<BC
?/.$./+@*##%..+# 12.+%2$. ,#8&;#-#&#$2/;/;,'(() $!;*
<DO<PK)E)K 1;@+1+3#/ 1#..+%@#R03?$3 $
++#%#$J#;%&'(((2/ 1.# #$/+/;#+#@@+#/;+# 12.+%2$.''
$#+$.+1+/+3.#&;$$#+/;.Q#*+#!#! '<'E'B<
+#12.+%2$.<K<)=E<== ,7+.<=>>&++@+#//.#$#@@/# 1$./+@
++#%J#;%&#$5#?+H0'((< *3;$.HR",!!O$P2.+%0#%.. 1
J./+@*##%..R,#.@+.#$1+ $".#+/; $.T.*+0@@$2//: $Mg
H+##DBBC<EB)< "#11.@+!+T:,+%#3J"#$4 $g"<=KC
+#!;O'((KP& ++.@$/0#%..#$J##& $! J*@3 1#./#1+!+#$!!;..$+..
7;#*##$"&;#@3##$F#S&"&879 +689 -1 $Y/;!%<(C')ECD(
8#+#&,#+.0#$"$T'((D+/;/#$ "#%5<=>CH3@+#/ 1+1+/.#$#+$.+#!
/*#@+N! 1/33+/#-#/6N+/11. @#..+#12.+%2$.<<B=E<KB
?@+/.3+@+/@. 1/;#+#/+ $U#%#$ "#$#F<=>=Y;##%.. 1..+%$##-%!+#A$
,+1+/<KD<EB< #+3 $.53+/.+#D)><E)=D
8$V&W6+XJF#.+2Y./;#0#+#;Z "#.:X!+:J1+T#$:-/p0'((C
#$L+#+$W'(()J+3# 1+*#..+% "#.-7;67;7 1.3#./#1#/.
@+@+. 1@$$%#3.O[\]^_]àb.@@P.!#/#% #$;..+%$*+.% 1+#$#$+%.#.#!.1+3
-#.$$./+@*##%..3;$!% $U#%#$ ;#..1&+#+#/T+@# $".#+/;#$
,+1+/<>BK(EBK= Y/;!%'''K>(EK>=
W#-0#$:#7..FY'((>,+../ 1/?11/. "*#0&;#3-+.JF#$T.*T'((KJ*@!
#1++#!#$7;.%+1+/.+# 1 $ #?/1+;Q#*+$./+@ 1+/;/;.. $
2//)D2<>KE2<>= U#%#$,+1+/<CK<)EK')
8 !+#Y++.FY&&+&Y+%.#3L#$ ".*6T& 72/T7#0#$:%JF<=='
F ..@'((>M.!;+1+$./+@*. *#+###%.. 1/*#@+N!$##R0/3
.+%##%..R+3#+#!#$17@ 1##. @#+. 1#5+.;#$#8 +7!#+#$@#+#
.@# *+;7-+# 12.+%2$.'< 12.+%2$.)=)E<<>
<>(E'(' "03#G/;2 +#,.:q@AY#3#3.
Wc#;%<=>C2.+%T*## 1 $#+/ T#$5?#$+#.2'((>2.+%/;#+#/+A# 1$+%
J66+879 +689 !.7;$11+*#@+N.+#1 $2//
d3e+dA-6,#$J+/;0'((>J*@!;?## )D2'>CE2'=D
#.# $++1+/.#$#+$1+..+%##%.. 1$+6! "$!g,'((B:?/1+2.+%0+-."#!
7#+4#+".#+/;B''K=>E'C(B Y?+ #$ 0@@#+#/02Y&J"W4.
W,f,<==<W3#.,+#$,#++.5G,.-.<B<C) & .;;/6,0
DK(>2LM+/;Y;8;+#$. "%0#$:!;4<=)(Y;$%#$!+$R03 $N/#
W+.6*/;J&g5,#$2;+#$4<==< 1+@++%!+$/;p5+.;+# 1,.%/;!%
,+/+..##%..#$.#@@/#. 1+/;/#$ <<)D'DED')
;+..+%@+N!HRFY:#7..#$5,g 2#T*0,#rg +#T#$#+8'((BH3@#/ 1
O$.P2.+%2//Y;+%#$0@@/#. $. ; 1#/+%p#%#$/;3/#/3@?% 1;Q#*+
#+/J66+879 +689@@D<)EDD> !#! ;?+ 171#.++$%!+.#....$
W+.6*/;J&g5,#$ 2;+#$4 -%;+$11+..+%3;$!. $U#%#$
<==(G#+#/.#../#$7;$./+@*##%..#$1+ ,+1+/<KCKKECC>
/;/@+N! 11+#61++.,+.$#HY0# 2/;/;,<=>=0 202SH:@+!+#3 1+!+#A$@+
!0#;3&0<CE'(<==( /+..##%..2TMLHc>=,+/$!. 1;202
W+3-#J63#06L02+$#2 T+@#M.+.L+@H+##& 1+/& !
:&;#3-+.T#$&;#3-+.J'(((H++#.;@. L+3#%#%=E<'<=>=
-7 $./+@*?+@+N..+% @##$ 2F2:!9#$F7#!H'((=0*
$./+@*#+-..+% @#*##. 1-1 3;$ 1$./+@*##%.. ;-+7$/11RY3
h]ij\^^\kl^#$mak\naio\i]^l^3./.#2//KB ./#!$./+@*##%..T+@# $".#+/;#$
D'KEDD' Y/;!%''>=D<E=D>
0121314516 789
!"#$ ( @%$#L#"? (#$

#%#$&'%"#$(#%)*+,) $"(#" $
((8D9K-D8DD
-.//0(#/12($3(45$6
S(SE )
=>DD 2L#" "# $
&%#1$&$'""((%4( 2(A?=>=>8-=>D=>I
778-79
S(SE)
=>K8* $#L#"%#$-
:;#;<=>892 $(#%? @%/13 #?(#%" 2L#"

#"$D8-=K
4?4A$4#5$6$)((#(
S(SE)
=>D7C(# #L#"((##(
2#A#$A?B"##$$C# " $
((I7I8-7I>
/##"#C (A8-D=
S(SE)
<$#)$$HH=>D7
EFFG)$H?H=>>9C4(# 3%# #$$#A( ((
4#$#)CHG
)"#A
#) # #L#"$(#F#M#F:(#%$
=9=>>9 #?#$ #L#"%"#" $
EFFG)$H?H=>>9F2 4(#

((I7>K-97
C4#$##/2)"#A)## 2
C"%*$
K
?
*)"7-D=>>9 %"# ##L#" 9>(($
#E<G$%2:HC=>II "" $B"#?7===-==7=

((A(A#% $((#((/13 2(+O)#$ OT2K
H2 5$6$)((#F#?G%)$ N"S# ?A$# (

(($!J7D7-7KD #AP2$
(($2(A?9D-D>
#E<G%2:HC$#<A P$=>IDL$$#??(?((A1

=>>94(#"AF$$?(#(# 3"F($#?#(A U$ ("#V(#"
##F"#" E#A,(7==8-K 4(#$4?(?(7>9K-9=K
#%

=>D>+$(#AL# ($#? P <=>8C(#$4#A"#(
#(4(#$4?(?(D8=-8D 3##
##A
%(/##"#PA#
#%

$EG=>IC$#?#(A 3CK-98
"(#A#$F?A#"$##4(#$ P))$)$*=>I9) M (
4?(?(K7K>-7I> #?? $$F%A/1)$$
#M#N
#M#)$$O=>I=<"( 5$64A%",(4($A #
# #?A$(*MH$!J 9# P" %",(
?"G%
#O$MH2=>>7C A$"(# )#$2 O#$78-8

(A$#%# ((A?% P))$A
4=>I92 " ((
"# $?$"(#"
?
#"$I A #L# ((#"
8K-K #
(( $$)A("#"7888I-8DI
#H
$+%
P?)$
A# P$N=>>8G (# *A(+A$GL#(#
,C=>K9
?%"#F?Q"###%$(#% #$#
?C (##( NG(
?$2(A?I9DI>779
2 !%#? "C"F+
#H$
$9
?G%"#4(#( P(2W
$ 4"#4>HM F
2 $G$#)($(+$ ("X C ?
#H
$$<Q"#=>IB"###% ("$ L#$ Q"#? $ 4 (
$(#%?C$P$8D9-D79 $1===DY: $Q">D=
"#C(%/

"%A#$/("
PA#J=>>9)##F"#3(%?$4(#"A
=>>>C#%#A($"#$$(
2 C!%#?/#(OMRE
#1G (#A ("
? ZEJ$PC=>KK2 (#$" :"$A

#"$=99DK-9> $(#%$2(A?7=8D-D=
? HH33*F(#*R"AC2

$,=>I8C $("
012345677
8594
652
24
!"!!#"$!%&'##(!)'"!*'!*+,!#!
#(!)'"!*!-!**!!%'&.'*#"$.*&!!)'"!"!&!*#(##*
/*##0!!)!!)&!1"!*!"2!**!&+3!#"$!)'"!0!'"!0!.
!-&&$!,!)'"!4"#-&!5!#*."!*!"2!*(#&&#6!*2$"!&%!&$2"!(
*'##(#""!&#2!6!!"'0!&*!#"$!)'"!0!'"!0!+
789:9;9<=
>?@<?;ABCA
=DB<EA@FGAHE
IG@H89JA@KCBCAL
M/&N##*&!.OPPQ1
0454R 77|78594
65}5~5
SSTS UVWXVYVZ[V\ ++++++++++++++ ]^_ h& 3! /QQ1! !!"&&$
SST] àXbcX\adec[\UabVUVWXV ++++ ]f] !!**!-##(!)'"!!(#&&#6!)
OO+Q+O g'&,!)'"! +++++++++++ ]f] '"!!!#"$*('#&0(!##(!
OO+Q+Q h'*#"$,!)'"! ++++++++++ ]f]
OO+Q+i ,&!,!)'"! +++++++++++ ]fj "''"&.0!&*'"(! "#!"!#((##*
OO+Q+k ,&!l*m!!&++++++++++ ]fn *!!!*"#'!!!#(%#.!".#'
SSTo pV[deUVWXVqVXVrV[ +++++++ ]st *!!+3!!#!##!0 !
OO+i+O ,!)'"!4"#-&!5!#* +++++++ ]st
OO+i+Q u!"3!#"$,!)'"!v%&'#
,!w'!+++++++++++++ ]s] +!)'"!!#"$ "#!"$6#&$2!
OO+i+i "'0!&,!)'"!5!'"!0!
*3!#"$x#""!&#+++++++ ]sj !"!%!***!"2!*2$'0/*0&1
SSTj yd[bXad[ ++++++++++++++++ ]sf *$"'0!&0!'"!0!0'2!"!&!*
zV{VV[V ++++++++++++++++++++ ]sf #!#"$"!#!+
2+!)'"!0'&"0!!""2'!+
+!)'"!*!"%!("#0!"''"!#(!(##*+
*+!)'"!*!!!*2$!%!"&!!+
h'02!"#(!)'"!"!%!6"&!*!)2##
%!2!! '2&!*/#'"!.QQx!.Q.QP
x"!!.OPk'"**5'!&.OPP
&.Qk5!.Qi5##6.OP
#!&.OPPP3!.QQ&#!&+.
Q%g&!!&+.QP1+
l+,+6&!.l+l!$0.9:KB<>;?CG?@AB:BDBB.m##*3!!,!)3!"!.
uO+OPOkkOPkOO.3"!"3!!'!5!*.xQO
012 33 4567895
87
!"#$ "HIJKLM

$
MNM O'"*!A"4#
#!#&PA"#(
%&"'()&'()# Q*
#&#$() *!# '$ >* *!'$#$
#!"!# G* * %'#
# !"#$*+ '%! F,21 0,EF
,-*!') .
% "'#'% ;R' F,3C 0,0F
.!!',! !! -*' 0,FE 0 ,DF
*'##$' ;*'
!!&/.)0123(,. &"$( S!T
!'4
4 * ##'
F,CF
F,16
*'#*'.. #"''$) S!!*##' 0,EF
%' )#$!!!" 4 +
%*'..4 4 4'' $' U%!'
*, Q%)GT
V "
) QW
F,1C
F,16
5''#4.&0163(4*%7 Q')XQ F,CD
#%48%9&"'(#8'9&+ >#!#*;<<.#=
'(#7 ,/'# G#"' ' $* &0130(
$* #!# ) **!A")"' '% *!'#'"'$
*!')4'# !# "'#%! *!4 !'# $**
#*#!':'$&;<<. W ###$+* &%%!
#=)0130(,>##''$)"'' &;<<(#

#4#+ %* E1#
! *'+ W#.)#
0130(
!#,?"'#+ * 3Y#&;* ##4#+ %
<<.#='$)01CD(
' !#"+
#!$ #!#
!, *'$%"*!*B#
-#*'$*! ! !#,Z*!' %%$&
*,@)'.'#A")+ !(!!)%&#(.)#4'#
:'$#$* # &$('$ .,G#)
# $) ##:'$, !# *! !+
;<<.#=&0130(#* !'$!#,Z*!'
' #*B4 , #$)#)#*$"%',T #',
#"#'%*' 4* &EFF2( #EF[ #!'$
4 #'4 4 '#*,G ''#)
*' ,>')**!'$#$ !"# *'!
*#*!$!'#'"'$**!+ !##*!' ':# 4'"'$
%'!!'&;<<. '4" 4# #,
#='$)01CD(,;<<.&EFFE( 4#)!')
*#"*B! #+ #!#*'*!,>*'
8!!''.'''# %*#!)!#4 :)
!'#*,;%$)%**$'*$# #%# # *' !!<%
4!#'*! #!' %;'$ .)-)#
#%%%%.%,9'00,0 . 4 :)$#4
# '"*!*!'# 4*' '''#,;<<.
" A" 4# "$ #=&012Y(*'#%'!!'
#, '#.!*#4%'+
G *#)!"# * #"#'# # 4*',
*!$!#,- \$##?'$&011D(*'#!!'
!##!"#4'#" #$*&**++**%
0010 2345673839
3 0
"%*$H-.55
!"#$"#$" .5:$!.5?
"$%&" ?5$%I$
$$# 1?
'" 40?%J
()(% G-.//5!",<<2
*"$+$ =-./89!>,<<2-./0.!B2%
$#,$$-./00!# -./05!>$+'$$@
"1/$" -*#..%1!%
%,2" K
$$$"%3'# '
456$$"+# %L'>A"%
%346$" &"#G%-1551!M
$###706 '#$DE
+"#74.6 -"!N,">
+##$7896 +" $>$@#"%*
7:.6 +"""% ">>
*& $""'O'>
$$$$ >$$"#
"% "'%P2-./:/!
;,<<2= "#$19%
-./89!$$"$ *'#"
$%*"$" 'I"2"$$
'-'156!$"#> %
>$2>"%*"
$0/>1.1?
#.55"#
04$%*$@# QRSTUVVWX *#$@
-$>>>!>" &>C">N,>
>->>> N,
A!%B2%-./05!,<<2 & C" # ./89 ./0.
=-./89!# YZ[\] L YZ[\] YZ[\]
"$$C""%* L ^_à P YZbcdef
$C" YZbcdef h
^_à gb[df g
^
b
_
[
`
d
a
f g
,
b[df

#-458!&" gb[df i YZ jklf YZ jklf
-0/!%*"##$$ ,2 H h , $
#"#2 YZjklf ,"" h ,
2
$$C"$' M YZjklf , ,
'&$%&> J YZ[\] , M$$
C"'# * YZbcdef L *
"2' J
m _no
"
$

'D' * H
2$C"""%EF> #3 $?5$-G>.//5!
,<<2-./0/>#!" $ C"
$C"%*> ./05!% $/0$-B2%>
2"$ 3

$-"
!$1/$
"% -,<<2>./0.!
G-.//5!,<<2= 3
-./89!#& $04$
-,<<2=>./89!
010 22 34567849
76
!"# A\]]eC1*+0[3*(75-B/)&?A*++G(37'+5I]C'-+23/13'+
!#$# '(+2)*13/,51:@'+9'15+-8.(+3'@5+30*<
g*)&?,1*'350*3&03'.-/+2+/,17.(+5+'(+2)*;
%&'()**+,')-&.+.)//0)*,1**2)*13/4310)'-5643&0 )'@(/+) +,1*'350*,-1/0@+:3&)71/3'+0)&0+7+&0+&'/6-8
&2)*)@/+)&?+*'+0,1*'350*40+h)iE+'3/<A\]]jC
'3,')/+7+5,+7')-&*-8'+9'15+)&:-5+0+'3)/3&0'(+& 8-1&0' (3''(+2)*13/'+9'15+-8'(+2)*)@/+,1*'350*
.+.)//0)*,1**(-.'(+*+&*-56*7+,)3/)*',3&:+3; ,(3&?+0'
*15+'(+*+7+5,+)2+0'+9'15+*)&8--075-01,'*<=(+ '350*<G35*(-+- 53/'+9'15+53')&?*-8'(+)&?+*'+0,1*;
1*13/*+>1+&,+-8'+9'15+7+5,+7')-&.(+&,-&*1:; '-3**+***'53.@'3 &+
+5
/<A\]]\C'53)&+030+*,5)7')2+73&+/
566-?15'*1*)&?2)*13/'+9'15+'+5:*
)&?38--075-01,')*2)*13/+23/13')-&-8'+9'15+ )&,/10)&?*7--&):7
8-//-.+0@60)5+,'A.)'('(+B&?+5*C3&0D-5)&0)5+,' 75-01,')*i+//)B+0+253+/*1*3)'-+&A`'(+0+?5++'-.(),('(+
A.)'(+3')&?1'+&*)/**1,(3*E&)8+48-5E4-5*7--&C'3,; *)-&/+8'3''(+*1583,+38'+5/)8'6/
0@ --E)&?3''(+):75+*;
')/++23/13')-&*8-//-.+0@6-53/F'3,')/+A.)'('(+/)7*4 1&*')55+075-01,'bC3&0*7--&,-2+5)&7?A
)&?3* --&81/85-:'(+
'-&?1+473/3'+4*3/)23C+23/13')-&*<G-&,155+&'.)'('(+ .(),('(+75-01,',-2+5*'(+@3,E-8'(`+* '(+0+?5++'-
7--&+23/;
-53/F'3,')/++23/13')-&A3&0*-:+'):+*3/*-.(+&,1'; 13'+0@6/)8')&?3*7--&81/85-: '(+*3:7
')&?D*'3@@)&?'(+8--0.)'(31'+&*)/C35+3/*-'(+3153/ =(+68-1&0'(3'@-'(*7--&):75+**)-&3&/0* +,17bC<
7--&
A*-1&0C+23/13')-&*-8,51&,(64,5)*764,53,E/64+',< ,-2+5)&?.+5+()?(/6,-55+/3'+0.)'(7+5,+)2+0- 53/
AH)/,3*'4IJJJC< '(),E&+**<=(+2)*,-*)'6-83[1)0,3&@+3**+**+02)*1;
3//6@67-15)&?'(+[1)085-:3,-&'3)&+54@6')/')&?3
,-&'3)&+54-5@6+23/13')&?'(+*75+30)&?-8'(+[1)0
KKLMLKNOPQRSTUVWQXU -&3(-5)c-&'3/*1583,+Ak/+i3/0+3&0H-E)&)4IJJ\l
H)3**+-?/-13&0m(+5:3&4IJenlm(+5:3&4IJ^^C<
o3&(pi+'3/<A\]]qC'53)&+030+*,5)7')2+73&+/'-+23/1;
Y3&6*1583,+,(353,'+5)*'),*-838--075-01,'0-&-' 3'+/-.;83'6-?15'*1*)&?2)*13/'+9'15+3''5)@1'+**1,(
-&/6388+,''(+7+5,+)2+0377+353&,+-8'(+75-01,'@1' 3*?53)&6-&/)03&0,-&')&1-1*[-. 85-: *7--&<
3/*-388+,''(+7+5,+7')-&-8'(+'+9'15+<G-&*1:+5* f++3&0m3'-A\]]IC1*+0373)5+0,-:735)*-&*,3/;
E&-.85-: 75)-5+97+5)+&,+'(3''(+/1:7**++&)& )&?'+,(&)>1+'-2)*13//6+23/13'+'(+7+5,+)2+0'+9'15+
'37)-,37100)&?35+3/*-7+5,+)2+03*/1:7*)&'(+ -85+3/'+9')/+*3:7/+*3*.+//3*7(-'-?537(),):3?+*
:-1'(<Z)*13/'+9'15+3**+**:+&'(3**-:+-2+5/37 -8'(+*3:7/+*<=(+68-1&0'(3''(+75)&,)73/,-:7-;
.)'(377+353&,+,(353,'+5)*'),**1,(3**()&+4?/-**4 &+&'*73,+*0+5)2+0@6'(+'.-:+'(-0*.+5+>1)'+
3&05+[+,'3&,+A0)*,1**+0)&G(37'+5I\C<%&'()* *):)/35<
*+,')-&.+.)//0)*,1**2)*13/'+9'15+&-'5+/3'+0'-
'(+*+377+353&,+'+5:*<=(+*+2)*13/'+9'15+'+5:*
.-1/0)&,/10+5-1?(&+**41&)8-5:)'64*1583,+7-.0+5; KKLMLMrQsOWtXuTUVWQXU
)&+**-5@/--:4-)/)&+**4?5+3*)&+**4[3E)&+**4*'5)&?);
&+**4*:--'(&+**4.)/')&?43&0*1583,+.+'&+**AG(+&4
\]]^C< %&*-:+,3*+*4,-&*1:+5*:36B&0'(3''(+*-1&0*
=(+*1583,+5-1?(&+**-83&-3':+3/-5'(+,--E)+ A310)'-56'+9'15+C3**-,)3'+0.)'(+3')&?38--075-0;
,3&@+3**+**+0@-'(2)*13//63&0'(5-1?(-53/3&0 1,'&+?3')2+/6):73,''(+(+0-&),5+*7-&*+*3**-,)3'+0
(3&0'3,')/++23/13')-&*<=(+@/)*'+5/+2+/-8'-5')//3 .)'('(+75-01,'<v&+93:7/+)*'(+?5)''6*-1&0-8
,()7*.3*3**+**+0@6_51.+5+'3/<A\]]^C.(-8-1&0 *3&03?3)&*''(+'++'(.(+&+3')&?,5+3:+0*7)&3,(
'(3''(+@/)*'+5/+2+/.3*&+?3')2+/65+/3'+0'--53//6 :30+.)'()&30+>13'+/65)&*+0*7)&3,(/+32+*<w&'(+
7+5,+)2+00+&*+&+**-8'(+'-5')//3,()7<%&3@5+30 -'(+5(3&04310)'-56'+9'15+,3&3007-*)')2+/6'-,-&;
,51:@377+353&,+*'106'53)&+073&+/)*'*(32++23/1; *1:+5*+3')&?+&i-6:+&'3*.+//4+93:7/+*35+'(+
3'+0B&+&+**Aàaa2)*13/+*'):3')-&-8'(+3:-1&'-8 ,5)*7*-1&0*3**-,)3'+0.)'(:3&6@5+3E83*',+5+3/*
?3*,+//*bC40+?5++-8(-:-?+&+)'6Aàaa5+8+5*'-'(+ -5'(+,51&,(6*-1&0*3**-,)3'+0.)'(+3')&?3i1),6
0+?5++-81&)8-5:)'6-8'(+7-5+*)c+*bC43&0-5)+&; 377/+<G-&*1:+5*-8'+&1*+*-1&03*3&)&0),3')-&-8
'3')-&Aàaa0+?5++-8-5)+&'3')-&-8'(+,51:@?53)&bC 8--0>13/)'6<Y3&6-81*(32+3//'(1:7+03.3'+5;
Ad-&c3/+c;_355-&3&0_1'/+54\]]e@C<f3**-1+0+'3/< :+/-&'-0+'+5:)&+)'*5)7+&+**A3(-//-.*-1&0)*
0012 3456789
64 977486 2
!"#!$%&!'!(!! +'!)"!*'.+ ))!$
!$)!*+)), 5B+!))3(!)167K6&,M#!1%!#!(
-*!./*!)%!0/)..$*) '.I$)!3(!)567KA&+*!.
#+!)"))1!*+))1!(%)) ), !)")))!*#+!%L%!)"))
2+!%.#!(+)!#)'.3(!) %*)3(!)567K8&%)++
4*!56789&,:%.+!+)'!)*!0 !%L%)))!!.$"!%*0
!)+!#+!#'.;*<!5=>>6&1 !)")),
#!('.:*.3%5=>>9&1?%! 3(!)E))!$56787&)*+)D
%,5=>>7&13!%%,5=>>7&,?*)! )!.+!!))))#+ )*),2+.
"!*'.$+%)*!')#++0! +"%))%*+)$%!.'#"!)
)*#)#++!"!"0+!*0++!! )*)"!*'.!*)+0+ #+"%!)
+!$1)*+)'* !$+@# 52'%66,N&,2+!)*%) +!)*.+
'+')+$%!5;!$1677A&, +!$.'#))!.+!!)))"!0
B!)"C!!*+. ) %%#0!)1 )*)1+)) +)*+%*D
$%.#)!.),?*0!D )) +)*,-)+%*)) +)*)
!)+)#.") )53(!)16787&,E !)+"%))O"!") +))
!)" )1%( !)+ !*)0'%)1!$D !*+))1!)"))1!(%))1)+!"))1'!D
") %0%%)+!*!0*0+#! %))1+!))1)""))%)!),E+
)%'%,F+!#!)1+%%)/! +)*)**)5&+"%))"!
*#!"!))*!0)+%%#%%),2+))*
)!**!)+*))$%!%% .#!D
G%%'%%)$0+!,E++)!**! PQRSTUUVW X)!*)+#+!*''!D"%!)"!*
))!.1'.'!(0!+#01+%%)"" !!)*)
+)"!*)),F!DG%%'%%+""D X ;)!"
"0)*)*+/"%)/") +! M!. 6#+%Y)Y'!.
$"!)))+'%%,E+*!0%%)+ X!)+%!. 6$"*"!"*%!)%(
))*+)*!%) +*!0!"!)D 4%+%!. 6$"
*"!"*%!)%(
$$!)!"%.'%0
)*!,2+$* )"!*)%))#++ # ! !N>)
)*!) +%H*)+0+,I/")0"% B!(! 6#+%?*)+)%!(!
%%))*G$)*!!))+*!0!"!)D Z!""! 6$#0
)*! +%%)!))+"!!)")) [* 6#+%X)+!O)3!0).%"*
+"!*, \0!)" 6#+%]')4!)00!)"
J++!+1/")0!.!)" )1%( X!)+!! B!))#))16$%0
()1!(!)1+")1)$)*!5+*$ 4%+!! B!6),A )#
$#
))16$%0
!&!))+"!!)")) + , 6,A$#1$$!)!"%.
2+)"!*)+!%%)!))*!!*'. '%0#!!6$
'!%%%!.#%%),E++)#%%)!'!D [+" 6#+%[!0%)"+"
(.!$0#%%) !0$))"'( Z Y*^"+" 6#+%[!0%)!*^"+"
+!!0%)+",E++#%%))"'('!D !"%0

%
* )""%

6$#0
)!*)+0!)*#)5)$%! Y"0%
*0 !(&,E++$)*! !.!)" %*)""% 6$#0
)!))1+#%%)!%))%(%.)"'( \!+$!(! 6#+%5$**!!*(#&
+$* )*0!)%)), _%(+% 6)H*! M!)+.O)$%(+%1
3(!)567K6&B+!))3(!)567K6& E!+)* 6#+ %
)+#+!)"))!*+)) )"G ?+!'!( 6#+%%\ : !
)+#!+)*
;5]')
)'!+')) )*%1
+')) !%L%%*)%1!+')) ?+!#+ 6#+%)+!(#+
4! )&
(
$' *!.!%L%!D 5]')4!)&
$,B!)")))$'*)%%.!% -" !$3(!)E))!$56787&
012 33 4567895
87
'&6!64/8998+#;"-4.
!!"# !!"-4,.!!,
$%&!'())*+ ! #A/!" ...-!,
!,-!-. !." -4"!
!,...,.! ./-!!"%-!,
-!-" !!!."!#I
! !/ .!,. !.!4 ! -%. ..-1
%.01"2!'" !-.!!! ..-4!!,/
(3()456+# -!.!4- ..-
7!%'899:+"!" !-.!!.4 ..-4.
!!6 - .,. " 1!! #
1"2! ': 456+/!! " >!4'89<8+" -.,.
!!6-. ! 4 !,"!"-
1!!8#(:3(456/!!4." .#;1J.#'())E+!-
!!6-. 1! %"!2!!.0."!"
.,."-!!#;!!!.. !. %%%-"1
%..". 1! ""% -0.-""!
"2!.8#9456/ . %!!..#A%--!
.,.."1!/ !%"!"#>!4
"2!8#9456/.! '89<*/-+.,.!%"
'&% 5%/89<<=>!4/89<*/-/ !!4.""#&".4!1
89<:+#?%"0!.%!! !!/!!4.!.-
"".%.! -!.,.#A%-
%". ! . %."!.
!!"!!- !"!!4.#5 ,/.3!.1
"4..@ '7!%/899:+# %"."
A%@ -4..-8B) %""%"#I
5$"2!%.0 -;'())9+,-!.!"
--/.C !! .. "..@!
!,-. ." !!"#
.'D/89E8+#F
.,.!,"
!!.%! " KKLMLNOPQRSTUOUVRWXU
%. .%4!.#G
""2! ..-%-
"!! ..-% A!.!-,.3!./
4..@ -#&%../ %".!!!/!.
.,.!, !,/!.!, %1
"!.%! .-@!-'"""-!
" %. .%4 !..YZ+ .#
'H.#/899)+#A%"!
,."!%!.-
%. ..!!! [[\]\^\[_àbcdaefgbhdhifj`h
"#G""2! ..
,.'7!% k.3!.!%...1
.#/8998+# .!%#A.//.%!/
I, "!!!".4 .,//,.,1
%12!"-4/"% !".#;'())9+/H".#
"".!""! '())9+/l.#'())<+/,m..#'())B+/
0012 3456789
64 977486 2
!"#$%&#'(#)*!+,-$,-.' #$(2)0(2/&2.-0#&''-$-,-$#)'#$#&''
//$#)+&''-0"1#'(-&#(-"#$(2//&(' /$'.(#)*"E77"E77FRM-(0'((#)*"
/#)+23'-)03(2+&+(-//$(4(* "F"@!*?0)(#)*F#!/$(2#(
5&&$-0(,#6)-((#)*7!#$(2'(2 /O@'9N&('.-(21#)$(-(-(2&2.-0
'8&/#)(4(+&+(--,),'-0'(- &3&)'(#$37*E&1L+#('/'.#)).-0#0$
93(2)-+'"9-(-093(2/(-&-'!((2"&2.: /1E@(EE*S0#)"-$-,-$#)'+$&-01
-0/2#$/$'93(21)#'".((-0.-(2'#)-,# '#)-,#($$($/.&2.-0&3&)'(#$3
#$;31#(-&9#<$."$/1#(-/'1-:')-$ #+-&/$3(#'(/'.#)).-0?0)(#)*"
/$'9(.(2(0#$2#$+#)#("1#-+: F#!*A2'#(2'#)'/$(2#(9((-0(#'(
)#(-/(2/$-(#9)'93(2(0#$ $&#'$(219/&2.&3&)'+-('.#):
'.#)).-0* ).-0*>$2#$''-'#)'+'-(-,)3&)#($(
=-0-0'(-(2)-+'1#3'-0#)(2#((2/$-' &2.-0)0(2"&2.-0&3&)"#$1'&)#&(-,-(3
'(-&<3"')-13"2#$"0#-3"(&*>4#1+)"?0) #''&-#($.-(2&2.-0>'((#)*"J"Q(&2-0'
(#)*@!2#$(2-+#)-'('#(+&-,$02: (#)*"7RM-(0'((#)*"@!*)-''(((#)*
''#$')-++-''/&'(#$'#$1#3#-''93 @!'2.$(2#(-&#'$'#1+)'-;-(2-&#'
99-0(2(0#0#-'((2-'-$/(2)-+* E"" #0:G#,$A(3:>(-'/1T'()"
A2B'(9-(#)).'(2+&+(-'/2#$"'+-03" U<"(2V-($W-0$1!)$(1)(-+)&2#0'-
&2'-,"&19)3"(&*"(&&*A2/&#++)-$ &2.-092#,-#$(2#('1/(2'&2#0'.
$-0(2B'(9-(-')#($((2/$-(')/*C-&2 -$-'3&#(-&*
#$D3E77F!'-0+))(:'2#+$1$)'/$ 5 19/'($-'2#,'2..-$#0'
(2#(/)#'(-&/$1$)''-)-&)#'(1'!(2 -'#)-,#3G. #('#10-$-,-$#)'*?0)
9-(/&.#''311(-&"(2/$$-$(/#&(#$ (#)*F#!/$#1#G. #(/*FP
(2+&-,$2#$''.#')#($((2+&-,$ *F1)X1-/'(-1)#($G.#$#1#/E*F
$/1#(-$&'(#(9-(/&*5/$4#1: P*J@1)X1-/'(-1)#($G.'*Y#)-,#2#'1#3
+)/'&2#/$+9#9)3$'(4-'(9('1 /&(-'9(/1##)(4(+'+&(-,-(#&('#'
/$''&2#'0)#(-0)'&1&)'*>/$1$)' #)9-&#(*A21&-'0)3&+(-'!#'+'-:
(2#(.1+)#'(-&$(#).#4'!(29-(-0/& 9)/(2)9-&#(-//&('/'#)-,#*5''2.93
-&#''(-)#3-)$+-(-'#&2$.2(2/$ D-;(#)*@!'#)-,#3)9-&#(--'-&#'-0)3
90-'(G.#$(2/#&(*A23/$(2#((2 /B&-(.-(22-02'/#&'+$'#$-&#'$':
1#4-1#)9-(/&.#'2-02)3&)#($(+&-,$ /#&)#$*5/.'($-'2#,'2.(2#((021#(
2#$''HI*7J!*5#)/$4#1+)/#+)#'(-& '#1+)')#$(2-02-&+#(-/'#)-,#-((2
/$-'9((*>9-(()/$1$)'+2#1#&(-: 9)'+-('.#)).-0K)#$(#)*"FRC-&2
&#)(#9)('!(23/$(2#((2B'(9-(9-(-0&3&) (#)*"L!*A2'#)-,#3+Q#$Z:#13)#'&((
.#'(2'2('(.-(2#9+(-&#''#$$&#'' #)'#//&('+&-,$(4(*?0)(#)*@!
-/&#$#0#-+&-,$2#$''.#'2-02)3 /$(2#(Z:#13)#'#&(-,-(3.#'0#(-,)3&:
&)#($(1#4-1#)9-(/&HI*77!*K<-' )#($(+&-,$(2-&<1(2/)/&'(#$'#$(
###).)$/$4#1+)/#9-(()+$&(* +&-,$+-&<)31(2/)/1#3#-'*
D&-,$2#$''9#'$B'(9-(-&#''.-(2
/$(2-&<''50#.#)#$%&#'"L!*=M-N<
(#)*O!/$(2#((29-('-;(202#'(#./ [[\]\^\]_àbcde_fcgb
#&2&)#(:G#,$$#-3'1-:')-$.#''-0-B&#()3
'1#))F*OP*L0!(2#/#&2&)#(:G#,$)-8: A3)E77L!,#)#($(2//&(/'-;"'2#+"#$
-$$#-3$-<O*@P*F0!*Q.,".2(2' 2#$''/''+$$+#(-&)'(2#)+&+(-
#(2'1,$(29-(//(93'-0#+1+!(23 /0-((-''/'3+'*Q/$(2#('/("$$"
/$(2#((2'$-//&'$-'#++#$* )#(-,)32#$"G#(+#(-&)'.(+&+(#))3
K2.-0/#01('')-$#$'1-:')-$/$'-( 0-((3+(#9(Oh1*Q.,"2#$#0)#+#:
'1#))02+#(-&)'('.#)).#$(1-4(2' (-&)'&(-9($(0-((-''+&+(-.2(23
+#(-&)'.-(2'#)-,#(/1 #)9-&#($9)'/ .#9,#'-;#0/EEih1*j-&2#$'
'.#)).-0*A2-')#0,#-#(--&2.-0&3&)' #$(2E77L!/$(2#('1/(2-+#)-'('
011 22 34567849
76
!"# /*63/
!$%&'( !&/3
)*+,-%&./0*1* &3&"
'2))+0!/33/2 'H0*IF3.
4)%&/& 3<3
/// !" &/&'C$0
*54)%& //! &/3&/&
*6 &&& 'C20*J3!C2
/#& "/!!/
-%&'7.$8890*:;'$88-0 3&C$*
<&<&!/" C.&<
3! !&/*A 3/
$&=*:&'2))>0<! &/L<"&M&
/3 &/&.&/#. &
!)*)+>&=* &/?&3!!
;3?'=.$84-0//3" *
!&/3/
!&!&*5//33
@!&/! NNOPOQOQRSTUVWXXY
/ &!/&&
#*A B&/./!! 7!!
&& !&! 3&3& ,
&3#*CD3B// B*A B&/.&!
!!3&/#*E&&" //333
/#&3 &3//3& //3&/"
&*5F & &!/!
&/&3! !!&" < 3!&"
/# &33 "&&/*G!&!
&/#*6< 3./<'
!//.!*G!! &/0..<"
B/33!!!&/#B/" '0..
/3HI$848*J3 .'&//"
!!!&/#/ /&/0..
!&.5&. &'&//
/! &&0.
3.<!!<.J &3!/*A&
*J&/#-'&0 &!3
)*$2+.$*)).4*))& B/&" !!< "
F&/" //!/*C.&&
!&/.&// & !3!/
3#D!&/3" Z !!<.B&/3.
!!/D/ //3.<3*G &!&"
&/ *J! 33
!!&" H/2*
/#/!@!&/ 5 ';*.$8[-0.
#*J!.B//// JB=?&3
/!&/#*53. &!"L3*MI##<
3K*'2))403//!" '$8[[0?&!/(
/?? \3"'.<0D!!!
0012 3456789
64 977486 2
!"##$%&&'()*#+,#+ h 6jk(lmnot6u9
<9 69 i!
!,+-'%#"'%***'()*#' !.'%$/'% =$,0+de fg <plqrs9
,-$()$", !,+-+%+"*,+-%##,+-()
#,+-#$#/,' !,+-,+-%'%#,'() .$
#,+##+-"#/"+!,"'%,0'%
&'%'&'()"#$ !+%,+-1 ev,3,&$/,#,'
,+-()&$',##-, !2,+-%$1 qv"& ,+*.+$#, &$
,+-%$,3+$,+-()"& !$#%#() !#_+,"(+ $#+-
+ .+ !.% '(45#.,!6789( 2+ kv/#",#+#,,+$&#!#/+(
"+'#$"#$"4:,+#+ ;. mnv$#,#+#,
!677<( ",,"+,#+#,+-+ ,+-# ov +,'#,
",+$+',+- ,"+,#+#, .,$ gv/#,'##+-"#/"+
"#$,*,+#$$1&#4=$' wv#&&, *'#+-
#+ $$ ,+$-#&>,+-+'% sv#+-,!",+-,(,+#+.,$
+"*,+-%+ &,+4 $##
pv$"#+,/,'#" ,"
Y&#,+ #*';.4!677x(\.$,$,
??@A@B@CDEFGHIEFJKHLMHNOPJKQPJOEHRSFN 3 ,#+"'*#&#,+#$/,#1
IFTPOU #,+y+,+-",+-'"%,, #*$
$&"# *0+#.+,+&,#+1
=$",+-*$/,####,+$"#$+ ,_ "#+-+#'&+,4]=$%$,,"%
$#, V$+-$/+#*+1 $'#",+-,,*,+- #+"&,,'
, V+,/'4W+'##+-#&$$+- ,+-&+,+ ,&,/+#'/,#+#
,+$"#$ #$,+ "&,+$ ,"z,++,'"$###-'4=$$*+&$#
#/,'4XV"&$##+ ,"4 ##.1,"1 &,/ V+
Y"+,#+ ,Z'+ [,$'!677<(&#1 ",,+--+, ,&,/+',
&# +\,"4]=$' $'+", !Z'/{++4%i||<);,#+ }~+%i||8)#+
#+!$"#"+1#1"#"+$+-,++#' 4%6777(4;.4!677x(, $"1
V#+,+$"#$(,&#+,*#$ ,+-#,"&##*"### '"+
$,-$&*,,'#,"+ #$&#4=$ #+ $$,'"# * #'$V1
.#0*'Z$,+-+ ;,# !67^^(#+"&$1 + "&#&,#&4c$+-,+
,_,+-$ '+",*0#.+#$"&,+$ ",+-*$/,#% *' ,&,/+',
"#$ ,+-",,#+.*0$#-$$ + *',"z,++,'""+%. #
$#!*$+#'( #$,+-#-+ $ -#*,,#+*'*#$'&#'1
&$',+ &'$#&$',$'&#$,#&,/ "4W4!6779($ #+ $&+,#
V4 +#$ -#",+-,+$#/,'#1
`##","$+ ,+## "0,+-+ '&,$#++,+#,1,+1.",#+
&# /#&"+$*+#,",+"$ !&#,",#*(.,$",+-+-#
&#,*#"## &#4Z#./%$ 68z96c4
,&,",'&#+,*#$",+-#,"%
$##%'#-%4%,+$#/,'!;+
=&&%6779(4=$$$,,#, .,$ ??@A@B@RSFNISPGJHGGISJEPJHGG
&$$+-$# ,/ ,,#+,+' FJISFNPJHGG
&# /#&"&#&$"#$
$,,#.,$&#"&#+4 Y , ,+$,#+#+,#'V,&1
a++''b#0,+,+ c!67^<%67^8( +%+$,+%+ 0,+'$/+
#+ $$&,/ $,0+#",+-," ,#'#"&#++*$+,#+#$/+
,+$#/,'. #$##.,+-3,#+> #V#"&#++4$/,.*'#
012 33 4567895
87
!"#$ YZ[\]^^_` a*/ $,!$)+#
$ Q*#+# ?$#$+,!$)
%$$&''( )!! $$ b#+, a*$.!!! &/!04
)! #))$)$*$)+,! '/$2$++$+.$!
!#/+$)!$)*$)/))R$24
!$)!-./# .!$)/+0 #$!$) #!4.!#+0$2
+.$!!1$,!!#2!!#/+3 b/$ S/#$$/ $
$ 4$ 4#$!$ 4$)!)$ )3 / $ )+,&/!0 !4
$2$$$## )3)35$$##/ .!$#!)+.$!!#/+$)
!$)*R$24#$R$2#!!
6#+7#$,%$$484! #22)!0 !7#))/!
$$$)2$$!)$)#$!,) ! ##
)!$))#! / b/$ b7#)#!#2!&/!0
$)!$ )$$) $$ #$2 !.!0$#!)
)#+,2 ,#/ ).$$$$2#42#$30
/ "/$4/#$2!!
9$ ) )#.",:. 0$
$ $).",8;<<)$)3 9#)$ a*$.!!!2$/
$2$!)#!")$ !)3 )# #),#/ +$2
#) )/,=)$!48 0$)) "$/ +.$!#/+
$)$)*$)/))R$2
>/3?$(#$)!)#0 =)$ b7#) $2#$#
<@!")$ . ! /
" +.$<A$)<' c$ Q!/#$!) ) !
).$.!$$)#$2! $
",$! /$)$2
",/! /$
BBCDCEFGHIJKLMGNOPLLK 6)+, Q!.!.!!!)#$+
/$#)$! #!
/%$d<49
Q!$)) ##,"#)!#2! 9!0d'Xe!$$fe!$$83!
?#5d&g
!##$!/#$# Q0$ Q!)2.!!R$2)!!
04!+#)+,.!0$4 )#X/#$)!"$
!.!.!!0$ +)4 e!$$fe!$$h+,id4
+,/$#$+,!$)! $3 >3$dA<
$2, 04!)R#,$/ $2 U,' 4S T4$)
S)
) / >

!+.$!!#/+$)R$24
Q+&&8 #//5 .!$)
+#>#)$))3
"$"#!$2 $)! $,"#$! $$$!
*#+# .!#2!!$)S )$ .j.!# )+ /!
T#)$$3*#"#$!3 "+#, ).!+!$).!
5)#)!U,'#) / !$$$/#$2) $,
!$)+# $!"#$)3 .!$k,.!!)l$3
#U)$$)6!.5'#$)!! $.!$#))#?
$))"$,$#))#+, !$/$$! $. ).$/9"
++"$ 4 $+*#4 3 $)U# &''4?2)T4$)9"
$ 4$)#55$ #$2!R$2V #) &''T
(#)W!R$2"#! 9"$)U# &'')+!3
#/+*#.!+) ).!+$) /!$3
Q!*#"#$+7#$, ! / 4$4
$#) #!$2/$#$2!/.! $) $ 42/! #55,4
!R$2?#!!.0$!/ / 2,4/#,4.$)!/!3
!*#X!."4.!! ./!4$)$$3 #$)
0012 3456789
64 977486 2
!!"#$%$&$'(%)&$!%(*)&&+ $&#5#$,+0"!!#$,+$#/#0+6
+%$,#-,&.$$%+/#01$2&3%+% E#0/#&63BFFG40+%!%$"!+$,"#$%
,#,/#+#$,4,%,&0%+++$+#,%% #$)&0#$0',++,%!N&$5$),,%N,##%
+&+5#,,$+,%!+%2,#$,+ +6P0#&638??H40+%7#),Q+&5+!+(R
$##0#)&0#%67$"$#+&+#%, $!(+&%+!#$,$,%)&0#$,+6:,
&88696 ,$#+#0%(>###&D0&#+$+5)&R
:,++$+#0%+;!,+#/#&+,#+#+ 0#%0+,'!$&%+!#)+&+3;$+,#&6>
3<5#,%=5>8?@?A<5##&6>8??B> 8??B460#$+$0,%'$$%$&#$,+#5,
AC#+0%,%<5#>8?@@4D0,#2%,% #)&0+$#,%(#!,&,%5#,+#0",#&
$&#%+,+$()&0#$,+0&#+$#/#&+5# ,%,',%$"!++$,)&0#$,+61&!!#&6
,+#0",#&"+0",#+6#,D0+) 3BFFS4,%*%&&$#&63BFFH4%+%#0+
,/#,+)&(0+%>+#0%%>,%%!#%5#,# $',%+!#),&(++,#)&0#$,$
#/#&,%0+#(3E#0/#&6>BFFGA*%&&$#&6> ##/#&,%$$##$,##%5#%R
BFFHA*,#&6>8??BA<"#&6>BFF9A<$&#&6> ,#,%0+#&2,++,%,"&#(&$#,'+>
BFFIA7J#,%>BFF?AK%&&#&6>8??946 +!#)&(6
L#+,+$(#/#&"+0",#+)&+$, C#&638??F4$0,%###50&#0&
%)&$!%61%$"!+$,%+",#$,#+#+ %,+,#,%&!,+$",Q+5,#
),0+%#$++++#+#,++>+"$$#,++> +0#+6!,&+#+$"T0+#&>:,%>=5
,%+$#,++$$##$,+3MN!$,"5,>8?@@46 U&,%>#M,#%7##+>,%P$,'N$,'5,%
E0,+#&638??94$0,%##+0O#+5$)5% $,++#,#!,++%$,#)&0#$,$#
,%&#+%+%#+,+$(!$!#+$ ,%0+,'#%+!#$++&N,++>0&&,++
+%,#&(#,%%+0O#+5$$,&(&## 2",++>,%%!6!,&+#+$";!,,%#
+$#,%6(0#$,%##&$#$( 1$!&Q+V!0&$*,%,#,&&($,++#,#
#,D0+##$,&($,,##%$,,%"(,$# ,%+$"5#$!!$+#!,+#$##$#2+#
WXYZ[\\]^ 7&#%,%!$2&##0#%2,#$,+,%,,$+
T##0# _2,#$, 7&)&0 .#(!
.$#$$"!++ T"$0,#$$D0%#$$"!++ 869 1$&(+#$##$,9F9FN,##00&
'#%+"!&,!&"3&$5$#$ H6S *$##$,&$#''
'$4 ?6H *$##$,#(&$#
8S69 `8F*$##$,%0N''
V+&, .$5#5+"!&!+++',+# 86F 1$&(+#$##$,9F9FN,##00&
0!!%,%+3+%#$$&%% G6F .&",#,(&$,I6I+"R%0&&
$',&+!4 ##
8S6F _$,
7#,++ _'#$5+"!&&+!$,#%> 86H 1$&(+#$##$,9F9FN,##00&
%'%>,%N%>,$#$0,%>!&&> S6G CJ%$##$,!,#&$#
0)%3!&&#$+#4 @69 CJ%$"%$##$,!$!&,
8S6F *$##$,$',%(
-$"#&!$!#+
.0JJ,++ T"$0,#$!&>2>0JJ$,+0$ F6G _$,
+"!&3&%#$0JJ($,!!(4 H6I *$##$,,N&'0J
G6F *$##$,R+#>#00&
896F *$##$,a
-##,++ T"$0,#$+"&&!N(!#&+,+0 869 .&",#,&#$#
$+"!&3+"$$##$'##(4 I6F *$##$,&$#''
8F6F *$##$,!,#&$#
8869 *$##$,$',%(
T%!#%$"*)&&38??I4
012 33 4567895
87
!""!#$%&%&' &XL$,$ &&H&0&
&'(')**+$&$'&&&$$+ '&L0YK0,+kl
),'&-&$&%&$$$$%&$.$ &&.+-&$&0%&
$)&$&$$)&)**+'&'( &&- - &0& )&H&
-&$$+&/-01&$$&$ 2 -'-,,&-IK&
&13445# )&
O*'*$J!"5P#0&H&'$$K'&
$+$&%&,&,&-'$%&H&&I
66789:;<=>?@:ABC>:D:E<C>:F:;B< % +&'&$)&'&
.,!!5#O'&*&'0&H&
G+&H&&&,&$',%$$$&I '''&&$$&&''$
' )'&$$&$m%''
$$+&'($$'$$'%&&$&1 ''&$&'$%
$1 &'''&$&'$1&
J1$'&0&'($.H&))I '&&))$#.$' %$&+&&)&%$&
$$K'&)%&,$$
'$,&-&-$%$',&%$ )&.H&YK%
&&-I&&0)'''&$&.$&$ .$&$K&&% n&&1!"5P#
$,&&$'%&,&-&$%$ --&$'&&)& $&'(&&
,$&$'K&H&&&,&LH%1 %&'1&&H&$$&% ''1I
$&$',&&0&0$'$&+$M& '&)%&K$&,&&'$% $ $ '&-&
%&))'(&-()%$ &'&.&'(&),-$&M &%$I

0&&N O*'*$J&1!"5P#' )'N')%&L0YK
&,$J&&%)&'00$'$&+) $&$&%I&%&.&%$('
&-%$+$%$))
Q&$$$$,&(&)+&H&&&,&$$ %
$&MK$&,&N&$1&M'-N
&0$'$RH%$-,MJ &$ &'&+$'$)-,+&$
&SSSNM$'&SSSNT$&H&1$'+1 ,+H&$$0.
+&
&H&$$&$-0&
$&$$$&&I
$&$)-&$%&$&+$'$1 $'$.
$'$&$$)&$)'1$*%I 0,+O*'*$$ &&$'$-
,)$)'&&$1$&+%&) )&$&+$$&J $)
&!"5P#&'0&
)$.+$
$%&'&)('&G$&)
&$$%''&&&,&$(&&&I
',$+-J&$'&0K
)&'&U)'$1$+&$'&0 @$E% op:667q .H&'$$K'&&,&$%'I
$'&'(1&$','%$%',& &H&$'&O'$+
&$,&&%&%$&$)& Y%+&%$ &%$ j$%&%$
-$$-&0&)%) k$0$$ O &'J+1&'J+1+
)'&&'$&$' j$0$$ n&&$$ j%,+1''+1,&&
O*'*$J&!"VW#$'$%$&0I j-$$ .1'-+1&
&)$$&%+0)&X X%%$$ O&1%+1$&+1
L$.H&YK%& $,-#&+ R$&'&+ Y
%%+
$&'1$&'
)&&'$%$'$&$'$- r$$ O )&1K%1
$)K'&+-)&&H&))'&$& T$'$&+ . 1&'J
%&+M)**+N&H&K Y&'$ j1'+$&1
&& K,$1&'
Y&'$* j$1+1&&+1
ZZ[\[Z]^_àb^cbdef^g^`hdi $
L&'&& X$$$
&'
X$+
U$$ U+
..H&YK%&-$0&X G$&'&& l+1%$&1-&1-&+
L$j&&+!"54$.$'&$&$ k&)%O*'*$J !"5P#
0012 3456789
4 8446 84456 0
!!"# '!!,"".+@';#/'"")
$!"%&"'(") "!"+-+.,!
!(*+"*!'*,--) +".'!+"-"+.(',
*!*(.#$!!(" !","".#D,!('
!"+.",!"++. .-,"!"+
!."!"+) #/"".*!""!""".
-"".+.(-&-#/ " (;!"-(#
'"".-"& $!$&0"!+ -
"# ,"#>(""J;456<=9
$!$&0"-!'&(". + ! - !#$!
1"2! -+% "-.,-!
,"'&(-*&-"* !!."*,!(")
3"#456789".'!!'" !!"
!,,4:%%;*56<=9*:%%; ,!("*,"
3456769- "+" (""F-)"#/"".*!
'-+".4>("":%%;* !(-"'("
56<89#$!,"$&0"!"(.) '"+".#
,",!*&-"*!"!'! BC%456?79+"!+,!
( *,--!(*! " '!!.
! 43*56?@9#$!"+.>("" !"#3' 567856?7-.
:%%;456<895A)." !!, -" '
!+.BC%456?79+5=- ",(. )
""'!!.! $&0""!' ("+"
!"# .-#:!"-E! -+
>("":%%;456<89".+ !"#$+"55#<55#?)
!'$&0""!"+" "".+ !-(-!$&0"
#$!.,,,+ " -+.BC%456?79#
'!!,"!(,"&("+""" G!!(-!"+)
-#$!"!",!."," ! **&-"*>!( "#
,"-""'! 4@KK?9'! '"!'.
!'!!+".--, +L*!*;"#M
&"#0""(' !!!"-
*("+".**--) ."-!)
;""#D,",*!" !"#M '!
&!+'!E( -!$&0""
& !".," '-+!/-
!$&0"#$!." !' ("+" !*
!," - ) &-"*3"#456<=9*G,+."#
!#$!"'"",!, 4@KK=9I )*:!'%456<=9#
" #$! $!:!'%",";
-.+F&(*",("-!#/ . ""-
-," -!!(-; "#$!
"(# (' +.:; 456??9(.-")
G !"!+ *'!! - !!&"!#
"- ----@H8!". $!.&""" **
@';""'+.7-!5!) &-"*JN 4@KK59'!!
(-';I!" +, !F+3"B""
(",#G !!!- 4@KK59'!!#>!( "#
"'!&" 4@KK?9(" '"*
010 22 34567849
76
!"#$%&##&'
"#$ (&##&
)&*"!
+"&"!"%,'#-&'' ./0#%'""#&%#-#%"!#/'"%,#&1"/,,'#&1,$"/23"'&4
"3!"/#&1&#/&&'3,'"&%"%/&"!"
5#'/'#6 (13'#'"&/7,&'#%$6"'&
8""7,#/,'"9!:7'%7&,'#%3"#!%/&"#&
;"!
<&##"!!#/&"/
=%,'#-&''!#' (17,#/,,%/'#/2>"%,',!#'?,'"9!#'!"/%$7&,!#'"&%
/9''%&/'!#1,!6"&%!"'%"''''!#"%,'#-&''
@&'' =9&39#'/#-%&,'3"/3,%/47,&#&/&"/7#,,!#
<&##"!#&'#&#&9,
81,&'' (13"$"'#-&''3,%/A''3"/4"'/#-%$6,&1
B!3*"%,'#-&'' ./0#%'""#&%#-#%"!#/'7#,,&147,&,'"9!#'!"/%#&,
9,
B#&1#&'' ./7#,7,#/,,'"9!&'#'#1#&"!'#C>',"4"3"#"!/9''#&D7#,
3"#!E$7&,&1"&%,"!"
<&##"!$#
F,'#-&'' =9&3%39"#&&%1&$6,9"#"!$37,&$##&1/9!!6
,1,'"9!7#,9!"'
=%,'#-&''"!" ./0#%9-%//9!!639"!"4'#&1&14"3/9''#&3,
'"9!$7&&1"&%"!"
(&'&'' F9"/&''3,/'''/#&3,'"9!"3$##&1/9!!6,1,7#,9!"'
."/"$#!#6 ./7#,7,#/,,'"9!'7,&!"/%$7&9!"'"&%$#&/9!!6%7&
""3"'"
G"%&'' ./0#%$#/9!!6,1,'"9!!"/%$7&9!"'
=3/,7#&1
=%,'#-&'', =9&3%/"%,#&1&>#&,,"39"'#/"#&3,%/
F,'#-&''39"'' (17,#/,,9"'',!%'1,"39"'#/"#&3%/
+#'"$'#& =9&3'"!#-""$'$%$6,'"9!"39"'#/"#&3%/
=%"%39+HCDIJKLE"&%B,9"&DIJMME
/"/2!#&''4"&%/&/,#&''#&%6"&%73%' OOPQPRSTUVWXVYZ[W\]V^T_WVàbc_bTd[Y
D"$!II?JE?
F"%!!"!?DIJKNE'%3*9%!'9"1&#* ]VeUYdf_VZ
%'#9"#&'/"!,'"&%"%!
'/"!'3"%,'#-&''4/,7#&''43"/"$#!#64,"%* ,'&'6'/#&#'%'&,"-"#&""&!
&''4199#&''4"&%-#'/'#6?,63&%,", ','&'6!"&"!6'#'/,&#0?<
/"16'/"!'3,"%##&"!!7 #'&#!6''#$!'1&#/'&'6%'/##-
/&/"-%7&7"%7,&!%"1"#&',9"1&#* "&"!6'#'%'/#$%#33&/'#&,'3%*
%'#9"#&'/"!'?,#'#&%#/"',"3,' /'?."9!4@&&"!?DNgghE'%/&'&*
"#$',"&!#'',#$#"1"%#'/#9#&"#& ''"#&#&1"#&""&!-"!"9"6&&"#''4
",!7!-!'3#&&'#6?,#'#'""&/&* '"!"%%''#&1'4/'"%'4"&%7"9 '"/'?,6
'#'&7#,@$A'!"7D'F,"NE?@$A'!"7 3&%,","&!1%,'&'63
%#/''9"!!%#33&/,',!%'"!7!-!'3 ,''9#*'!#%3%'#&'#/!''D-#'/*!"'#/*
#&&'#6?,%"""!''11',",'!'39 !"%"#$'i'3"/3!*!"%"#$'i$!2
/"16'/"!'"&%9"1&#%'#9"#&'/"!'" ,91&#6*!"% "#$'i"%,'#&>/,'#&*
%#33&$'#9#!"? !"%"#$'i7&''>%6&''*!"%"#$'i
0012 3456789
4 8446 84456 2
!"#$ .!%&'!%%'()&%! /,! 0'#1&,%
%!%&'!%%'()&%!(*%!*"(%+ 21-!"(3!*!"" 4#5 6+1'#7!*%!13!7!%)!#(
*,-#'" 8!1(& 8'"-#5%#(*7
6(7- 0!*&%)&%%!'
21-!"(3!*!""%#(" 4#5 .#%#
8!1(& 9'!1"%(,:
6(7- ;(,!,!'!
21-!"(3!*!""%#%!!%- 4#5 <
8!1(& ='-,',:!'
6(7- >&?&)!"
<#-!"(3!*!"" 4#5 <#'*&$@*
8!1(& A'(!1$'&(%
6(7- <-!5(*77&
<#-!"(3!*!""#$"" 4#5 4(,#'(,!
8!1(& B'*:$&'%!'
6(7- A#&7-
A!*"!*!"" 4#5 C-(!1%#(*7
8!1(& 8%!1(:)"
6(7- B'&(%?!(!"
B',%&')((%+ 4#5 <#'*&$@*
8!1(& 7(*7!'"*D(*"(1!'%E
6(7- 6'1,*1+
6'1*!"" 4#5 <'!,-!!"!
8!1(& B'*:$&'%!'
6(7- 6'1,*1+
8*&1-!"(3!*!"" 4#5 8'"-#5
8!1(& A#&7-
6(7- F#&7%
8#("%&'!)"#'%(#* 4#5 4(,#'(,!
8!1(& 0#%%#,-(
6(7- <',:!'
;#&7-*!"" 4#5 =!%(*1!""!'%
8!1(& 0#%%#,-(
6(7- .-(*)'!15$!'
/!$G1-!"(3!*!"" 4#5 =&(G)!'
8!1(& 2!'(,*,-!!"!
6(7- <'!
/'(*7(*!"" 4#5 <'!,-!!"!
8!1(& 8'"-#5
6(7- =!%(*1!""!'%
C!%*!"" 4#5 <',:!'
8!1(& 6
6(7- C$!'
21%!1$'#8&H#IDJKLME*18!(7'1!%NDOPPME
*1$%G'!%!1%%'()&%!"EN.-!"!&%-#'""&)"!Q&!*%+ !%NROPPPER:!%,-&DW'!!%NROPPSER#%)'!1"
&"!17!*!'(,1!",'(%(3!*+"("*!"%#!3&%! D/!*:(#G8'%%(!%NROPPXER,'!+$##1"
5(1!'*7!#$"!(G"#(1$##1"&*1!'1($$!'!*%,#*G D.#&'*(!'!%NROPPYER,'("*1,'&*,-+1'+$##1"
1(%(#*"D*7!!*!%NROPPSTC!!*!*!%NROPPUEN DA(?:"%!'-&("!%NROPPYER*7#&'!!5(%-11!1
V%-!'"-3!"#&"!17!*!'(,1!",'(%(3!*+"(" )'(&"&$%!D:)!'7!%NROPPKER*1+#**("!"
%#1!",'()!%-!%!%&'!#$,##:!1#%%#!"D.-+)# D.!'"%'!%NROPPKEN
012 33 4567895
87
!"!#$"%!&!'("%"%!"$)*%+)**%
,((-&( .)! /"&)"!(& 0"1$2"%!"$3"$&
4%+)**%5
6 .!7("( 7$$*889:"(($#/; <=>-"
? @-+-"% A"":"*!#-&;B <=C$!
D /&$(8"1!!"# E*+F"'"1 G&"#'!"8*;B <!"#
<H :(2I*((*(*($$"!' @(*B"+J"$$"IK <!'
<? 7(($' @(*B"+J"$$"IK <!'
"!#$4%+)**%5 <=
6 $&-!"!# 7-$:"(($#/; 6!"!#
<=
L /&$(8"1!!"# E*+F"'"1 G&"#'!"8*;B 6!"#
M BN((:&("!**# B&:"!$*"0" <=DO&"
<=
<6 I!&( P"-!*=7")(@**%'!"8*;B Q
<=
-*#R('8"
<? F8" ,!'R"+:"(($#/; 6!**#
,%"(%)*1'"&3("$S46HHD5
TTUVUVWXYZ[\]^XZ_à^bZ\[^ 1"8"!*)**%*%&!(p@*k"1$ ('

c^_Y\[^]^XZY_Xde^XYf[g *+'"%*)!'('"1"(''"%*)
"!**#*('!3%o&!"*!"(%R('
hf[[^`_ZifXY "8"('"1"('"(!3%!**#%("#p
, 8$"!"(('$("(&R*&$%%!"(1"+
jIk(&"*+*(+l402!2"#46HH65 k"1$R'('"&('*&('"1R*%4S8S
"%('&('8*"$*)(&1("$j(k(&l1"&m '"%5)*('1"&1(*)-*('*+"%
1((**%&!"1!'"!"$((('"(!"$"! (&1("$(k(&""1((')**%*%&!(S
*+"$"(k(&3"$&"(*(**$SI'%(* I'*-$1('"(('1"&1("*)(*(
$"!('*+"$&&"$$+%&(*!*(*)nm '8'$+!*$"(%R('*"*('SA'('*!!&
!!+S:"!O&(*('"('*&$%-"#%"R'"( ('"&('**)"*(*!*$*"'*&$%-k(1$+
"1"(-+*-o!(31!'"!"$j(k(&l*( !")&$(*%(8&'-(R('*+"%('
"%%*"*+(k(&"$*(+'"3&3"$ (&1("$1"&1(S@8&<<S<%!"("
vs
vr
|}~|}ÿ}}
vq
u
t
©ª« B"8* s
"%!*$"(*-(R(' ÿÿ ÿ¡vxwzÿ¢ÿxsÿ£¤¥¦§¨§
*+(k(&"((-&( r ÿqw{qy
"1$!*3+"%(' r
(&1("$(k(&""1(
84IN,¬(k(& q q
*n$""$+5S.%"R" qwx qws qwy qwt qwz qwu qw{
%))(*("(*)*171 |~~}
("$S46HHM5S
0012 3456789
4 8446 84456
!"# -*!)---&"-+
$%&'#()*$'(+,)%& ,,&,
&*"-.)+/0 *1" &&,
234556+2
7,,&,!.)/&& *+$,/#-,
//"/,)&""8 &" /"-&2()3
&"29)3")&"%&"8 ,'R/#
&"!"-#-" ,-3,&/,,-&.8
*..&+!,)%& ,.&") ,")
.&2:&;<<</%")/#& /"2
-)&/"&=&&" *+$,&""&",&""
-"<<<>*:&?&3455@+,& /-,&.&
)3"),. %-", ,)&2$,&/.8
,!& ""&.&*A23 & ) ,&,,"-!,
455B+2 ,,,/"&"
9/,--&" "0) ,,)29,,
%&-"!,)%&&! ,3/,/,"-.&8
C"2*@6BD+2$,&,!-/, ),!!,,",,
&-.-,EC$%&'#2 /"&"") !,
$,)! -/=&-"8 ,-.27/,"-.8
,%&"&"# )FGG0 &." "-&,,
*@6BD+-,)"&"2$,EC &""&",& ",
$%&",-&!,,-,/ ""-&2H-,..&8
!)3,-/! /,",!)2
&/,&"%&-#! S)T*455@+-&,
!!,,)%& "8&,&"%&
-" ),$%&'#-8 "&"/=&)-/",!
2H&,/&/,&"8 #,,)%&"&,
%&-",&,,,,8 0#,27,&&-,&8
!,)&""&8 !,!"&"
"2$,)&#,$%&" !2
-&,&" /--&"8 *+F,"-/)&,
)%&&*FGG023 "&",""- .&8
@6BD+2$,"&",=& , ) ,",2$,/,"-/
$%&" "0!,$%&'# ") -/,-"3-)3/,
()*$'(+!,,///" ,) .,%&&/
$%&'#",2I3,$'( ,=&3 "-/","&2$,/=&)
.-!,,$%&"3!&!,, -"!,""-!,,8
7J.$?,, !,"&.)&)
=&-"*:3@6KLMC)3@6B6MFGG03 *U.23455L+2T!8&.",8
@6BB3@6B6MA23455B+2 &,8/--)*:G=&G
FGG0*@6BN3@6NK+&)-8 23455B+!,)&,
/)8 ""-/ ,,&"%&
)&"&2F,/ "&" &")/,",
&),/ ,/)2
.& )O8:G,3,8 *+T&.)"-"/
/#/"P526QP52@B2F, ,!)33.)V!323
,/&", ,,)&8 /,!//,=&)/&"8
""&"!-/"& %&,-2
012 33 4567895
87
!"# $ O'"(3%%G'!%!.3(%&%
%&'()& '%&'!$*&% % (%$.-%&(,%''%,(
+&'(,'-,'.!(/0$+&1 (U!()&BD!(/,566=C
'(!$*&%'%& ((12 .$3%!( E#(#1&(,566=8CV%%&(/,
..%$'$&$%+($ 566=CD%0$38,5666CW$"(/,566K/
'%&'!$*&!&((3+&'( J$566-%$%%&'(!$*&%
'%&'/4!(3,4%%566 %$%3 $&%%$!$!#+.!-%&+!
%$3'!-%!''%&'%7&% )&/
%((%'%&'%++'%(%
.88(3+&'('%&'%+)&/9'
(/566%$$:1.8"%&% XXYZ[\]^_àb\]a
'%&$+!&8(3+%$1%$.+%%
(%+&'('%&'/0$%!$*&%
&%$%&!!%%834;<!566= %&3 0$%%3-(&+)&$%-!"
.!-!%.%%.!$.%/>$'- (%!$'(+$%!&3,3
!$*&%$'%&'''!%$!%+$ I(M#!#%F,%&(3$3++)&
$$;%&"$%%3'%&'% $cM%,!&(%((%$P$%((
!(%(3%.%%8(/?3("@(%566 % '3'.".%$!%&'G)&+!
$$%&'(2'%%-(&+%(' $."%%$%F.&.$$$
83!'.%%%'-'$( +%&'()&'%&'%/QM$!&%
$%%32'%%-(&+.%%"$ Pd&-'%&%+$(-'.!+
%('$$)2"/A)'.(%+'- )&%.!2!+!"%%$&(8-(.
!$*&%$.&!&%$!$'%&% ($(-(+)&(*&(3/Q0$%%+
$+!*&.&!$$($+ ++%%'.$%#83J$566 $%$
!'8+%$!'.%%+!%,$&% PTTT$&"$&%"+$.!.(%
+%&'%&'%"$%!%,B&#,566C '!$%'%-(-+(.!%%"((8
4%%,566C9'(/,566,(!'3".$3 %%(/>$&%&!$F(",&%&%++
ADE/0$(3F83<!F%!F% )&.88(3&("+/QE-$'.1
%<!F%,=8 &%"$%&%%%! !++)&+*&(3!!.!,
$8"G!$"3!%.G!&!$3+ $%%(("(+F$'&%8
'.!-!%.%%!&!$%% $%/
$%8).$&"$$&%++%H&
%+' ("$'%I(1J$F(/,K
+!((3%%(/,L,E#(#1
&(,566= (3#%&+*&!% efgfhf]^fa
((/,5665/ME#N(#(/566
+-%+ADE +)&-(&%/ I"(,9/4/V&!%,i/>/566:/0$'!$!%+$
I(+'ADE !8+& 2%8 /i!"%+$43(M!3,V,,
H%(/,566KCE#N(#(/,566LCO% I(1J$F,!>/
(" (M!!,56,5U5=5/
,I((+,9/',I//K/J$!#
(/,566,566/ +8(+.&!%jI..(!+!&%!('%%
D%%&'()&'%&'%'.1 '$/&(+0)&M&%,5,:5U:L=/
!(!%%(3P%(Q!%%+ I' %,E/,E'k#,I/EN'8,I/566K/O%&'(
.&!%/0$%%!%%(3.8(' %! &$(% !$!#()&+&(!(!$/
BF(/ %$P>$(+&'1 ((,J/m/,J(0&)%%&,?/
+ M&%,:,ll:UlK/
($("!(%-('.+' *&(3+..!F"A/ M,A/H/l/H!%++!"
'/O/V%%+"$%&3+
F %&!& '(!&("' ),I/?/
&/H0!$("3,,5=U5=L/
R!$%S,TTT'.!()&-(&%F L/J((,J ((,I/</,V%$,V/V/0&8,O/I/
*&((3((8.!"%%3)& +!'3(,'
&( !$!(+(&)&(.!.1
)&%/&(+0)&M&%,5l,
..%/Q Ul/
0121314516 788
9

!"#
(( ' #

(+'"("+ '
"
0
$!
%& !
' (('")!# 6

'")!544.45
*
+*!
,-./ ;K
!

)NC6
'"
3D4--4*(
'
9
!01
2 3

*4--56
" (
+ (
:!

N
$

# ' '

((

#7
2
(
'
+
0!
#+!! ' N
#$ '
0
6

'")!55,5.8 (
P(

!

")!#C
99:!
9;3
+<=>;"3? *'
'
8544/.44,
3
"!+;*"
"117
;=>3
;N@1!+
'"
2D

;@4--8
#
+ !

0!
6)7?
4--5& (7
!

(

('

'"

$+
#

A
''
'"'
9 #
'" ''! '((

#
(
''(+)!#3 +6

'"5B,.8 ' (+'
! #)!#
0!
*!
,
9
&@ 9)CD!#
C+&)) 8,.54/
4--51##
'#$! 9
'"
# ;7 ;
!

)NC;Q 6=
;
!
!
7

)!#=9 +/ 4--B"

" (#9$
"
&NK

B54.B,4 "
D
N99
"

+
'"

9!
4--/6
$

0!
'

#$
$
G
)!#!'
C
)!#
0!
*!
B4B./ 844.4/4
9!
;4--4C
0!
? ' +%;'
( ;"!7 &F!'
C6;*94--,

!

*
'1 &'
'E
F2 *'
#' (
''2
'!'"

9!
;,4C
0!
? ' +%;'
( +
#%6
" ( "'! '

N

!

&'
'E
F2EF )!#
0!
*!
/B.8,
9!
;*7&6? $; ;"
)4--5*!#'

0!
##%@
'
( '"'N
9!'2
*5B
+
0!
(G
(

9 ; '
7
C*'
'
E! /5

7
(
# '
; $ B,.B,
)!#
0!
*!
/.B4 ;"
)4--C ('
.
7
C
9&*2
1H;
)&8
0!
+' 4.4B
(G

")!#C*'
'
4,/-/./- ;"
@N92
6* "<D*'!94
9

D5B&(( ' #
0!
(G
+
#
#2 #$ '
0
6

'")!84
!
#
0!
7! )!#

0!
4--.4
*!
8B.,4 ;"

;,/C
0!
(
'
( J%
9
)<) 655*
#'##
' ;=;" '"

129**'"

"
$I
' 7
!+##
0!
J%<<9 '") L
M&7'
C6

'"&'
'=C
<9
K)@2
L
M*
+@(

# ((B.
C&((
*'
'
((445.4/, ;"

;? '2
H,6
" (#
9
! @
!

)NC4--&' ( #"
'"
!I!

##' (
)!#
!

#(
'
+
0!
(G
# C*'
'
/8B5/.B55
'"

)!#
0!
*!
4/.BB ; 7
<? 2J5B G'
9D13!'"
;D2
J8JA!
'
# (( ' ##"

#
+
0N
'"

#G'
'+#
0!
A7!(
'
( !
(G

'" :!
)!#
0!
*!
8
#)!#
0!
*!
45/./B- .
9D1
+K6 &'C
) ; 7
<?'
NC$ 'C
=
8D2"(
/;"'
#(
#'"
$
"7 ! (

$+*
+*(
'! E '2E7
$
,.4
"! !$I
'"
A!
'

0!
(
'
( 8
)!#C
0!
4B/BB./8, ; 7
<?*9'9
2&*5<!

N
9!
N)'<
)C9!D)4--5C!
0!
(G
(
)!#
0!
*!
/
#
+
'" '

G

!
4-/.44
#'2#
0!
CO! +@
#

'
, ; 7
<?3!;&-3
7
(
#
N
,-.-- +
' $
"
"#
((

#((
#$ '
9!3;"
)93;
93 )!#*
+*!
B.4
)K
*9B*
+
' ;!
F)
("; '"
4--B
$!((
N

' ( ##$ '"@
'
(!*2 ,

"

0!
'"
'
0

4-.44 *'
'
5-8B.5
;
&?&*

+),4"
3'
;B*!'!' ( #
!
'
#
0!
%6
' $+ !
)!

$+' (+'!'"+''2+#)!#
#C*'
'
/55,.5/-5/4
0!
*!
4845./
;
&?*
6&,1##
'# (
3'
;;9*!7
C; $!
9
( '
0!
I!
)!# # N$
N'!' "
'
#!(
N
*
+*!
/., '
7
!
+
7! ##)!#
0!
;
&?D
"
;*'"!9<4-- *!
44//./B8
@
' "
"
' ## +'" 3+2CK"+ K4--,!
#$ '#
+ !
%3
7
(
#$ ('
!
!'
$+"

+
7! #
012 33 4567895
87
!!"#$" @"M)":" E7"* /$$(C
%#&'%(( G8!6696
)"* +,"-./$$01234 !99.86>8!7637"0#"
5637658!679 D<(0'D<;$
5:"/(;"%/&0'%/0< @9"??":7".C+,,4" +%0(D
)=6">"?4"@A=94"./$$/ JG89T895B2G8
-!B839A7636!!653 989.86@+"/&"D0$'D0(
8=79C33"&%"/0;'D$D I,U6,"A"*7"1-U6"./$$%A2OJ
):B4"A "E6"1 "C36"F"2)"A* 86973!7G89@
E,".@/$$< 9838 +J63716";"<(%'<;%
576"G8E7637=2"&/" I,U6,"A"*7"1"=".A7" /$$<
D0;'<$D P56S8336973!7!
):B4"A "HB6">"*"*"I"- G8268@A216"/$"
E,".@/$$&J275," %;'D/
54E7637=2"0#" I,6,L="V=86"@/$$&E!L
#/;'#D/ 66W!!5895!!836
)B48"I"F87"?":B4"A*B"./$$; 286G88C8!@A
725867!87796 J637"//(";;0';&#
@K867E"%&"D;'#$ I,6,L="V=86"@/$$&5)9
)*"A"I"J"-266"+*"E/$$& 8953286!!3L3
E3G85899886"63L 5625793G867.86@
696!!,68!.86 C33"&<"<&$'<&&
JG8+8"D0"<($'<;0 I8".LX*,,86"A%00(J7!!L
)67"F*" 4"M-95"C/$$0 36 G8986J@+
6G767G8!!6! J637";"/%D'/%0
!8.86+7+8"/<"#&%'($$ ?"+".")"A9"?H9" /$$<
)4"=%0&0+7G8N636!!O !67L A6! 5 27 !
366.86JG8+8"/$"%'/; G8 !.86JG8 +8"D#"
)4"* "I"E)"J83"P))85"-A (/%'(DD
%000A6!589679L ?663"E%00#F9V+9!
89G8.86JG8+8"D$" @J637"<0"%00#
<#%'<;( ?8"."-":>" %00DE7!76
)8,"F*/$$% 286873 965B25268O !!6L
7!!!"87467G8 +2W6.86JG618"&<"
J@+J637"%/"%;'/< D#<'D(D
C9". P4"H*%0&#189686 ?83".=F66"E.%0&&J!!
98!.86JG8+8" G8T!6!754!.86
%("%#D'%(; JG8+8"%0"%$D'%%#
C453"Q"=R6"*"C49"+"?66"I"+3"* ?83"+-"=68".C"F6"AI"@"M)"
:"M/$$0C58986989 .".A*3"*E/$$0P5
6377G85896 ,7!59!6
A63"#$"%D%'%D& 23998@K867
C6B6"--M4".F%00/J!7!7 E"/$"<#('<($
!83"!3L9827.86 ?7"A.:67"+ %00D)79O
JG8+8"/D"D%#'DD( 758!6567 !!"/%"%'%(
C36"F":B4"A"E,".@"." *":" ?763"I>6")/$$% 278L
?=9"@/$$DJ6!89L 9698268G8986
!8!!S2G858 .86JG8+8"D/"/%0'/</
9L6 !!"<%"/;D'/&% ?72Y"F"F"*"J86"?)B48"I/$$D
C36"F"@BLJ49!"@ 2=6" /$$# E!963S269L
JS8!8666L 37!.86)7
3 2Q6=637"#$";D0';<( +"&("%%D$'%%D&
C36"F":B4"A "2=6" "E,".@"." 1"Z"?"*+"A-"),"*"
*=9"@/$$#5A63!6G8 M967"CM+9" -/$$;JG86L
!!E7637=2"&("%%%'%%; 7688OJ986!698L
C36"F"2M758"E *":B4"A "P9" 92868@K867E"%&"
C-1">8 93" P"=9"@"E,".@ &/#'&DD
2=6" /$$;J629! 1"Z"+"."A"-"?"*+"+9"
G8!!9L6 2Q6=637" M967"CM/$$0C6973!79L
#/"#%&'#/# 7986296
@7"CC%0(0Q5B2989G8 7!!!!6@K867E"
.86JG8+8"%"%0'D; /$"/$D'/%#
0121314516 789

!"#$% 7),)#9)#/)# 71A$$!:<!
"&&'('#)*+ ),!-.# C=D
/$,!0!1#&!2!3(4#!5" !5 !JA!0@@H) #$79)#7
"%$((%$ 6 )7($ 8) 7( 9 + #5#$(2 $!:!D:=D<<
4%#'9 5#$(/)# !:!;;<=;:: 7!1!#!+!")!!#*!+0#!
/(>?!'!"#!+!1#>!0
!5@@: @@"#$7# 7,#-'($$7
A#%#( 4$ $(#$#&7 -B7## 7#?$#%2!;!:;=::<
%4)#/)# 7'9)#A)!C<!<:= )!FN#O!
P@@<I77$77 6 (#B
/-!5<D'(#* 4%779)#/)# 7 $7G6M6#$(*$ $$#*-(#-()
'9)#A)!C;=C;E 7**$)4$#6/)# 71A$$!
F)#!F<1#8)$%$#4#($ %7,$B <!A;;=A:@D
)$$(-4)#-()# #G$ )$!"J!"#&!/1!24#- !F5#)$!
+
/)# 7'9)#A)!!;@=: @@1(%$# (% 4%1Q) %
F-,!A-!0E1,#$#7#*$$ (B "#7#$!C!@C=C
4$ (4*)7$)#/)# 7('9
)! )$$!"2'#!/D1# $#(
E@!=; 7)$ %777'#1A$$
F-,!A
)!0-!0B #(#% '$( 4%!;!=
7,9 7#*7#9 B-67,#$/)# 7 )%!6H !'@@:2$)$*!7#$)#
('9
)!EC!@D= ,(6#*#( 4%7#%$#%72$)
F-,!A!
!F-!0,' #4#$ #$ /)# 7'9)#A)!C<!=D@
$# /)# 7('9
)!EC!C:=C<C 0(#!'/!J( -#4(!"!3(4#!"" !5@
F4 )!H3A(#*!"EC'(#( 4$ $B 0)#4##47,#$ - #*$($
$-(?)4*7)#, %#B )#7$##"#H1,#$( #,)
, %7 #4/)# 7'9)#A)!D! 8) %$##'9 5#$(/)# !:@!<=<
<<=E 0#!/'(%,!2F@@@24#*$##)$B
F $!3A#%$(8))%79)# )# !#%#)* #$($#,

2/5( !1'9)#0)#* 9)#,* -$(7) '$( 4!CC!
"#$A#4#!-.# D<=DE
F $!3@@D'9)#1A 1!H +5+! 0)#!0F-,!AEEA)%7(*$($
-.# ##7-6 7,#$"#
1,#$*$(B
F*!/B/!.!AF*!I@@;A# ##%6 B $ ##( $(#$#&4-6 7,#$
)7$)# 7,#$)48) %'9 /)# 7('9
)!<!D;E=D<;
5#$(/)# !<;!DE=DD 0$F!0@@C'9)#1A*BA 1!
F*!IB/!+##4!HF!# %!3
!0!! H +5+!-.#
J !2/0#4#!0"@@"#$4 0 4#!0!+6 !NH+##!'@@:A#%
(#%9)#7$# $,#)4#)* I6 )'$(8)!1)#(I+5+!$5!
*)#*/)# 7'9)#A)!D@!D;<=DE 1
F( !!K4!L!K()!34!.@@:A),?$6 0 !3/!4 %!F50#!2D77$7#
,?$66 )7,#$(1#**)7$)## #* *#)##%*77$
$)*#
# -6'$($ +7#$! "(% 4%(6#!;:!:;;=:;E

'+@@:!=< 0$(!!)# !" 0#!A@@C+(-4
F!/+) #!IEC"#$4(9)#7 ,(6#, )7#*)#4*$7*
8)* 4*B 7/)# 71A$$! -(77#9)#2#$(67R# 4%!DE!
DE!=D C=@@
F!/+) #!IE<"%$((%$7M)7 0$(!"%#!02;7#$ %)#4
9)#
0-&!1'9)#
#)* *7(#6#)9)#$92#$(6
A#%0)#*0#$ 3#!- .#! 7R# 4%!D@!D;=DC
<=< 0DA*
4#R6#6-!AB;@E'(
)!!3 #)!/!)%!0$(!+@@E )#A-F $+*%
#)$9)#+## ,-#)* 0-&!E<1'9)#
#)* A#%
#%$(#$#&)4M(#G /)# 7 0)#*0#$ 3#!-.#
+# A$$!DE!CC=DC 0)S&!20E:36 * $79)#
- !'!')# !!/) !F!H#!" #7#$$ /)# 7A#%A)!!;;=EC
#!/:I77$74)#4)**$#$#% $ B R4,%!!2!/!A%B3-!I!4!0
) #%(#* ##7(4(7* 2)!5@@;A#%9)##G 46 *
7%*/)# 7'9)#A)!<!DC=;: 7##4$ 7#)%*/)# 7'9)#
!JI!
!A$(-&#!02!0#4!N0A((#! A)!C:!D<E=DEE

3+@2 %77$#)(4))#4*B "( !1!A$($(#!!2 (!3+3$#*!+
$' ) 6 )/)# 7'9)#A)! @@D'$ 7 4A#% % 9
!:;=<E 4'9 5#$(/)# !<D!@::=@<
!+05)##$$!2H2@@
*#6$#B " 4!0EC'(*$7#( 4%9)#1
# ,-#%#)* *)#* '$( 4%!6*,#EC!;D=:
012 33 4567895
87
!"!#$%& &$;YBH9941#2,G/%,")ME
'((')*"++,-./#,,01#/++ ,+$,L@V"WD22/"&+,-D/-++,0>,,"+<,/)
%,"/2,++"+/,3#+,#/,041# ::,,>!229H7B(F
&#"+55'667'89 &@@+$D&BH85:/++G,,01#/++
%,*::+,*";"+ ,#/,0>,,"&'9596759H
<=$>?".</4'((9;00,0+#* &@@+$ D & BH88 41# %+#%+ >,,"
#++,-@,,0+2++,0,+" 4,/,3-,IBH886'6676869
#+$,//>,,"+),/ABA9(7A98 &@@+$D&BH89:,/,+IL,I=."+*
@>"C=$D"E#=+!'((F!,""2* +,-1#%+#%+>,,"4,/,3-D#3#+BH89
"-,0,0G,00,,0,/%#,+>,," AH76B6576A
E-",,//,"+'BA('7A(9 &@@+$D&BH8H4L-+"L+,0,I=.1*
/"!#BHHB;%2/%,"/0,0I" #%+#%+:")+#,0>,,"4,/,3-
))&#I=.++++%41/+,#/8B ,#/'B6(7B68
FH79' &@@+$D&BHFB:,+#%L++,01#"
"+,J",,KDBHH5#/2/2-+/2* ,0,0,,"I#+)),#/,041#&#"+'
+=#"3%+,0%-1#,0%/$+"%+ BH87'(8
D+-,/,39AH'7B(B &@@+$D&BHF6?/>,,"+ 1# 2,G/
,%EBHH(:,+#%L++,00,,"1#D#+ .+"X-+2+2.,#/,041#&#"+
,#/,041#&#"+'B58575F5 B67BF
,=,>"<><'((9>#2,2+,0 &@@+$D&BHFH"./,2%+ +,/.*
2,,+2+),/,#/,0>,,"&" 3,+#%*,"1#2,I/%+>,,"4,/,3-
4,/,3-A5F6'7F8( T,IBHFH8B788
,/"D//2+!?",, BHHH;00+ &@@+$D&BHFHI:/++G,,0%,#0/*
,002/+,++,-2,2+,/,%/3 ++,0I.3+)M&%V"W>,,"41#"
""++,0%,#/,0K-&9' ,/,3-D"%JLN,$22B7'(
'(HA7'B(( &@@+$D&BHHB41#/22,+"0,,"Z#/-
,+/DBHHH>,," 41#M+#%" ,#/,0>,,"[#/-BAF6796
2,&23JLN,$ &@@+$D&'(('41#+++,-2,2->,,"
,++:>'((H&+,-+#%*%* [#/-"0B5'B67''6
04"+>,,"+"4,/,3-'(857F' &@@+$D&",#:BH8H&+,-./#,
,#"#?K%,:<//O+P%+:,/+" ,00,,"G%++,#/,041#&#"+B6'78H
!+'((':+2++MD/.L,++,- &@@+$D&"D">"%EEBH85
"%/+22,+4"+>,,"&" K./,2%,0+""3+/+0,%/2%*
4,/,3-8QF'BF7''F +,01#",/,IL,I=."
#I#%:'((F -:,+#+D003C ++,-%,"+,01#./#,,#/,0>,,"
,"-MD;12/,,-&#"-C4I/C+& &'95HF7A(5
4++R.+-,0:/0,K.+R&D &@@+$D&" ;!BHFB:,+#%L++
&/%$//,*/, !"34,#+# ,0"#"+,0,,"1#MD"#/+,#/,041#
D#,D"!S%S$!'((AD22/-3% &#"+''9(7'H6
/3%,"++#"-3/,+2+IL%,#* &@@+$D&" ;!BH9A41#,++
0/"%,+##,0,I",#/,041# ",%I,+MD./#",+#%I#,#/,0
&#"+56''67'6( 41#&#"+B6'9675(B
&/.",D</&@4">+@%&'((H &@@+$D&" /-KEBH85:,+#%L++
R"+"32,,2++2-1#I-+%#/,#+ ,01#",0,,"I#+>,,"4,/,3-BF
0#",#+%+#%+!C4*>,,"&" FA7FF
4,/,3-A'F857F8F &@@+$D&!,L"&$;YBHF6
&00%&BHFF>,,",3,I-/"/-,#/,0 :,+#%1#2,G/Z#,#/,0>,,"&
?,,/,3-5'69876H' A(B'657B'68
&L@JTBHF6D"2,,0++,-1#2,G/ &@"!'((H4/0I,%0,2",#+3
%,",+$2,"#+,#/,041#&#"+B 3++,/-++C&;D&4+,+,:,%2#+9
557AB 'H'75(B
&-%,#& "E%KKBH99:+2++"#* 42+;//%E++D"C=$D
++,0+/"/,L %,+#0,,"+,#/,041# @>"."!";'((H",,01*
&#"+BHFH7H6 #22,,0%-,++0,%,/,3/",./
&%BHFF&+,-2,2+,00,,"+LU,L +#%/%+#%+,#/,041#&#"+A(
)M???" $V"+W 9'7B(9
&+,-,2+,0>,,"D22/"&!,", 4-I,D%);+";3/+&
225(575B6 '(((",,0++,-1#,0,,$"2,,+#+3
&@@+$D&BH9F:,/3++,-L+#%/ #1B/,%2++,0"+2,+,2-"/,L
1#%+#%+X,..L,0"./,2%+ G/" EJ+2,+,2-!I+%/L+++0#"
,#/,041#&#"+B9B7B6 4,/,355B(57BBB
0121314516 789
B> KC 28&5 +)P
!"#$%%& ' ( /?1 )G6222394223&
) * +,- ./ ' B> KF 28&8 .Q
++ 0 .1 2&3334356 ))'/+ +1 )
.72889))):+ )+ !

; 2%928499$
+ ; +'.- B1GB288<
QR) +1 )
7.(<6222422< )G-( &<25$425<
=>+ ?1@28<<- ))' ? B1GB$%%6-++))
'. ) ' .! ; ))( (G -.225834&%6
7 '?'A/'A/ F71E ?"1?N2883-
)()' 1 )
; )S?' ) Q) + .!
&83<%43<& /1 )
;<56&46&3
B C-(D7 D1/ E FE/ D1/0 10A>
F-''>1E$%%5@ +.(. CFA>$%%9
; ))
7.(.C/ <8$$4$& ), . (!
BB
/->-1(EE1E " ? 1 )
; 96
1$%%8+); + +-/ 23842&8
)23%$&46% FE1 "EFA>"-$%%3 .
B 71G: 7/.1$%%5 + '!Q) . ) .!
) '.( ++ GT.
; +,
; . 7 ) 2525942&%
1 ) $%92249$% F>0A> C>$%%%
B 7H ' -#:-0 HD G ) ;
G
$%%9 . ; >+
(.2266$463%
?)) > 1 )
; F ( -C(7EEC
9692&499%
> $%%3))) .?(
B 7/ -G: $%%8@! .> - /)@ C(.323%43&
)) ' )C( ?(I F ( -C(7E
>
G" ,2%2%25JA) $%%8 $%%6"(?!'A/ ?(.
%<%%6 > 4++ - /)@ C(.
B> K28<& .) !) 6888242%%%
)++ +1 )G F ( -
> C(7E
3$29<426% $%%&)))': ?!
B> K28<&' + 4 ; 'A/ '.)?(.> - /)
'? . +,E">?: @ C(.3995849&3
LMG
; , . UFDC 17( 1G S0"V
0>> N?O >++2634255 @7-1S 1-$%%<"/?)
B> K28&8 + )), .. . '
7 LMG
; "(.-
.692939429&3
D O>?N

O28&%
B> K28<2"+)?(A(! R)? ) +))
) + 1 ) ; ,)? 1 )
; 2
G6&2$242$6 69&466$
B> K28<6 >,"+) . K ++$%%6
) .
A( A( (+ / +)+
1 )
; 233843< +1 ) .2896&4959
B> K28<6' + P)) K(0!FK(D$%%5"++!
+I1 )
; 2323&4259 )(; ) /))Q4
B> K28<3
+)+! /?
G
(.2&
(QA( ) + 22942$<
))1 )
; 23<3483
012345678
965
5655 54
!""#$%&&&"#$!'
(&"()(*)")(&$"+*),")#-##(++."$(,(/0+&"#!&
&(,,((!&"%()()$,+-'(+"$,+("("#)
&$&")-"(1+.&"%"(,+())((()&"%"((+.2)((&(
+#)(3"4,5","((,"/0))"(+$&")()#)
+#)!(*!$"&(%",")#-&""!1"&&1()%""$(
!"#,-()(!&)!",/
6789:;<8=>?@:AB8CD7EDE=8F9G>HG8AF9ECD:AI;EI<8C:J7D7ED=ED78=:9F;K>=DE9DAF??8=89J8C
;EI=8J8FG8:9A:8EDD89DF>9LMDFCN:FD8K>CCF<H8D>K=>A:J8CDEDFCDFJEHHICFB9FOJE9DAF??8=89J8CF9
K=8?8=89J8?>=K=>A:JDP7FJ77EG8HFDDH8K=EJDFJEHGEH:8<ICF;KH8F9J=8ECF9BD789:;<8=>@:AB8C
D7EDE=8:DFHFQ8AL
RS/T/U&#2VWXYWZ/
0[454 ~~h`bd]sahbkg~mhaqcmd̀agdm
~~aa`mjd̀udh
\]^\ _àbcdaèfa`cghihàjklmcda` ]n] hqhhà phqhh`ht~da`//////////// ]oo
\]ô phqhh`hrhstlhldhu ///////// ]nv h //////////////////// ]o]
Ww/x/W y&z({"$,("//////// ]nv
Ww/x/x |((" ///////////// ]nv
Ww/x/w U"${(#" /////////// ]n}
\]^] idg~mhpdhbphqhh`hrhd̀ ////// ]n}
Ww/w/W "$$))")# ////// ]n} 78
Ww/w/x U((+z(/////////// ]n 7 46[ ¡4[¢0[¡£56¤5[6¥
¦§2¨¡24[
Ww/w/w "*)($,+ ////////// ]n
Ww/w/ #+U((+0,,"$("/// ]n
Ww/w/ y&U,(+{("#'(+
y!/////////////// ]\n {"#$".'(+#(" ##(++.,"$)
\]^ a`qahbphqhh`h /////////// ]\\ "%()&)"&,")#)'+",$""4
\]^v h~mdhbphqhh`hrh///////// ]\] $#+(".+/0&$&(+(',")#
\]^} h~mdhba`qahbphqhh`h////// ]\] ,"".,&('##(++.1("%))"%"($(4
\]^ thhmhbhab/////////// ]\v (!(1+#1&"#!&&#"((+.(+".
Ww/Y/W (*! ////////////// ]\v /©#+.-&".! "++"%)1.
Ww/Y/x 0(+."(*)((/////// ]\}
Ww/Y/w z"U(+! ///////// ]\ ())"(+!)"&"#!&$(*(&/y&
Ww/Y/ ()!"()&
," ////////////// ]\ 1!) 1%"#$".()$(*4
\]^ fa`mcda` //////////////// ]on ")#( ! &!&(&".!(++.
)%&"))-"1())-,")#%&+
~~h`bd\sahbkg~mhaqhhddd
h~hhbphqhh`hrh_`mcbd̀h $( * (& $"#+.)"%&1())
aphqhh`ht~da`//////////// ]on ,")#V'(yª,()U&-WXXZ/0+"-
~~h`bdosrhpmhaphqhh`hrh //// ]o\ "#$".((+.&'!(")
S/y/(%+-S/S.$(-®89C>=I¯GEH:EDF>9>?°>>A-©"")UyU- «¬«
±W²/W²²Y³XY´4W4WX4µ´´4¶Ww-·U,!U¸z#¹)(-{x²W²
012 30 4567656896
6 8
(% ""%
!"#""$ "%""!#"%(
$"%""&'( " "&C ""
%""" """6"
$")"*"&'"( """ "
"%$#""%+ "D""",(
$""),"- ""!6"-&'""
$""!"&. &/"%
"""" $%" ""$!%
")""))&/!"" D"$")
""" ",7"".@FGG:-&.""
$"$% ")"")& ")"%""$"""(
.%0"%" ""6"&B
" $"!"% ! "%
!""%",1)"&2345-& 6""$
/"%"") ( ""!&
"""" '%"#)
$" 6" )""&H$(
") 0"%$%") "! "$
,7"23548923:;8. )"@&'%$ "
"&23<2-&') "") $"$% 0&'%"
)"""""%"% )",0)
0 " !""" ""-
")"" 6)")" "+ 0%
0"%)&="$%"6% 0"% &C$!%"
&>6""$$( "" ""%""
%""%) 6" "" "!")
""") &.%"%$
") ,1"&233;8 "%") )$%$
?@23<A-& ")%$)&>"
B """ 0"%)")""
""%)"( 7"25$"0"%
"" "( 6"""$%
",923:;-&B" ")"#&
"""&C '%)""#"(
$!&B """%""%$")
" "") $"!")")(
"") ""& ""*,I""123<48
="""$)( ."&234F8'""&233;8J"
"0"""( ?"2334-&/!""!
&=#D !!"%E"
"""(" %"" "&B
% & )""""(
1 % )"" !"""$"!%
" !&' "& "" ") "$("$
')" $% """ )""!(
"""%$"&& "&'"$+!#
)D"$" ""( ")")""$$"
""""E"!"6&> "%$"&.
0123 4567656896
6 656 1
!"#$% '+"(!(#+#
%&'()"#*( #(%-]%%%',%
+(!#"%%, 'i(-
#*( #(%-
.,'#%#,%( ##%
')%%!"')%(#$ JKLMLMjRZTRkTW[S
*(#')#%+"-.%/
#,%( # ##%-0 l# ##%'++%'
# %")#" ### #(%#(#+-m+#/
#1##%#++#/ #1'),% #(%###
% %"-., #!11-]#+#'+
+#++#%#' #(%'#(4(("#5%/
2/##%34#!5-6,#%/ (#1,%1!,%
' #%% +# ,#-.1,!n,#%')!%(
# ##%)!(##% % #('%! %%)+('
+##1'#%(%/ 1++%%#
#-7,#1& # #+%# %-on,# ##!%
%,- %#1')--)"(%1 #
#1'", #(%-o#1'
!n,#%'#%(%##
%-
89:;<=>?>=>@A>B>CDCEFG>=GH>I m# #+# ##%
++(2/##%3
JKLMLJNOPQRSTUVWXYRZTSW[ #% #-./
##% #+##)!(%/
%%"-0'#"+
\##%#%%++'% # !" #(%##%#('+#'%
!,, #(%#'+# #(%-0 (!)"!#*(##'#
##, #(%)1, # #/ 4]7.p)abbq5-r%% ##/
#%-0 4(#5" %-h,+#)#%##%,
1 #(%%(#- #%,%%(#s%#
\# ##%##" ' #(%%%%-])#%
-] (!%^_à%#! % #++%,###!'/
4# #%'ab"#c5!"d#'# %(#, ###+#
e "##,(#%+ # #(%-. ##"' ##%$
#(%#% #',*(/ 2,#)3--) #(%,%'$%"
#4f#g)^_à5-\## (# '# %(##+#
#!%(# %")!%("% -
,%(#, (#%'4%'' .#+!($%%" #
#+5)!%( !+"# ##%,#/#%/
#+%%%%-6+ (#4eg)^__tueg."#)^__t5-
#(!#!(' .%(#)",%(##
(+" !, #(%'#%+ ,#)%(#!" ## # ##%
%#%)!('!'" '++#!#(%(' #"!4
####-h,+#) ! "!!((#5-
#(%%(,%%)!(!(/ 6(#/
',41%,#"( #%(#",#1,,#
1#%!(+!# %(#4eg)^__t5-\# ##%
,+5-.#%' ##% ('%#%+!(%%("(
012 30 4567656896
6 8
!" #"$ %&'&&( U' !&-3&) 3)+)$3&&
)!)$&(&)!$")$*)+)$$) '''&>"&$ $&- !$%)! !,
$!$"&#+!),!"-'&)$'$&$$)&-'( !&3!)>"$&!?!$")$)&&-)(
+!)%) $&,")$!! .-/01& &9+)!+& ))$3!&-!'3)!"
02'& "'302')& $%'!=',)-),!,V&-&$!
&$&&9&")&#"$'!"-'!!)%!$&
4"&33&) 3)+)$$+!)'#$&$+!) '))&$&$,$!&$ )&9'&) &$&!$
%& !+')$'! $'!$"))5&! W!%9)'$!!)%!$&'))&$&$,
')$'!*)$!6778-9-)!"3$!+%& $!"$!'))&#"$!#9&% &3!$&&9
!$")$&)$&-9$!+)&'!)!&$!&%& &-'B3!+%'&$ ';'!++<
!), : '9%'&'')%$!( $?'3)T!)'$)$!$&&$ &+:"!
$&$3)+)+!)%&%&'!"'&$;!)&< -!$")$!"),B3&'#$&$+!)'&)
='&$'&>" &$"$$ &?'3)@$'!" : '!&R'!"-'&&$+&),!!!$+!) &-(
%&33&&!&!!&&$&%'&'''( !$&&+!)&!&)- B3$&9"&(&,
&$), !)&-&$!+9)&!"$&$ !++(A9!)$) !$"): $!$'!" )!-&V';+"VV&(
%")$!! $2&#677+!)!') $$<!+'&$&+!)&!'&# &+:"!
B3 &)3) '&,!),!#"$+"&,
& !+&3!)&$&!&-
CDEFEDGHIJKLMNOHPQ ?'!&$$ )&-$&&' &)&!!+
3)+)$+!)'3)!"#")!$&- !:
')&9 &++)$&3)+)!-'3)!(
R!!'!!!-&3)!#!$+)!' "$S!')%!)$')$"$-&9!&&&!!+
3&!! 3)+)$' !+$! '$&V!+'3)+)W!%9)&&$3!$$&#!
!')& !+$$!),$='&$$'!"#9!&S&$ )&9='")$!&$9"$+)!3)!3!)&!$-&9(
9),"%&$!$+!)3)+)'!&$+) &+( &-$!&&&!!+'-&"!+' &++)
+)$+!)B3='&$&$!)! +!) !-)!"3#$&$T!)B3.-/!3)
"#)!+)$!$T&)$'3)&&3$&'%! ''!&)-!+!!)3)+)!-"$
$$)!'!$!'$#$&$S3)+) !')-!+&$ &$&$+!)'&)"$&-
$$'3)&&3$)"$)$!+'3)!"%'& &&)$9"$#$ !"&33&) !3)(
& &$)&&&!$$3&$$)$) +!)"( &$!$R "$$'!% %&))-!+3)+)$!
&,!)& !$3)!")$ ,9")-! '&$#$&$
)"&),)&&-=' &$)&&&!3&$!
)3)$&9!+!$")$3 &&$"$",
!& !#$!='3'$&$&$ &$!3)!( XYZY[\]^_àb\c`dacè`c`fg`hìj\fk
9&&-$$&&9!!'%&+"&!$$+,
& &- &++)$2!3)&&3$) CDEDEClJmHIIJPnJnopHmJnMpJ
&,&+)!+& +!) &$)&&&!%'&
',)&&-)&9+)!+& !!&-
'3)!"$%'!"$", ))&-'!( S3&) 3)+)$$'3)&&3)&9$%!
&,&3)+)$$='&)')&$!-!! ! $3$='%!$3$)3)$ !'
$!"&!!'>"$&!!+%'! !%&'3)+)( 3&$$&"!"$, '3&$&$$ !
5"-$+)!!))9)$"$&!))3&$$ &&+,'$3'&$3)+))q+!$&(
&' &$)&&&!$.9&+ )"$ !, 3&+,' ,$&$ &)3)&!'$"#5
+)!'!$%'!-!'$!))$!!+') "$'!&+!) '!&'!"-'&&$3!$(
!$&,-"$$&-!))#,'.B"$&!!+ $&#!&"!(3)+)!3&!&$"$$ )
3&$$+)!!$")$!,!')#$&$' &'&$'3)T&-")r$'!%$$3$!)
'&)3)!""$-&$3!!)3)&=','9# $'%&'!"'!(3)+)!3&!='$$!),
$ #$ !'&)#&-)3)$&9$3 $3&&$$'!" $")''!$")3(
!+')--)!"3 &$")$$'$ $)&# #,'$!)$'
0121 3456789
4898 88884 1
ÿÿ!"!
#$ÿ%&$
'())))))))))))))))))))))))))))))))*!))))))))))))))))))))
+"!ÿ',(-))))ÿÿ))))ÿÿ))))ÿÿ)))) .""/ÿ0/ÿÿ)))ÿÿ)))ÿÿ)))
1"ÿ"ÿ2/,ÿ(/,!3ÿ4!3ÿ4!ÿ-/ÿ"!!$
1"ÿ!"!ÿ!3ÿ!4/ÿ"(1"ÿÿ!3ÿ/ÿ"!5ÿ/(ÿ1!ÿ!/ÿ$3!6
7/,ÿ(2ÿ8ÿ"ÿ(,3ÿ"ÿ2/,ÿ4/,1ÿ185ÿ-,!ÿ2/,ÿ(,"!ÿ/",(ÿ!ÿ1"!
31ÿ!3ÿ"(1ÿ/&6
9ÿ2/,ÿ3&ÿ2ÿ:,"!/"5ÿ1"ÿ"8ÿ!3ÿ"&ÿ/4
;<=>?@ÿBC@ÿDEFG@=ÿHIÿBC@ÿJKFL?@ÿMHEÿL=@I@=
N2/,ÿ(,"!ÿ(8ÿÿ3/O
PQR STU
+38ÿ2/,ÿ/ÿ2/,ÿ!!/6
1"ÿ!,ÿ2/,ÿ-11/!ÿ!3/,$3ÿ!3ÿ4/4ÿ!/ÿ!3ÿ"&
VWXYZ[YZ \]^^_àb]cdâe_f]d]gfahdfaeafaijaàk`lmai]jm_gjagke_fjahn
omad]gfahdfaeafaijaàk`m]kl̀_d_kkgpâkafqgir ]ih]j_tiàfp]^]ijgirkjmacaka`tde_fm̀a]ầfi]ỳ
kastaijakuv\w\vxnomakakastaijakkm_t^hpaf]iy gird_kg̀g_ik_em̀al̀_df_htjk̀n_iktcafkctk`pa
h_cgzah]jf_kkd]ia^gkk̀lg̀m]iast]îtcpaf_e fajftg̀ah]ihkjfaaiahk_m̀]`m̀a]faktg̀]pâe_fm̀a
d]ia^gkk̀fajagqgirag̀mafk]cdâv_fk]cdâ\{fkǹ àk`tkt]^êfastai`tkafk_em̀adf_htj`]fa]ddf_dfgy
omakàdkgika`g̀irtd]ihj_ihtjg̀ir]d]gfah ]àn~im̀gkàkẁj_iktcafk]fae_fjah`_c]a]jm_gjan
dfaeafaijaàk`]fakm_ligio]pâ|}n|n~`gk]^l]k akd_ihgirlg̀mi_dfaeafaija_fast]^^dfaeaffah
]ddf_dfg]à`_j_i{fcm̀aàk`_pajg̀qaklg̀mm̀aaih gki_`]i_dg̀_in
tkaf_em̀aàk`fakt^k̀noakg̀irj_ihg̀g_ikkm_t^h]^k_
pac]hajâ]fgiàfck_em̀a]c_ti`pagirkafqahw
àcdaf]t̀fa]ih_m̀af]kdajk̀_em̀admkgj]^kat̀dn
~`gklgka`_lfg̀ah_lim̀akaj_ihg̀g_ik]ihm̀adf_y
jahtfake_fkafqgir]k]k`]ih]fh_daf]g̀irdf_jahtfa _fd]gfahdfaeafaijacam̀_hkm̀adf_p]pg^g̀_ekaâjy
uxwk_m̀]`m̀ak`]eej_ihtjg̀irm̀aàk`m]k]jâ]f g̀_i_e]kdajg{jdf_htj`gk_iajm]ijagil̀_noma
tihafk`]ihgir_elm]``_h_nej_tfkawm̀ap]^^_`m]k it^^md_m̀akgkk`]àkm̀]`gim̀a^_irftiu]jf_kk]^^
`_padfad]fahwf]ih_c j_hak]kkgriah`_df_htjk̀w d_kkgpâfad^gj]g̀_ik]ihk]cdâk_eda_dâxlmaim̀a
012 30 4567656896
6 8
!"##$ $%$ SJ #+!*)$%$%$T!"*%$3
&'()##*+# ,$*!$-#+-*#)+-#*+3#2#*$%$$
# !*$.!#+,#/'"'-%$.!0%$!
!#"123+' OSJ$)#)+0%*#G"3!*)$%$%$
4'5"/#*+6-7+!*- 3%$3*$%$$ #*+
*$#!$-'1'
8'9$$!3#+# $!)#++' K3%$*!+#-UY $)3*#3*##+%#F
:';$!#+!*$' $#+%!233#+!+# #UZ/U[&1Z/U[41
<'$%$$!!#".!#+%0' \\\4Z&']$P#*+-3P#$)#)+0%,7!
='!!$)#+# $$/>?-?>1'
@'>"$#*3$ "# +#)+#*+
%"B3+$%$$"#*+7$#3$
!A+#' /&A41^/BY1A/8Y1/<Y1$<=A4<=$E'4&D'?#$*
B'C!' 3#33$)#)+0%T!!*/
D'>#+06$!+' $*#)*#+P#->P?1#
&E'C**!#$!++$F!$' 23$$# )#G#!
"%$3",#%$%$'H3,$
%33#2$3$)#)+0% "
!$+0"!+# 3#7#$%$%$ # !*3P$*_à_`bbcd̀bab"%$
",#-33$$*%#$3$!3#+
$!7$33$-!*$2++G#3 % 3 $)!*!#+)# 3$)#)+F
$!#.!#+!*)$%*'H3!3$)#)+F 07
0%3!++303IJE'<';**)$3# #+!#++#!
'!$P#*+-22!+ #+3#7
I-3$$3#2#$*#G"#%$ B-# B3! $)#)+0%""=!%B-@!%
#)!-$%$3$$22!+ $%$ 3#3)%B# # 3*#++"3$'e!#
$!7$#3$3!$+0"!+#F !*$# 37 * $0!*)$*%$+#$"F
'K3$$$%$$"3##*+ #+3!"33$#$"##$#3
$#$+0 )03#-
!$ ##'L#3*##++0-3*#0)2$#MN )*#+$)#)+/;+$+##)+'+-&
# 3#!P#
O/>1JO/?1J&A4'H32#+ $$ $#3#32#+B1'-H3 D@ F
3!03P$*$ G2233 0!3#7# #++3#$0$*-3#+
#*+2++)$%$$ )03!*$!+#' )+03#+3!+) !)+';**)$23#)7#F $
H3$$"3#2$Q#*#$%'H3 #$$$3#32++"2#0$
#+$#7303%$3#$ $%$ 33$# 32#+'H3$#F
2$#M#NO/>1RJ O/?1'H3$ ####+0# +# 3#+!+# 7#+!32H#)+&8'4'
)! )# 3)*#+-3F.!#$-$$*#+ H3#)+"73**#+7#+!%$##+"F
$)!-$7+0'
H3)*#+ $)!#++23$0F K%
,###%!%3!*)$%+'
#+ $*233$3$!+%3!02# +#$"3 )# !*)$$%$$"$!/3
$%32-''-3*#T$031.!#+
!3##+$233$3#+#!#++0 $
3# #$%$%$#*+7$33$'H3 $%$P'3#)+ 7#+!-3$#",#
%++2"%$*!+##++23P#$)#)+0%
,!*/)!37$#++$)#)+0%$3
3031'H3.!#"73$)#)+0%
,"ST!"*!%U)+-23#3# fghghgijklmnopqrstm
$)#)+0%F3#+%N
UW K##$ $%$!":<!*$#+-
OSJ/&V41U/UXS 1WSW /& 8'
&1 4:#+$%$$ #*+>'KH#)+&8'42,
3#3#)+7#+!%$:<!*$#+23#
23$ #+3#$$%<u8E'H37#+!+#$"$3#
4:# 3$%$3!*$#+ 3#7#
$%$%$#*+7$33$'v!#!*
UJ#+!*)$%T!"* 3## %%$!0-8<%<E!*$#+
0121 3456789
4898 88884 1
!"#" $%&#'()*+',#+'-./+%01!23%!4
5+'.'+'!3'
5+'.'++'-0%"5&'W/$'+0%"5&'XMT!Q%D&'9<M7YE'
6 ) 6 ) 6 )
2!-4C%44C'4%D&'$%&#'./+:85%!'&040%4%&5C%/.:Z
78 9: ;8 <= 988 ;9
0<<MQC'/D4%!'-$%&#'/.<:01+'%4'+4C%!4C0"!K
79 9; ;9 <= 98: ;> "#"%!-4C'+'./+'4C'3/!0#"'+0C%-%01!23%!4
77 9? ;7 >8 998 ;; 5+'.'+'!3'./+0%"5&'W/$'+0%"5&'XM
7< 9? ;< >8 99: ;=
7> 9@ ;> >9 978 ?7
7:
7;
9@
9=
;:
;;
>9
>7
97:
9<8
?>
?? [\]\]^_àbcdefgfh̀fhigdjkklmnhogfhmp`
7? 78 ;? >< 9<: @8
7@
7=
78
79
;@
;=
><
>>
9>8
9>:
@<
@:
/040'!0/+S05'3%&040#0'4C'0'0"5&'&//q#5
<8 79 ?8 >> 9:8 @@ 4%D&'0./+2!-!14C'01!23%!3'/.%4'04/#43/"'M
<9 77 ?9 >: 9:: =9 r'"'"D'+4C%44C'4'0404E/4%&'-%!-4C%4S/#3%!K
<7 7< ?7 >: 9;8 =< !/4#0'4C'4%D&'0./+4C'5%+'--..'+'!3'4'04YEC3C
<< 7< ?< >; 9;: =; %+'/!'4%&'-MT.S/#-/!/4C%$'4C'4%D&'0C%!-S/+
<> 7> ?> >; 9?8 == E0C4/3%&3#&%4'4C'5+/D%D&4S./+0/"'$%&#'!/4
<: 7> ?: >? 9?: 989 !4C'4%D&'YS/#3%!%&0/#0'4C'UK03/+'./+"#&%./+
<;
<?
7:
7:
?;
??
>@
>@
9@8
9@:
98>
98?
5+/5/+4/!00C/E! !FC%54'+>%!-W55'!-IX./+
<@ 7; ?@ >= 9=8 998 -03+"!%4/!4'040MQC'D!/"%&-04+D#4/!D'1!0
<= 7? ?= >= 9=: 997 4/1$'$%&#'0$'+S!'%+4C'!/+"%&-04+D#4/!./+
>8 7? @8 :8 788 99: &%+1'0%"5&'0s'0Y%!-0!3'"/043/!0#"'+4'040%+'
>9 7@ @9 :8 77: 97@ &%+1'(6t988*Y4C0%55+/%3C 0!'%+&S"%4C'"%4K
>7 7@ @7 :9 7:8 9>7 3%&&S',#$%&'!4MQC'./&&/E!1./+"#&%3%!D'#0'-
>< 7= @< :9 7?: 9:: 4/3%&3#&%4'4C'JK$%&#'%00/3%4'-E4C4C'+'0#&40/.%
>>
>:
7=
<8
@>
@:
:7
:<
<88
<7:
9;@
9@9
05'3235%+'-5+'.'+'!3'4'04(u4/!'%!-u-'&Y9=?@*M
>; <9 @; :< <:8 9=> QC'./+"#&%0D%0'-/!4C'4'04./+%-..'+'!3'/.4E/
>? <9 @? :> <?: 78? 5+/5/+4/!0(4C'/D0'+$'-5+/5/+4/!5+'.'++!1$'+K
>@ <7 @@ :> >88 779 0#04C''I5'34'-!#"D'+/+/!'KC%&.4C'0%"5&'Y%0
>= <7 @= :: >7: 7<> ./&&/E0B
:8 << =8 :: >:8 7>?
:9 <> =9 :; >?: 7;8 JL (v/D0wvNvx*Pw9
6
P76 L ()w6vN *w8M: (9<M7%*
6vx
:7 <> =7 :; :88 7?<
:< <: =< :? ::8 7== y+
:> <: => :? ;88 <7: ~
::
:;
<;
<;
=:
=;
:@
:=
;:8
?88
<:9
<?? JLz{)w6P7|w8M:} 8M:N6 (9<M7D*
:? <? =? := @88 >7=
:@ <? =@ ;8 =88 >@8 EC'+'
:= <@ == ;8 9888 :<7
A/4'0B6 04C'4/4%&!#"D'+/.3/!0#"'+0 )L !#"D'+/.5+'.'+'!3'#-1"'!40./+4C'"/04
) 04C'"!"#"+',#+'-!4C'&%+1'+/.4C'4E/0'1"'!40
FC/3'0./+3'- 5+'.'++'-0%"5&'
G%&#'0/.)E'+'3%&3#&%4'- !HI3'&.+/"4C'JK03/+'%55+/IK v/D0L)6
"%4/!4/4C'D!/"%&-04+D#4/! 6L4/4%&!#"D'+/.#-1"'!40(#0#%&&S4C'!#"D'+/.
G%&#N'0/.6%!-)!/40C/E!3%!D'3%&3#&%4'-.+/" )L 5%!'&040*
8M=@ 6O6P7O8M:
F%&3#&%4'-$%&#'0/.)"#04D'+/#!-'-#54/4C'!'%+'04EC/&' vL5+/D%D&4S/.3C//0!14C'"/045+'.'++'-0%"5&'
!4'1'+ DS3C%!3'
Q'040E4C6R78%+'!/4+'3/""'!-'-D#43+43%&$%&#'0 xL9v%!-./+5%+'-5+'.'+'!3'4'040BvLxL97
3%!D'./#!-DS+'.'+'!3'4/4C''I%34D!/"%&(3#"#&%4$'*
5+/D%D&4'0!Q%D&'T
G%&#'0%+'D%0'-/!4C'4E/K4%&'-UK03/+'/.9M=;./+VL8M8: QC'5+/D%D&4S%00/3%4'-E4C4C'5%+'-5+'.'+K
F+43%&"!"#"$%&#'0./+VL8M893%!D'./#!-!Q%D&' '!3'4'0404E/K4%&'-0/4C%4%UK$%&#'/.9M=; 0
012 10 345654578595
7
f"#'#-#%1#
!"#$%&"'#()* #%1&%'&
"%'#+%#, 32H4#&#%
-.$"#/#0.!"/.&1" 1!%"%
%# -#%'-##f###f##2:)4
/#%% #"-&1'-##f##
"#%#1%#%& 2:64!1""1-%#
"-#%&%#' ""#&#0%"#&&#&
0%2134#%!% 0&#=
%%#"%%&#'&
2/&##5.()674.%#%
#&"8 %##1%%#:!%## g@hiBjBDkmDhiDBjBGkm 2(7?4
1%%##"%9 "# l l
%###"-#%&
#%#';#"#
#-#%#% 2h41&"3%&-#&
0&#%#""1&%(<:.&0 "#2%#-#'J%"%>g-4.
&1/##1%'#0%- 2k4#-.#%.#
-+-=(4"%"--& '&2l4&#
0%-#>"-.:4 #2gnoH4j#%
"%%##1%%#24.74 #%"#"-##
-.?4-"10%-. 4& #2"%4p#0&.&+&
-&? f###f##2:)4
A
B2C(DE (4BD F: AB2C:DE:4BDF: ;##%#>.:)
9@ G E: 2(774 *% #&#%#7 %
:
E( q#'"#1r s-
#%t'#2.##
C(@"-#&"%%&( "#? # u4.%#&&
C:@"-#&"%%&: &f2(7?4
E( @ 0%#&"%%&(2# ,.%##1%%#:)
@H<:4 :)?.'-#'#?)6M"%#'
E:@0%#&"%%&:2H<: '-j-8):!##
'#4 &"1
!%%-9:#' v?2 4
&76?!"#-&0%76? l@ * @:7(
"'#+%#I
()*.%%J%!"#&J%#
%%1&&%1-# #g%%#%=
#'""#%##1
%%#2K./#0L.M%#L*4 i m i m
g@h B)::BD Dh DB):BG @:?
7( :7(
NOPOPQRSTUVTWXYZ[Y\T]^_àXbYZTcWT #-:?2gw4'-#%
RT\dXce ####
-r."#"###
/ 1 -#% &## &1 " ,#&###%#
&##%-#'%& %&"##'-##x()
0123 4567859
9896
100
0*%(&7.%0"%."!"'*)%#%4'"7("%'#$
'*("!'%.,%",$.)&F"'+&##"!"%6"!'&'#%." !+*$%1
!"#"!"$"%&&''("%&("'&$)*+"+&, &7% G#!%*,%")/%.&''&%*,%&&'%!"')6"+0/##"!
,&!"+ !"#"!"$"%"'%-&'&%"#%."#,$%%.,%&%$( %.,%.""
!"#"!"$"%&16"!/!0*'%H+&7&'
)"4&)),6&+(,F&7%.&'!"''"-"6"4&%.
)&$,%"'%.",,)/'&'$'&+"!,0)/12,/ !,$%&%&"!' ./'&$,))/& +"%&$,)',( )"'8I.,(,,+J,4)"''-
3*"'%&4."%."!&%&'4!%.%.""##!%154"6"!-&% ;<<CKI.,(,
(,/0"!"3*&!"++*"%%.")"7,)!"7*),%&'#!$),&( )%.*7.@!&+" 7"
%,)1-;<<LK2,!$.&',"%,)1-;<<M>1
(,4,'$!!"$%&'%,%&7%.,%%."
'*0'%,%&,%&89:2-;<<=>1:."'&?"#%." !"#"!"$"%& ##"!',++&%&,)&#!(,%&-%."
!"#"!"$"!"''"7!* $*)+0"*'"#*)&#!(,%& !"''"%&+"' %&#,$%!"')6"%."3*"'
&&%'4!&7.%1)'-'(".,6" !'"+%.,%&%$*)+ %&#'%,0)"'"7("%'1N
.") +"$&+"&#,"3*,) !"#"!"$"')&%4,'+*"% !"''"(,/0"!"3*&!"+0 "%.")"''-%."!"#"!"$"
&+&##"!"%!"''"!4."%."!%."!"(&7.%&#,$%0" %."%"'%!0"&$)*+"+#!%./' (""!"3*"'%&7
'%,0)"7!*'#,0*%"3*,)'&?"4&%.'%!7 !"#"! 89:2-;<<=>10,))%#!,"!!) "7,)$'&+"!,%&'
"$"'8@!&+7"(,-ABCB>1D%."!4!+'-,C<EC<')&% !"#"!"$"%&&''.4&O&7!"1A$M"%
"#"
1;
"'%4&%.%."
1
&, !"#"!"$"%"'%&'$)",!4&#!"&%."! !+*$%-
PQRSTUÿWSTXTSTYZTÿ[T\[
]SQY^Tÿ_T`TSQ^T
aQbTccccccccccccccccccccccccccccccccdQ[Tcccccccccccccccccccc
eT\[TSÿafbgTSccccÿÿccccÿÿccccÿÿcccc hT\\RiYÿjiUTÿÿcccÿÿcccÿÿccc
PkTQ\TÿSRY\TÿlifSÿbif[mÿnR[mÿnQ[TSÿgTXiSTÿ\[QS[RY^
PkTQ\Tÿ[Q\[Tÿ[mTÿ[niÿ\QbWkT\ÿRYÿ[mTÿiSUTSÿWST\TY[TUoÿXSibÿkTX[ÿ[iÿSR^m[p
qifÿbQlÿUSRYrÿQ\ÿbfZmÿQ\ÿlifÿnifkUÿkRrToÿgf[ÿlifÿbf\[ÿZiY\fbTÿQ[ÿkTQ\[
mQkXÿ[mTÿ\QbWkTÿWSi`RUTUp
sXÿlifÿmQ`TÿQYlÿtfT\[RiY\oÿWkTQ\TÿQ\rÿ[mTÿ\TS`TSÿYin
uvwxywÿ{|}{~xwÿÿwww|~wÿÿ
{~v{|ÿ|wÿÿwÿvv{|ÿwwÿx|ywy

aiÿPSTXTSTYZT
,))%#!, ,&!"+ emQYrÿlifÿXiSÿlifSÿWQS[RZRWQ[RiYp
!"#"!"$"%"'%4.",$.&$" PkTQ\TÿST[fSYÿlifSÿgQkki[ÿ[mSif^mÿ[mTÿnRYUinÿ[iÿ[mTÿ\TS`TS
&'%#!$"+,+,
!"#"!"$"!"''"&'
,))4"+1
012 10 345654578595
7
8&)*%+.
!"#$
%%&"'$%&"($ 0-8)*
)*+&",$ *&%+)
%+-%
+% -W)%9
.-/0% "XY#55$%.
&+%1 *+"'5:$-%
2+&%3 )*
%45645 ` a\'],^"Xb^$c
-7+++.
"--&%%$%3 \ ]'^_ \' X
'
%- Z8[ 'dX]\'e "#(-4$
8)&+%+-
+% ^%.
.%- &X%.&
%9 \';*'4-SS-M3
))- 0%M0(
%.% -
.+&+#5:- M.%&
8& 3%
+.*'5:&.) 0V&&.
%%&%+.; +-
)!1<=>?@AB>CDDEFGDCCH?@ .0%
IFGDJDGD?KD&) 1;2"f-&
L2"MNO&'55P$- $-6
9 'g55,%" %&h3$-
-
8& N%Z4%%i
-9&g%"g$&
-Q+%+
%456450- ; !
.*.+
%%9.
%456450.+ l
-M* j#["bbgi$k ' "#(-R$
*0
-&R56,5%
0&* j'["]bgi$kl' "#(-R.$
.%
R56,5-.%) j# j
%Q3+"#SRT$& Mj' j%
0 MI?g*-
*.%. *; 3&
-U*& .*g-Qg.i&
i i
).;%+.* j.%g*
9"MQVO+&'554$-8% )+%9l-i*
*%+&+ "m$%* nkX-j[g6m&
%.%V& %.#-SR-.3
**+*) *)i-n*"
"MNO&'55P$- f&'55R$g*.Q-
0123 456789
5 9555595 101
!"#"$%&& )*3))*15(+8/.5)*/_P;/(1>143)14-)*/0/+0(/8*+
581)2*/72*+12/5G<
'()*+,-*./0(123)1+4154+)+6)/47+4/81)*0./6/./42/ 3475'A3.1/);+6+)*/.300.+32*/5*3A/P//45,--/5)/79
)/5)59)*/./3./36/8-++7./35+45)+2+4517/../0(1: N+2*./3A4///.3(3././A1/8/714)*//U)/451A/030/.P;
23)14-14)*/7/51-4+635/45+.;)/5)<=1.5)9)*//66+.) 7152,55/7P)3 (<CDEEIG<'P/)3:P14+>13(300.+32*15
3472+5))+./2.,1)952.//49347-/)32+45,>/.034/(15) 343(;515,5/781)D*.
;V1C EE`G951>1(3.)+)*/P/)3:P14+>13(
+4:51)/1563.(3.-/.)*34)*/)1>/3472+5)355+213)/7 300.+32*15146+.>3)1A//3 0(123)/77166/./42/)/5)5<L*15
51)4+)+4(;-1A/55+>/+A/.:
81)*)*/2+47,2)+6)*/32),3()/5)<?+8*;4+)-/) 3((5)3)15)123(51-41M2342/( /A/(P,)3(5+23(2,(3)/5)*/
5+>/3771)1+43(146+.>3)1+48*1(/)*/2+45,>/.53./ -3>>35)3)15)128*12*5*+85*
0./5/4)@?/2+479)*/./15/A17/42/)*3)>34;0/+0(/ 515)/4)(;-.+,0/7)*/034/(30+08. / 3
347+>A/.5,52+4:
.
5)+P/P/*3A14-<
81((2*34-/)*/1.>14756.+>).13()+).13(<B+5)/./)3(< T6)*/./3./5)3P(/5/->/4)581)*2 +4515)/4)0./6/.:
CDEEFG5,>>3.1H/75+>/73)381)*2*1(7./43474+A/( /42/59351-41M234)C4+4:H/.+G-3>>35 )3)15)122+,(7
6++753476+,479+4)*/3A/.3-/9(/55)*34IEJ2+4515: P/+P)314/7<=,.)*/.146+.>3)1+4+4-3>>33
)/42;6.+>).13()+).13(14)*/>+5)(1K/7C2*+5/4G6++7 P/)3:P14+>13(15-1A/414N*30)/.Q<'(5+9)*1534437) */
3475+>/8*3)*1-*/.(/A/(581)*37,()5<L*15./5,()15 6+.>3((;./2+-41H/5)*3))*/73)3+4)*/)8+).13(53.1/5 (;5
51>1(3.)+M4714-5+6N*30>34347O38(/55CDEEIG ./(3)/79347)*3)1471A17,3(2+4515)/42;+.A3.13)1+415
8*+6+,473P+,)QIJ 581)2*14-143)/5)+6>1(K5 03.)+6)*/012),./C1</<9a+A/.7150/.51+4bG<T4)*/4/U)
)*3)3((+8/7)*/4+:0./6/./42/+0)1+493()*+,-*)*/ 5/2)1+48/81((5//34300.+32*)*3)2+>P14/5P+)*
>3.-143(0.+0+.)1+45+6)*/-.+,05)*3)0./6/../7/32* ./0(123)1+43473((+8342/6+.)*/4+:0./6/./42/+0)1+4<
>1(K./>314/75)3P(/<=143((;9./0/3)/7)/5)14->3;P/
)*/+4(;83;)+3458/.)*/R,/5)1+4+68*/)*/.3IESIE
50(1)140./6/./42/./0./5/4)5/R,3(300/3(+6)*/)8+
0.+7,2)5C+.(32K+60./6/./42/G+.8*/)*/.)*/./3./ c de$f#e" !"#"$
)8+5)3P(/5/->/4)5+6/R,3(51H/0./6/..14-/32*A/.:
51+4+6)*/0.+7,2)<T4+)*/.8+.759./0/3)/7)/5)14- g/U)8/81(((++K3)0/.*305)*/>+5)2+>0(123)/751):
2+,(7;1/(7/A17/42/6+.5)3P(/5/->/4)59148*12*235/ ,3)1+4914A+(A14-P+)*./0(123)1+4347)*/,5/+6)*/
P+)*A/.51+45+6)*/0.+7,2)P/2+>/2347173)/56+. 4+:0./6/./42/./50+45/+0)1+4<'51>0(/;/)/(/-34)
6,.)*/.2+4517/.3)1+497/A/(+0>/4)9347>3.K/)14-< 300.+32*)+)*/155,/+65)3P(/0./6/./42/5/->/4)5
'43(;515+6)*/5/73)3234P/51>0(/+.>+./2+>: +,47143^hIY030/.P;W/+.-/i<=/..15C3(5+
0(/U<'51>0(/300.+32*15)+2+4517/.)*//U0/2)/7 1756
A3(,/+4/8+,(7+P)314+43./0(123)/7)/5)160/+0(/ )*1/j152,55/714V19DEE`G<'43()/.43)1A/343(;515,5/5
.12*(/):>,()14+>13(Cj[G>+7/(934/U)/451+4
8/./P/*3A14-.347+>(;C1</<9*374+./3(5)3P(/0./6: +6)*/P /)3:P14+>13(300.+32*)+)*/>,()14+>13(51):
/./42/G<=+./U3>0(/9+43)/5)81)*)8+).13(59+4/ ,3)1+4<L* 15157152,55/714W32,(3/)3(<CDEEhG347
8+,(7/U0/2)DIJ+6)*/2+45,>/.5)+2*++5/0.+7: 1((,5).3)/781
,2)')812/9DIJ)+2*++5/0.+7,2)V)812/9347IEJ 3)).32)1A/P/2)3*38+ .K/7/U3>0(/<L*/j[ >+7/(15
)+2*++5/+4/+6/32*<W1A/4)*/5//U0/2)/76./R,/4: 2+45,>/.-.+,091</<9+8*
,5/1) 2 4517/.5)*/*/)/.+-/4/1);+6)*/
21/59351>0(/2*1:5R,3./343(;515234P/0/.6+.>/7 /)5+61471A17,3(581)*2+/4)*5/15.))/*4/).0/3./7166/./4)0+2K:
)+5//16)*/./5,()57166/.6.+>2*342//U0/2)3)1+45 32.+55./0(123)1+459>,2*(1K/)*/-3)>>35 3)/.45+../50+45/
)3)15)12+6
CX3.K/./)3(<9DEEYZ[/;4/.59DEE\G<L*/300.+32* )*/P/)3:P14+>13(<
234+PA1+,5(;P//U)/47/7)+)/5)581)*>+./)*34)8+ =/..1523((/7*15300.+32*)*/k:A151)>/)*+7+6
).13(5<T4)*/235/+6)8+).13(5)*/73)32343(5+P/ 2+45,>/.)/5)14-347)*//U3>0(/5*/-1A/53./14
235)143D]D2+4)14-/42;)3P(/C5*+814-)*/0./6/.: 5/R,/4)13(*+>/,5/)/5)56+.031./70./6/./42/981)*
/42/5+4).13(^14)*/2+(,>45347).13(D14)*/.+85G 34+:0./6/./42/+0)1+4C6+.+,.0,.0+5/59klDG<T4
8*12*2(3551M/5/32*2+45,>/.14)++4/+66+,.2/((5< +)*/.8+.759)*/./8/./3)(/35))8+5/03.3)/)/5)5
'2*1:5R,3./343(;515234P/2+47,2)/7+4)*15)3P(/35 2+47,2)/7,514-)*/53>/2+45,>/.5981)*30./6/.:
8/((<T471A17,3(P14+>13()/5)5+40.+0+.)1+452343(5+ /42/2*+12/P/)8//40.+7,2)5a'b347aV<bL*/./
P/7+4/)+2+>03./)*/-.+,02+4515)/4)(;0./6/..14- 3./5/A/.3(./35+43P(/355,>0)1+45>37/98*12*3./
')+)*3)2+4515)/4)(;0./6/..14-0.+7,2)VCP,)4+)/ 356+((+85mC^G0/.5+4581)*32+4515)/4)0./6/./42/
012 10 345654578595
7
#
!"###$ .H%-&/
%&'$& (!;;I".BJ
##$($)$* &9'$$".2
# +"$& " $##$(&
#$$$& $(&$&.2BKL&
#","#&- #&#$
$..('"- 3=$&>.B
$$&&$&)$'* $#&#
$&-#./-# 20+.+".
-#&0"$$&# MNOPQRSTS 4&'&*&$
'1(!"$$& $&
#'1(+" 20$
$$&%'&%$$*
*$&(*#(* 2!$ C& 4$& C&
".2&'##$ C&
&-3.4 4$& BB B
B
B
B
&'5&&& C& B

B B
('678)(
9$ U@
-%'&#-#.
2%&#- # =%> .
'./($$ 0"22
%$&3(-

$#&&

.
#:.(!;;+".
/)#$(<#-1$( BVW BXBXBXB 0+.Y"
-&$(&-'(*
(.(<'' #$$-&#$&"
#<$=$&.>6(
$'-&# & BZWBXB 0+.Y"
-(.-.($$#%'&
13(#'-.2( #$$-$("
$$&**- [WB\B\B 0+.Y"
##"$$%
#$-&#'1 -$&"
$-'1$$%#& !"U'$#(].
##$#$'.<&
'&(# ^_ !
#$($$&*#1- [\ [ \ BXBV`!"!BZXBV"a
$$. ]W !BXBV
?)$#9@< 0+.b"
-3$&A<-# +"/-#)## 1#(3
%&=>A2-B $$c c c.2
#($$&$ $$#'-
#$=C&D(>=C&EF> &# '.4c 9
$&9.2#$% BdB.<--5
&&# ] ##&
$.2* B$#'%#
+G+( *$&-#5'-
0123 4567897
57 756 10
! ,'"##(*G()#,
"##$%&"'()#'* AV:11($!+AV011
+#$$$)() .26W$((&$#G#$)#"*
#(,)&-,)+./&01123 ',#!)%#&$
4)#&01156#$# ((#####$,$)$#
)#)+&7&89.:5;6<,# #(,)>"KRX##('*
+&)""##(#$#,#$ ")##)#$,#(,)
(##("#$)#)+!+##* "#,(#,T*"
"#,#$#*$)'#( !)&"&=> ? 1KR.KRX "
#,#$)"#( G6
@A.:B706CB@.ABA!!670C T? =U> B1KR .:5:06
=> ? A.:B0706 .:5D6 N.=>6
@A!!.:B706CB@.ABA670C S)##',*
=E ? A.:B0706 .:5D!6 !+T
()G#$#*$)"
*,"=# (!)%#*
&"')#(()
#*$)",!#!#
=#?:B=> B=E .:5D)6 01XS"#')#((
F#(-,(###=> ( ),(('#$%#&$=#
=E +# #"'", 5RX& 01X (=> (=E !#"
$#G#$)#"&#+ #(,)
"#"#'##$
AHA#A!!HA)####$,#* >#(#$+$#%
(+I)#J)#,#"( Y*'#(#>(:(#$
#"(#!!+(!+#* )
(##(##*$)#(+"#

.K6#&##(#")#*
L()((##)"& Z[\]^_àbcade_afga_`hfi
(#'!+&##F&)),*
#')()#')," jklmljnopqrps
)),(#$M
= >.:B= > 6O.5= #706P0 W)#,(# '
N.=>6? A .:5:16 #(,)()("#)("#(#$
P $)#"),,+#
N.=E6? =E.:B=E6AO.5=#760 .:5:1!6 #(#'"&,#(
0
.= 70P06B.=>=E6 ,,+$#)()#)($)
/QN.=>&=E6? #
A .:5:1)6 ))#$$)"&*
)(# #+#F#
.R6F#)$$##(,)># +($"$(
##"'()#$G#$
)$)&#,"#!#
$##(,)ES$#($$)#$=> ', ('# ,#)',$###*
=E,"T*M #<$)",'+$#
= > B=
T? UN.=>6ON.=E6B0/#'=>=E #
E ) ,&)!(#9,)+('+
$$#!(#$#$$*
.:5::6 )#)L)#$#(,)(
012 10 345654578595
7
234546789:;<=<>?@98ABCD9E9
!"!#
!!!!"" "$ -"%
"#$% F!# $G$
!!!&"! %H#$.H#*/IIJ$1J
#"$""% $K(L).*/IM1.-$N1
# "$' ).OPJ*/I01-
$!"( $Q"(!#
)*++,"!!# #-)
!$-"" !$!"
.*//01 R,"!(
STUVWXVX P"$!
#
0123 4567897
57 756 103
!"# aNM'NO(*
!"$%&!'()$*!!+!"#!"!%, _E>;2` B '(C-(D*
#&%-"!" !"%#
%!!!.#"%-/0! +"
!!#.!!'*-1!"0,
%0.&%&!. N2#%
+!"(2%#&!&32%#&!- M2#%!&
45++&#& `!"!,̀!b2 cd &&#
.!&5+!"50%!!!"5 #&2(
0%!!!!"!!&+!"
!"!-1!"0%.!!#!"%&! !!!"0%%5&!0%
6!""7#+8 !"#!"/+&e=:.
!+++ !"&!=&
9&! 6 : 4 ; < = 7 !
!-=@!&!!"@!!#&##,
H 2(c-DC-"!H ,#f!cd
>.!! (? @? @A (A @B C@ CD &0&##&(@-cg-"#!"
EF % #.#!"&!!&##F!5
%
!hicd-j+&!&!+""%&,
"!&!!"!!"!&## #!+ .&F!5""%#
#0&%&!@@-9&! '_E !"-"!F!.&##,
+!"!"!!+ !". ! <G-'(C-(>;*# D*
- 1
!"=&!!!&#
!"0%!"_E>;2(D-3
F!5&##!5!"!!-9&!;!" !"#+
F!5%#&!9&!7+!"!" %#8 %!!#!!!&##
%#&!"%-
"=&!!!"%!G!
!!"!+,+55# +!"!!@, 9&! 6 : 4 ; < = 7
!-"=&!!G!&!"H >.!! (? @? @A (A @B C@ CD
&!!8 EF
IJ %
Q N VXY
@ RS UT@W [CM'NO(* 9&!"!"F%!!
H@2KLM'N*'(NO( *P T2( Z
'(C-(C* "
+&##.&%#-"%!!
k&#+89&!;F,
+" !5%#&!"%&!:<=&7-
9&!7F!5%#&!"%&,
N2#% !6&;-/!!"!!"!#!":.
M2#%! !+!!"!"!#!"
UT2.!!& @ =&!!-E"!&##+!
&##&#H 2'N)(* !"!+%%"-
\+&!!"!!"H@ !!F!
!"%#.!!%!&! lmnonmpqrstuvwrsxyz{|}~
&!+""%&##%##
!"-1#5+"&!" !"!&!,
!"F!&##%#. :!)+!'&0&##
!"]!F!.&&##^_E>; 0,&##*!"G+""!"!+%&,
5!!"_E;!!&!!!&### !&!"%"!""
#!5#- ".!&!""".!"!-
012 10 345654578595
7
$1""&"&5
! 4"
"" '4"
#$%"& 0#"!&""";
#"#& "!"5
!"'( $@4#&
$)**+,-. )**/,-$ ! &5
)**+0$%!""" '/5&
"! "'270 .3A05
""'12 "'-.)**/,
3//)0$%"4" "!5 -$)**+0$%46"&!"5
&6" /527
5 #&!($')**+0$(4
" 4 &445
"$(4"!4#44! !&9
#4'7$)**/0$ "&;B"5
%"4#48 ""
"4 4/5""#!5
9"&"$: $%"&
;"5 4&C4"!"!&
4&"& ""$
"'-$)**+0$%" %""
"45 $1"!&45
4#&4# ! 4'!
$%44! "05
$%"""&" $< "&#
4#&&" &&4"!5
"&"4#$%5 &"&
"! 4&"$:
4&&"! !-$')**+0
!"$ &C4"
!6&!"" "5
!44!5 $%4&6""5
"4" 4"""'-
!&$:" $)**+0$%4&6"
" !&4B
""! "'.3+ ;
'!4 "0$
!&&0$<12
'3//)0!""4
"&$
=>?5!# DEFGFHI J
JKLMNLOPLLQRSPLRLPLTUL
TMVKWLPVXKYQTZ

"""5
$%!4&#9 4&
""&"&5 #444&[!&
"4 "!&!$\
"&"4"65
&" '13/]^0$%"
0123 4567897
57 756 10
!""!"#$%&'("))*"+!+"!"#$#)#+"!"#$'#'%&),
)%# '#)'$(%"#*(!($#%-".%'#+ '# (#> $("8''%'"# ? (!"+

'#,/('""!"#$%'*(%""(0 @-#$0$-#%'#$($*"0#)# =-% (#
'*( (1%"#*!1"'#* ("#'1%, %$"8'#$%(""-#>$)"#%!"%#%
/* ('%'#!"'#"'#*%$$-)-%) A"(%&*"-%(%"#*0"!")'!!"#")+
%%(&#'#2'*,34,5,6!(%%'7'%"*8% -$%B,$(!!.A3CDEB"%)##0%'%!1"'#$
''%-%-0'#$#%-"%8($('$%$# )!"(%?'#)%!),
"#%$"')%94 :4#) ("%-'#*)'%"'-'# <#("'# ( '%%'%-%)'%F)'%'?
%)*'#%-'#1-!7"0%' (@-0G#)F'?(@-0G%1"''#
;+%,<%'# ($%!(=-%+"'*( %$)'%$-%%) !(#+"!"#$'#,/('%1'-%0"1')%
012 30 4567656896
6 8
! HIIJ2K7L,MN1OPJKQLRSI4J1TUVJ
" WJOO76XYZ767U[JIJRUJK\OJTJOJ23J]J6U
#$!%"%&! ^2345K72_UVJ`1Y\OJTJOJ23JaIU712
"'#(")$"$%*
%*$*%"
+'+'#$# &"$b;;%"%
$$'
c9CdE='>"#'%!
"%"&%"efCgE=%"
,-./01234567126 $'%"'h:?fCi %
jCdb:kEeCg;;lE@%%mgnmlE
8"'%#%
""%"&""% 8& o%" 8e
%"$"+# 8&
'$9&$!""' o%" ppqh :m pqdg
qdh pqgh
pqdh
pqdm
%":;<:;% 8e pqk pqdd pqrhr
#%=%"'>"#'
%"'%":;?@$" Cpqb;;E
%"#*$*% B"A$9
"
$"%"@$*@"%"! psqptptptp
">(!%")% puqptp
!"%"+ vqpnpnp
*'""'!
#'"#AB$*@"%" y
#%"*% wq vx zvgxCp{ps|gECgpu{psE}
@%& "" gp{ps
"%"
%"C%%D"E#$ ~9
%*""'%" vqd;;C$'&&ee*@
F""%%* @"""$%!
!%"%@*$%#"#" "E
%"@#A
""%" %"
F%"!" ppusqktd hqgg
!%"%=""! wq;g:m gtdmtddq:h
qd htd
#%"""="
$#*%'$ o$$A
%"@%""*" "'<%%9
$'$'##
$*%"$ zpCdxwgE}xzCpxpEwg}
''""G & q
#%"%"*@"! zpCdxwpC
g
dxg
E}x zwgE }
Cpxp Ewg
'!# e q

"#" pCdxgwgE
$'*'*'%!
""#"' *
%'$%%" qdx& xe

011234567 879

JKKLMNOPQRSTLUVWXYLZ[\V]L^L]LMYL
SL_`
!!" # $ $#%
&'!'!" # $ $#%& C, ,%#$% )#) -
(''(''" )# $ $# $ )3)$0a%
)+8!bc )+8B4d, #,/))/% 0

*+ %, -.)/%- 0 #)
)
)
#) *# %%$$ +, ,
)#) .+, %) - 0
$, 10 , *# %, # -) C, /+,
3(5 463 ' 78498 )%#-% . ) )%+
234 ,,# $ , # $ 3,0
: # $ #4)%. %e - +,
$$ %$,% )0
3(5 463 ' 7 9
848 - )/))% , )/.
23&4 & & #) %G, +, /
: #$ ),)%% , # )+%%0
))/+% # , /
3 7 9
8 8453&4 % %)/- , %$,
;<23+&4
:
fgghigjkl mno7pqmjrqnkgd, $ 0
# $ ### %. )
2?348=3 !">(!"+(! , # )
.) $DED# $ )+,,
% )$.) $
8=4 , , - 3# ,#4# $ $ ,$
23&48! , - + # - )/e , %
;<23@+@&4A(B $$ 0# $ #,, ,
$, # $ 3, ,.) 0

$, ,/#, 4 #% ,)%#$, 0# $
# e - +# - % )#)
5 ))-%, 0# $ #!AB" $,
&
1 ?234623&458;-& +% % ) ,,# F
, H $.) .
. %./,##,s-% $.) 0

,,)) %$$ $, #) )%. $%./ #) % ./0
.) $# %# $ C, #) H
-

-% $%$$ I%$ %E

, ) $, ##)) tnggqupijkjpvgigC,. #w )0
-) $DC, ))/ $%$$ 0 , # *# # $ $ $
$, ##+ % *# , % )# . +, %,,
DED#)/ ,, 0# $ # #% . %e - +,)% )
F, 1$ $- &- 1=!+ !AB"$, C ))/%-.)
.-$#%
G))/+ H %##0 ## 0 ) /, %,,
%#+%,##, )
. ,I %-%), $#$, # 0
5 D - ##) / ) , )

1 ?234 F,# /. x. %y/%/$0
)/s- % )#) / %$$
10 $- . ,I$D3D"4 $ G, +%-%)
011 20 345654578595
7
! (%<,
" ,",(
#$% %6)*!,,,,,
&'() ,'%&+,
*!+)*! ((%&+
(,(,%- (,+!
(, )!
!(.'/(0.!(! %
)*!!%
1(23%4
"*!% EF<:B
5, 63+-13+73$"
*!"89.!$%5 E+ 16 - 17
,: 6:+-1:+7:$(+) <3B
*! -
63
< 3
-13 73
!$% 1 6: -1: 7:
<3
16 - 17
; 63 -13 73

< 6: -1: 7: GHHIJKLMNOPQRSIKTMUVHWIQX
1(2:%<!) YZ W[LJQVLUWGHHRQU\][QGJUW^_LJàU[U
bL[][]IcQdeRIXIRIJ\IfH[LQJ
(%
&!/(0
+ < ! (!
:8=(+ > ' /)0 +
,,,,') g!( +9h3ii+
!"$(% ),! j:ik$
<!16++17% l!(;!+3mnA+%3m=+D*%:%m$$%
6)*!,,!,) 4,>'(!
, '(
)(
'!,"/0 t u: w
%<'' p :q:rs p: p qAr9vr$
"(,! 5&o x y 9
> : 9qp:
%
&>*!+!.!) ,
>(,,%&+!
?,%< p: oz%mm{zk:+
)@3=%AB ro

!((''
).!++ +
C)'!+>' 9o ?%
!%
D.!>"+63+
6:+73+7:+-13+-1:%<>, |}~BB9o3n:+r3o=+r:onz+
, o3A%
defegehiej 010
234567896
87983864 44 96 K4.89GK!GK4.89G[!89\389G9!",!M38.

35 88 >83839/3 .356935!S4 938.5GI86
F985386
F5555/865 6
Y6
J65
% &!""#'() J834 8/86S 63 8! //3553 8 6
J4 8
//83635G\]/; 4.GW !G!"T,!
!""#$ !"" K3GX !00!M85 4>V3543/3863 8956589H54/865!
*# !""+ K P-GF/5GF!
K48GW!GI .]GH!GM
839G^!GNOH
8>GH!89
,!""$
!"
! I 55GK!00'! 44 ; 4638.6
<F289<F
9
456 838/ 95538.; 6
/ 95>56/8954P38.
-64!2 9R36>89S448GG0T0,!
-
3
.35838644 /0!'120!03245426781 H 49 GF!W!89M
6_GY!U!00!M85 4>538L
! 5438. 85/468!8LY89; P 2 9
M38G9
8 .>89J8.38438.GIS455GK
9
87896
87983864 44 96 I6 8G2\!W !G
!!

35 GX!",!9
4 7/6
9 ` 48>535!F985
382 9I54
G,GT'0!
% & )
/8GZ!=!GU4<H4638GZ!89\-55GY!9!00!
!""#$ !"" !
""#'
( J] 663 858956;336> 448
3!X 48
M 85 4>M6935G''T' !
*# !""+
/ 8GZ!=!89\-55GY!9!00!M 45 44 489
6
8 <448 63 838934>4 9665638.!X 48
!""$!"
0!"

M85 4>M69350G''T'!

48G!<a!!GV;83PGM!89\ .
38G9!00!
/435 8 /6
95 48>_38.4369448
-
3
.35838644 /0!120!03245426781 66_565!GY! X 48 M85 4>M6935G0G''T0!
:9
-4; 89 6

3.
44 463 80!' 48G46"38". !9
5555 9693494<
89
9 83/6
95 46
63 8
45-36
6
4; 89 6
-44 4< 6;336> / 3_/4 9938H-3!F;5646G
63 80!0!=6
4 4 896
66
4358 6 2449M 85 /6435H638.GX8"TG""G^865G
8 .
398 48>9348386
448 6_8GY! G!!T!
899> 4GX!I!^!""!9
5899663 8
4 463 85! 6
349< /435 8/6
9386
585 4>63 8
Y/;4.4<6>6635;>336465/3<3646 85/45!
2 9R36>89S448G,GT!
J8835GV!H!00!9;5 4436>65638.!X 48 M85 4>
?@A@B@CD@E M6935GG0T"!
J8835GV!H!G89J8835GX!H!00"!Jb38
> 6
<
53565638.!2 9R36>89S448G9 3L0!0
F/438GH!F!89I 554GJ!K!"!=385L9
34M85 4> c! 9b!00"!0!00!
J63 8!24/8GM824835 ! J8.8G9!",'!H6
9896
4>386
569> 9 4
F8. GN!89NOH
8>GH!00!9
349448656 4485!8LF!94PGX!=!X
85 8GX4!89V!U!
896
8 448 63 8L=5N95P> 446Q2 9 H 6 8J95!GY/8I5 8556 J834 8/86N9 45!
R36>89S448GG'T''! F9/3G^-a 4PG!T'!
FM9H 864863 8!00!M68949 U39 4M85 4> 24435GU!J!"!9
P<3536/6
9 85/465638.!
3/ M;5686363 8!V53.863 8J"T0,!W !!0 K3 /6435G'G"T'"!
F88K P FM9H M689495!FM9H 864863 8G 2388GF!89\ 34GX!X!""!V64/3838.6
4 436
85

P8GSFG!T! 45 856 398 ;3 848L9
5 9
K5P4GV!"!43635 93485/ 8.48P 56>!X 48 S;3S 3>89H4P638.GGT!
5/5 4/63 /435 85!2 99
8 .>G2;44> 233 G2!",!N4.8 63-38<b36>63 8!!
"G,"T'! M689495 b36>895 438.5!4638.55!2 9
K5P4GV!";!43635 93485/ 8.48P 9
8 .>GG',T!
5/5 4/63 /435 85!2 99
8 .>GX>"G UGH!!89M38.
GX!"'!M663563H6
95382 9
T"! 89 85/4I54
!F9/3GN489 G2\!
K
GF!!GJ8.89GJ!GX
GY!X!GZ4356858GZ!89 UG H!G M38.
G X!G K3G X! 89 F68G M! 00"!
S 58G!M!""'!R 9LN63/953.8 94 9< M663563H6
9538 2 9 89 85/4I54
!
65!HFSS= 4P38.S4"GF4
5M
K53855G J534F9/3GF/5649/!
F4
5GV8/4P! U486_GU!"! 85/468 65638. 66

KP/8GZ!X!G
/;45GJ!W89U4.3GH!H!"'! 4 Z4 .4 9 8963 8!8LS4 938. 6
I54
95 434944889

9 8365638.!X 48 848 6
F/438H6856366G
3. G\G
2 9M38G'"GT! !,T,!
012 30 4567656896
6 8
!"#!$%&$%&!'%%# $44;6/4;A*JQR"%:4#9!
(!%#)*+,)** '4%"4 !"#!"!'#%&45##!9.
-./01!#023&%1%456(3!&4# %;!'4!9/!!<+J+,J+
6+77*/48!99#&%!!'#9!!# : M#.5<-=J-4%&'%48!%" !E
%%'#"%#'#!9 !#'"9# "#!"!'%%#%;!'4!96.%0#&>
;!'4!9<#!5<%'#+)+,=> ,+7
-?1;<0@(A4&'%5@ #!%%;!#2?'#%6?<+77Q#%%
6+77*@!"#!!9B8 !! "% W!#'O%!%&#&!4X9! !.%%$&%$
"9%&!#/!!C'4%59 /!!C'4%59*)>,)>
=,== 0!##4Q(:!06<4;2<%!-
;<0@44!1D+77A%%&!E J*QR"#%%#%4%:4#9!#%%#B
#4%&!#F1 !"#!!9%&4:499%8 "E"9"99'!E%!%4
%'G216H#4:#%E$!#%#4/!!C'4%5 %#%#;!'4!9/!!<>)>7,>
9+7+>,+* C'#:5@-'#%A@=>2#"4#'4E
;<0;I#1<11#456A( %!'# !B%84#9!'4%!4"!"!%!#
32+77*(#%E$!#%#4F1%!'%! &!%#=,
%4 !"#!$%&! %9!"94E <4:1;('44!.;-3;+77@!#'OE
%%!$%&9!!"!'%#/!!C'4%59 %!%&#&!49!* !4G'45"%!4H1'#%4
J,** $/!!C'4%59+777,7>
;44.=>K%!'%!%! %48$!9! <&99N-+M45##!989!" !"E
%&!#!9##!545##G!!'%#%L8!'84'E #!#;!'4!9%&1<%%#%41##!%!>J
%!H$%&#"4"&##!#"%8##!545## )*,>77
GL8!'"!B4%&!H;!'4!9'%%!A%%# <&%-;('&"?**1'4%E4.!!
+,+=7 "98#!#%&E5E!4 &4@&4
?-<2-<MN6&!56??M A84!"%)=,>)
+77*"9%#%#'#"4:!"#9E <&%6/#%$%& &4F%%4:4
9%#"!#!"%!#9!&"#!O'# !!.# 9!%!9!.#1<6<%U%!$#6&
;!'4!9<#!5<%'#+)>J,>)* >+,>
?4<1<E%6';EP> <&'%U-J<!'#!984%5%&##!584'E
<#!5%#%$%&5!' &4/!!&!4!5 %!!99!!#/!!&!4!5+),J
>*G)H+,>=, <4;2<%!-(4!!Y'#%;*Z##'#
?!#%Q@!'!2!/285@64! !9##!584'%!#&Y'4%5 !%!4;!'4
1<G+77)H0"%:4%5&!##!5#'%F !9A5<=>++=,+)7+
1 !"%'4R"4!%!/!!C'4%59> <4;<%!-3!!4#5165;6J+
=,J= @!4%!:%$&! %# !#'"%!!9
2'#/(44#4/*J&#'%!99!! :;!'4!9/!!<)J))
"9#&'#FA!%#%#"%%#%#"% !E <%!-<4;2J*@!"'%R%"!::4E
#'"%!S&5#!4!5(&8!)J),J>) %##%!%#%#;!'4!9/!!<>)
6&#!@2;@&!-<<'&A<;!<T? 7+*,7+
?MMN6&!56+77)@!#'#"!%"9E ""(;<&99<Q<&@6><#!5
$&%&5#&!'4!%F1 !##E'4%'4#%'5;!'4 ""%!!99% !!9!!#'#%$!#4%&E
!9<#!5<%'#*>*J,= !#/!!C'4%59+>,++
65#6+77JQ#5"!$9'445##!9"4% '!4--58!2D!2>4#%##!9
""9%#%#'#%& +%#%/!!C'4%5 !!.#O'%&!:%!#%:9%#%%#%#
9*)*,>* 9!44!$4:%' !#'"%!;!'4!9<#!5
6!#.!$%U-0*)!'%#%<#!5Q84'%! <%'#+7,+=
!9/!!#6.%0VA 1""!&#/!! DO"-@6<&99#%-;<#!5
'%%!3#%"!%@ 45##.%"%F@!"#!%%!
6'44</#K22!.#&2+77@!"#! ;!'4!9<#!5<%'#7+J,>J
!9:#%E$!#%&! #49!%&#'%!9 D.#[6'442J2. !#'"%!!9
!#'$"9#1'#%4;!'4!9" 9%99'449%&#/!!C'4%59*
30#&, ,
012345678
9
5342
55 4
!"!#!$%&'&&&&("!&%)##'
' $&& %&(&!'"$( '*( $&*(+ ",&*,
&!!#-.&!( !!#!&(/0"&!"#%&&#!$%&'
!!#-.&!(+&(,""$12#"$!&!,$&#$
%&/.#& '!!#!&!&&",/0"3%&2-%2("4567
&!&%&$$&&&!.&&$&!%&&$/
89:;<=>?@ABCDEFF9FGHFI<FCAJKLMN:O:PQK<R<SRCTRCUAVC:<FR9::WQIVNQMNNFCFJ:C<FU<NF
CFK;O<K:XQITFK<QPR<Q:IK:X<NFJOFRKRI<IFKKY;IJOFRKRI<IFKK:XUQO;<FK:O;<Q:IK:XK;MC:KFRIU
K:UQ;ZMNO:CQUFDSFMROOFUNQKZFRK;CFZFI<KNFU:IQMKD[OQWFU<NFV:CUAVNQMNQK9:<NNQK<:CQMROOL
RMM;CR<FRIUI:VVFOOQIK<ROOFUARIU;KFUQ<QI<NF\CK<:X\MQROCFJ:C<:I<NFIFVKMROFD
]^$_.)+`6a!&b4c1d17
0e454 c// !& //////////// ygy
fgh wkjqmnkj //////////////// ygy
fghf ijklmnkjopqrlsntnjuvrmwxknr// yz{ rrrjr //////////////////// ygg
fghz |rlkjnpqnjuo}mjn~nkj
krnqn/////////////// yz
fghy rkrjlrlkrlmr////////// yz 78 7¡46e¢£
4e¤¥
2¦5¢§¨©56 £
c//c #&//////////////// yz 01e
5
c// .&& ////////////// yz
c// 6#!///////////// yz
fghg xrrjrpqr ////////// yz 0"#%&!"#$,"!&%)&&2
c//c !&///////////// yz (!"!&) (&$*( '2
c// (%$&) //////// yy
c// -$(%$!& ////// yyf #$%!&/"&!"#,, *)"$&
c// _!!&$0&( '&!("$( '!!#-. '' $&/ª
,""$//////////// yyz ""&)"$&$ «%!"!-,
c// $3%&-!&////////// yyy
fgh{ mkmnuxpqr ////////// yyg &/".+!!#!&!(!-$
c//c ^&!#///////////// yyg &(#$%!+"%(" 2#$%!&&%&%2
c// )&///////////// yy{ ..')$!*&.-&'
c// & 622$
!(/////////////// yy ) &
*(!)#&&/&!")&%&"&2
. ##'#$%!"&$&!$(&
c// .&& '56^ //////// yy
c// _..&& `^#b/// yy &)#!"!&*/ &)#.+#$&&)
c// "_-)&$&&%&,"56 ') ,"""#$%!&*$ $&*$
!& /////////////// yg &)-&%&&/#'!&+)("
fgh rxnkqjlwkjrrqrl
kxr //////////////// yg $&*-"#$%!&$!" &"& ''&/
c//c $!6(!407 // ygf &%!"!&, %$ -%&.-( $$
c// ###&&!& /////// ygf #$%! . &"& '"#$%!+
c// !!# ////////// ygz -%"#'!&$&.%"&'!/
¯/0/,&&+¯/¯.)+°FIK:CL±TRO;R<Q:I:X²::U+ $!!0³&+ ¬®
^c´/c´´µ¶1µd2c2c12dd2·c+¸#(!!¹%&&&$+ć´
012 34 566789
677 9
! cdefÿfhijfkfcl
""#$"""" cdefÿmfjlÿknop
$"%&'(
( """"# cdefÿkqrfjsifcl
#"" "&)"$ cdefÿtcdupicl
$#*$
+$ $" vfdipfjÿcdefÿvqjÿrdtcdef
"$,"$-./& rdtcdefÿtcdupicl
0""#( "$$
$# $&1 # rdtcdefÿkqrfjsifcl
"%""$ " rdtcdefÿmfjlÿknop
23 3#.44.5&6"
""""7"8 rdtcdefÿfhijfkfcl
#% $ wPUIGHIG B 4!""
&9""#$ " """&3"$$(
"*$2: ;<"' %$ .4.:(4
.44=>9&.44?><"@+.44A5&B :(&
$$"*$+%$C
" $" " ":"$$ "
$"2-.45&1 "#"C#(#
$%( $" &B"a$1$
#( " $" %$"$,#!
"&D#$ ($$%(2E(E
$""# .4/?5&6 +% "(
$" ! 25 +!
: "" "&
"&E"%$"$"# ! 3"a$1$"
F"%"$"+" ","+(" "
"&B%""$ +"+ C(
" $"C&0" %"(++
""$" "$,"$&x&
"%$+ 2.4//5$$#"+4
& .."y%(-"
a$1$"$
4!%#"$,#(
GHIJKLMNOPQRQSTPOUVWXSOYPZQSYPNON[ 2E(.4z45&
\QQL]YS^PTPY_ {(""%"
|@"( |E(
%+"#+
B"""4!! " #("$ %
""+'&.=&.&B+ #$
+"$
"&B""%+ *$%#&) "" # $
,"% +
$("+,%.4=` (z}}++$ (%"
'3"D%a$' C"(+ #&6 +","
-1$-"12E( ,""(" $ " %
&&b~
;.4/b5&D%E("
""4!"$! B+"""+#
$"2E( :(*$"#(
;.4/b5&B""$"$ B$2-?5&B$"
0123 45678859
5
765
5 3
! !$%!!'!7'!(
!! !"!# ""'# 7'"
!$%&!'( !!#'#!((!$
! ! !(! "!(' %!""("*!!+)'##$
!!# "( %(!#!()"!
!!"#$%!! '!!"!!#(#
#)!#'#*!!+, (!#! (!
!'$-(! "(!! )##$C ("!!
!(!"(, !'#"
(#"#!'$ !!!#!'!!#(!#
%.! '( !# (!##"),$
((!$- )!,###!# B ! ,!!
)!,! )"! #! (#7!"#!)#!
(! ""#)'!#!!"##" !",#!(!#(
##$%/0(##) !!#! (!!, ###
'#!#)!# '" !#!*(!)+!#*
')!,##$1(!#1( (!)#$+%)#""'(
23.456! #!!!)""# !#DB< ##!#!
)!!!'!(!!#!2"!! 20:!#@#A>>36$%
!#7#!#!'"!#!6) #'(!)#!#")#
'#"("!, (!#"E!!#2@
! ""#$-!,#)(!# 56$0! #'(! "'
" #!(!#) )!#'#!(!&!)"
(#('!$%$ !!!#,!?#!2E $
23..86,#!(!##!## A>>56!("2@# $A>>F6
!##($-)!# !!2G$A>>56#!!!
)!#)" 2H $A>>I6!#'#
! !#!#! "!!# ( 2@#$A>>46$
9! "!:$;,')"
!!!#'!)!
,###!##)#!
!$
%.! )!:#!' JKLMNOPQRROSTOTUVQPOTWVO
!#$<,:23.=>6# .!
#!!)(#, XYZ[ZX\]^_`
! &
&#!2*!!#+6 %#"!#!#!
("7! !#!( '("(#"!
!# '# 9($;,' 9 " !&#
!)!,!## "!#'##!$%
!&!'! !!#!!! ,!!
#!(#( !# %)38$3$0(()'#! (
#!"( ! ?# !&#"(!#!(
(! !'20: !# !$B!'
!#@#A>>36$%!! (!)#!! )
(! '#! # " &! ( (
(!$B(!,*!# ##! &!!$%
'#!+',!! , #!) !!$
(# ! ",!!! %"" , !a'!(
012 34 566789
677 9
%&'($$)*$!+,)",-!$*-)$(*-$!"),
!"#$$$ ."$*+$/&#&-0,."1 0,!""$#234*
5&6!#!/$)*7-8*")-),$",-&2
9&:,,"4$*$!,"($**
;&<,"$*"),
=&,0,,!""$#."$*01
>&"",($*$! ,!,
?&@#,$!)4,A!#!$!*$(*$)*"B*$
C&D !"
E&@$*17,"*
%F&D ))"$,"**,!",
,8,!" $$)$!,,84)$0, $**1$1$!8$$#4$)#4",0,4)

$4,&@$)*($**0,$$$*# 4+,/b $*&-5FFCf:$V*#
43G#&%;&5& $!g$G-%EE=2&
HIJKJLMNOPQRSR hijklmmniompmq
T4!$$0,)4EA$*$,$#!48$*A
" 4,"#4-""$**1(#4U*V @")(,04,)4!4$8("!."$01
W,)*1X$#,1&T41$,4$$*17!"# "),$$$!44"*!,4$(
$,$),$-YA )$0,3 ,!A ",,##84$) " 0"),#!
"-,$$*1 08$,$0**3!(1)$, (1,18$*"$,*$3**$)$,V#
0)$0,) ,4$3 ,!"&Z84"#4 ,$,4,&T4EA$**04$() !+!
4$*)$1$48$,",8$**8*0)$A 8$,"3$1$) ),8,!"!A
",)-4$,$),$,$4""$**1["+! ,)$&G,W$)*-r$ $*&/5FF920"!4$
($! 4*$,#,$)*7$"),& $","",!8, 04EA$*/*$V#
48,($**$(*2,!"!$)34$3!,,$#
0$*"$#$) #@),$$!s$$/("
HIJKJK\]^PS_O`SaN e ,$2"),&d4$,*1!8*)$!A
# 08$,"4!$*-t,1$) /%ECE2!
@$! )) *18*8,*$! $4 $4,"*!(, ) 0,W$-$**1
$# 4$),!"(14$)"),& "(4$ 8! 04$*-$#-U@CA
b 38,-4,$,8,$*,$!,,*$A #,*$3$ !$*34) ,u*Vv4$u!*Vv$A
&T4+,4$)$1,8!)$!!$* $!$,!EA43
-("
()34$(,4$4
*134!$*#34,*$A
0,)$&c1,$!$*$/%E>%2"!,$! 8*13**A*V!0 !X/
(,/["!#3,(*! 4,*$2 (*3-33**(,w1%ECE!-
&5;2&d4
,*$*-)$#A
8,8,$*!$1 04,3,1)$ "!)$-*$(*!)$#"!
,$# ,!,$#*V#0,)(,$ $)($ 0*)$,V#$!,$$*A1-34 4$,
! 4,&!-$,*$)$1#,$*1 $!),*$8$*344$**3$!["
$*#-
) 0
,$4!,)$ 0,!" A ,8",$#&
"),4$8$(,!$ 04,$# 0,!"
(8$*"$!&d$"!1(1b$*&/5FFC2-4
,$,!"!,)$3$$**1,A HIJIJHxSN]y_OP]R
"!0,4EA4!$*&T4,!-4+,
["!#)(1$"),)$1(3**,!8 0
*$,(4$8,/e ,$*&-5FF92&G$**1-,*$ T4,$,$")(, 0"!34,4$*
3**$**3 ,!"400 0,8# ,!,- 3,$V! !$4,4!,
0121 345678996
496 96
ÿ
!"ÿ!#
$%&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&'&&&&&&&&&&&&&&&&&&&&
!ÿ$(%!&&&&ÿÿ&&&&ÿÿ&&&&ÿÿ&&&& )*+ÿ,*"ÿÿ&&&ÿÿ&&&ÿÿ&&&
-+ÿÿ!.*(ÿ(!.ÿ*(ÿ+""ÿ*(ÿ!ÿÿ*+(%!ÿ*/ÿ0-)1'ÿ0213
4ÿ5#ÿ+ÿ+6!ÿ*6ÿ5ÿ"!ÿ*(!ÿ!+ÿ*+(%*+ÿ*/ÿ5ÿ
!*"(3
-+ÿ5ÿÿ&&7&&ÿ%*+58ÿ*(ÿ5*6ÿ*/+ÿ5.ÿ*(ÿ+ÿ0-)1'ÿ0219
:5#ÿ*+;
&&&& ÿ5+ÿ*+ÿÿ%*+5
&&&& %*!ÿ5+ÿ*+ÿÿ%*+5ÿ(ÿÿ5+ÿ.!ÿ6#
&&&& *+ÿÿ6#ÿ*!ÿ%*!
4ÿ!+ÿ*(!ÿ%*(5ÿ65ÿ6!ÿ/*!ÿ!+<3
=*(!ÿ+ÿ!+ÿÿ+ÿ%ÿ"(!+<ÿ5ÿÿ/ÿ*(ÿ+"ÿ*3
4ÿÿ5ÿ%ÿ*!"+<ÿ*ÿ5ÿ+(%!ÿ*+ÿ5ÿ<3
'*ÿ$>ÿ<*ÿ#ÿ+"ÿ!?ÿ5ÿ%ÿ*+ÿ*(ÿ5.ÿ(!+"ÿ5ÿ<3
-/ÿ*(ÿ5.ÿ+ÿ@(*+8ÿÿ#ÿ5ÿ!.!ÿ+*63
ABCDEÿGHCÿIBJKLCÿMGÿNHONDKMCÿPGQJÿGRCJKSSÿGINHNGHÿGTÿMBCÿIJGOQDMU
)%ÿV&&&7WX&&&
&&& Y#ÿZ!%
&&& Y#ÿ.!ÿ%(5
&&& Y#ÿ%*"!
&&& Y#ÿ<5
&&& $5!ÿ#ÿ+*!ÿ"#
&&& '#ÿ<5
&&& '#ÿ%*"!
&&& '#ÿ.!ÿ%(5
&&& '#ÿZ!%
4Y1)1ÿ[>ÿ>ÿ\1ÿ$1]ÿ4[13
^_àbcad efghijklgjjmnompgqgrrkinglsjsntnkfnuvwkfghx gq{gpknwzf~zgjsk{nmlksqnwknkfnuzrwrmqphgnsmqsfpkrx
ijkfgpkkygjzgnk{mqnwk|ximsqnlgjgqrk{wk{mqsrfrgjku}grw mhhkq{k{ompfngq{sqigqkjffzrwgfkhijmtkkfmprmqfzhkpf
fzlfk~zkqnigksjjwgykgqkfrgjkompgipm{zrnsnwgqk rwmfkqopmhg{gnglgqwkpknwkipm{zrnxzfgk~zkfnsmqqgspk
nwpkkx{ssnrm{kuvwkmp{kpmokygjzgnsmqsfnwzfrmqnpmjjk{lt hswnwgyklkkqjjk{mznfmhknshksqnwkigfnuqgrmqfzhkp
wgnsfipsqnk{mqkgrwigksnwpgq{mhsgnsmqpmngnsmqmp kj{nkfnnwkzfgkopk~zkqrtmzj{rmqph wgnwg{lkkq
rmzqnkplgjgqrsqmomp{kpfuvwkpksfgjfmgrwkrmqipm{zrn {knkphsqk{sqgnkjkiwmqkfrpkkqsqmpmnwkppkrpzsnsqsqnkpysk
zfgkopk~zkqrtgnnwknmimonwkigkuvwsfrmzj{lkzfk{nm kuuhgjjsqnkprkinu
rmqphnwgnnwkigqkjsfnffkjkrnk{gpkfnsjjzfkpfmonwkipm{zrn
zqfnpzrnzpk{jsqkfrgjkffmhknshkfgqrwmpk{ltjsk fmhknshkfpkokppk{nmgfysfzgjgqgjmfrgjkfmp
gq{{sfjskmqkgrwkq{mzwkngju||gjkff euvwgnjsqkfrgjkfmzj{q{fmhkgiijsrgnsmqsq
|mwh gq{gglkp||usqkfrgjkfgpk wk{mqsrfsfqmnfzpipsfsqgfnwktlkrghknwkfngq{gp{
001 23 4556789
5668

!"#$% -%'(-/
&'( (''(-+)W
)*%' ++%-+(
+,( - )U( ("/
(- .((( -(
-"/ (-(
(0( %X)1!"YZ7(
1.(23%4$$"5.( '
)%4$$67)8'%- / -%%)8'%
'19'%4$$"7):0 ( '--'
-(19:; 7 (- ()U-(
'(- <14$$#7' (-((
=-+1>+7?'( ('()
((=+1+7&)?3 ;('(
'@)!A)B) -((
(-(19+[%!""A5
;23' %!"Y!7)U(((-/
'(%9'
CDEDEFGHIJKLMNOPQLKRHLKSJ 1!"##7(-'\(/
)X'(
:(T#%( %&%=&
('( (/ + '(%++'
- '+ '(
\ )?V--(
-'()U/ - (-] (
% '(-++' &-
)
(+' V
()U-(%/ %-( (((
("/
(-(+ -( +
+)V((/ ( +>+)
'(
^_àÿcdeâcÿfghÿeddaijek_^_jlÿmnaog`_dcp
qrÿtuvwÿxyqtwÿzu{|ÿ}u}x
ÿÿÿo_c^_~a à_jnahÿ^_~a ÿÿÿÿÿ^_~a
ajhaa^l `ghÿo_c^_~a ajhaa^l

rÿvqwÿtuvwÿxyqtw
o_c^_~a à_jnahÿ^_~a ÿÿÿÿÿ^_~a
9 ajhaa^l `ghÿo_c^_~a ajhaa^l
-)179
'
+%.( yrÿxtqÿxyqtw
)14$$67)17((
)1730 ÿhaejacj à_jahÿ^_~a ÿÿhaejacj
-( _e_èkâ `ghÿo_c^_~a _e_èkâ
139:;7 ÿo_c^_~_` ÿÿÿ^_~_`
<14$$#7)
0121 345678996
496 96 0
!"#$% zxÿ
&#!$'(")"* vwxÿz{|ÿ}~x ÿÿ ÿÿ
"*!"!+,-%.
/"#%".*"#! mfbndbldÿconmcanphbÿhcib
#""."##%"#"*"%
&"*&%!!"*&#/$.# hcibÿbqdfbobhr
01#'%2334567"#1#%23389'
hcibÿsbfrÿotue
:;<;<=>?@ABACD?EFGHICAJK?BAL hcibÿojkbfndbhr
/# !!& *"# 0+,-% hcibÿlhcmedhr
M*'8N'N9.#&#/$67"#1#
023389""&"$#O" abcdebfÿhcibÿajfÿkclhcib
*$!"*!#/$0 kclhcibÿlhcmedhr
1#"#67%233N9'.
")""!##+/#
-*"#60+-69#/"#! kclhcibÿojkbfndbhr
$$""$ "*' +-6 .
#&#/$P""#*"#./# kclhcibÿsbfrÿotue
"./$Q*!$/#R*$S
T0Q*%8O425P"'%8OOU9'+,-
/"#!&"!""* kclhcibÿbqdfbobhr
!01"*"#V%233W%/9"#"
"!$"*"##"X"*! mfbndbldÿconmcanphbÿkclhcib
#!!""*Y0M#"#Z"$%233N9'
#&"*+,- . 67
/#!!&*"#0+,-9%!
#/!."*[M%/.# %"#1#023389'M.."
/"*&""833""#233"
"*.#"#/$*"#" /0S833 8339%!1#"#670233N9'
0"*""9"*
#. $""0R\#,
.#.Q'T9'("/# $#'1"/.""&
")"."#! /"#"..3'OO!`8!#
""/"#."'6"#%#* ""#3'O4!#!#'6#!
"#"0**"/"*9* !""./"/'M
&#!""$!" !#"%./"/!
"0#/$Q*!*"""$ $#!!"!"0N_W0+,-9&
9"#$.#/."*" NÒ0O"9!8%2W`/"9',
$*""' #&"*!+,- ./&#"
Z+,-&#"$#&" "*.#!#%''%/&*"#
*&#"O"#"] "'*"#!"*.#
""!"#&"*. !"$.+,- %"".
"/^"#"*#!!""*# #*/&/!#!!""*
0+."#-"%8O4_9'("*" .#!#'
!#%!"!+,-"#O 6&#&)"#!
".067"#1#%23389' "!."#"'P"
.#"."##%""&& '0233_9)"#"X""
"*`8!#""#""&&"*`!#. ."#.!$!"#&#!'
001 23 4556789
5668

!" )<V$*W-.6%$!
##### $% 1!0#
&'( %# #!"#
#)#*+,-## .'$1,
./%0#$1 ,.X)*5-%
,. ,$YZX%0
234,)+55-#&'( , %
%!11, #.81%
##!%0#.6%$ [!.*\.].X!# 7#
71%
!$&%.)+55-#1
&'(%
#,.8%# ,
## ,% ^_àbÿ
%,0# deef
11#)0#
0-.
'#11
$%!!&'($% gahiijÿ
,, deef
$#%0# ^_àbÿdeefÿ
,1.8,1#
%! #! gahiijÿdeefÿ
1!,0)0-# 0# deef
7)# #- % deefÿ
,1! # "9# k_lmÿhÿinmmiaÿoeefÿ
0.:;!171)1!
,7#0-% phjqaÿ
deefÿ phjqaÿoeefÿebÿ
1# !% # ebÿÿ rhjqaÿqhfÿ
$7##)<.$+55=-.> phjqaÿ
1!, shf k_lmÿhÿinmmiaÿqhfÿ
%,, !# #$
#71! ) sh f
shfÿ
9!1!,0!#7:-, gahiijÿshfÿ
.
^_àbÿqhf
gahiijÿ
shf
?@A@A@BCDEFGCHIJDHIKLHMNOKLECMPQCER
SRCINGKM
6!,!#$ ^_àbÿ
shf
#T191:#)<
U-,1,1$
!1)(0%V$*=W- 1 tuvwxywz {%.8|}~}%
#)(!.$**-.' ##,<.)*W-.8
#,#/ 71
}%!,,#X
)*=-.8%# , )*5-.
0121 345678996
496 96
!"#"$ ""$3%"(K"9"*+,65*+,6;LM
$"%"$ &'$()$*+,,+-./". .*+,;=4+,;1-!#"9"%J"9(K"9"
(+,,0-$1"!$" .*+,65-.
23$%$3%$10456
#"$$"*&"!2"%$
3%$564&,#"$$""$ NOPOPQRSTUVWXYZ[\]XZ[V
6"2"%$3%$$"&'"$
$.7!%$3'/"."
"89.+5.&.7!2"9'! 7J"$!$E%32$"
%"%$'!!$!""$!$3 %%9% J"$9'2"
#"9$"(:+,6,*:.*+,;<*+,;=-. !%"$$"""$%.^'"$
9".(=<<1-"$9".(=<<&-%$ _$$#$!"9""
">"2"!#233 %#%$J$%3$*
%$'3'!%"#"$ "$9!*
###3""#"9>"$". 3#*"$""!
?$.(+,66-!%"$&"(9* "#$#J"(^'"$
9$*@*3$*"$ 3-%$ 3'3' _$*+,,=L_$"$^'*+,,&-*#%J"J
#"'$""9$540"$"$' "9.7$2"9!#$#"
$""9$+<4+=. $!!"#"9 "$""!
A""2!""9$" "9%*2'"% "9
#$!".##. !.7$$2"9%'
(+,,5-"%$$$"'%"9=' 2*..*"3%"!#"3%$9!
$%$ 3'!#$!"! J"93"$*"' 2"!$
"2"#""$"""23 %.A%9"3$2%"9
%$.7%!%"$ 2$""'!"3%(J"-*"$
6"!$""$$ #"3!%*"3%$
"9"9!#B% 9$CB% 3$C%$3 ! $"(`"%2"$a)2*=<<,-.
%$""'3'$"'%"95'$. A" $E%3"9#$B"J
)#$"$:%#(+,;;-32$ "9C#$(# "$@F_"'*+,,;-.>"
1'$$"%$ 3'"$! #$*9' "$'
"! $%"9$"22"9%. "3"!"!"%#.K$
:"(+,;,-"'D$!"E%$9#"! %$"$% #$"
#$3'5"$&'$$""$3';"$ 3"9'.A"
,'$$"%"9#%$#"""9. $%9*'$"9
:"$"#$!!"$"F!" "!$%$' $.7%
"$#!"3$""'!!2 #"!$%%$
".%#!# "'""$%$#"".8*
'%"9"$$$"#. "%# "2%$%
!" ""9$"" $"$"$*'""9#"$.>!
3$"! #$G"!# $% G$%$9 "
#$.7!""%$!"9H(+-" *2"!"$$%*
"""9"#*"%#"* %""3 #$$.b
%"$"$"9*"$#"#D"I%"*(=- $%$2"9% #"3"$
$"" "$ 23 "2"c"2 $".7"
*(1-#'"$3%"$! % c#"%"9""' "9!%"
'%"9 $"*(5-$!""9 $! 2*$G"$9""
%3! 2''%"9$"" "93% "!d/3"9 $ "(#"$
54&'.:9*J"$$J#%3 @F_"'*+,,;-./#","$
"! $!#32*%%""9" "2""'#$2"9(!"'-
001 21 3445678945
579
!" opqrÿtuvprÿwxyzrÿq{t|}q
#$$%&'()*+"#$$,-+./01
023*)"4*04 {~
25*000 ÿÿÿ
510402 rÿ
-+620401400 qÿÿ
172200800 oÿ~ÿ
9-+

qÿÿÿ
ÿÿ
ÿÿÿÿ
:;<=>?@ABCDE?ABFGHIAJKLMN@ |ÿq~
ÿ~ÿ oÿ~ÿ ÿ
qÿ ÿÿÿÿw q
OPQRQOSTUVWXYZX[\
]04100* ~ÿq~
/400/0/01^]_- x~ÿÿ~ÿ
00_4`"#$$9-+a000" x~ÿÿÿ~
0!b+c,+d1"50 x~ÿÿ
0+a0 ÿ~ÿ~ÿÿÿ~
ea fea4f20g1 ~ÿÿ
000ea0!feh0! ~ÿÿ~
+fa11e/0f ~ÿÿ~ÿ
110022e/0f GH<:;< ^0100+04*0+
4*04442 ¡¢£¡¤00+¥££¦2
0"00*00 0+b74012_4`
e/01+fa/0000 #$$9-+
4004(00g1+
b74"/00!eh
0*f2"e^0f f0042+
i)0*i!001*j a/0"!5"e2f02
+c9%9-500+ 00!b+c,+§+j44022
a/0002 004"04
24*50g10 45g+
14k+`50 024 a4015020240
040+b4"2 4k2(0*10+¨0*
524 0100! /40415
!0+b74"44* 24224+¨
0*2200! 040"012g g0
!00!+l! 0000514
!2"01221* 24+a^]_0502
!!7!!! 1 4107*2
40200*0424+)* 5 0+©
00*544 e100f 5400"00102
)6"c99$-+60248 4225002+]
5*228*105 02 2241g
200*1024 506" 2220402040!22
c99,"c99m-+n00"e100 250200*1+l41
0123 456789
578 76 3
!"##$%& #ÿ&%'(
ÿÿ
31 +4ÿ%2ÿ2+ #ÿ0 #%
1'ÿ:0'%;&<0%;2%ÿ
ÿ&%*;%;+&ÿ'#&
!.$%/0/(
)#*+ÿ,% '%-
8ÿ7%
1 +
ÿ2+ #
5"6ÿ7#& 50'%ÿ62% 3%& 4
8ÿ9& .$%/0/
ÿ3<''ÿ$%4
)#*+ÿ,% '%-
=>?@AB@C DEFGHIJKFLMHGNOPQPRSPJOPJHFKTGUFTNKHFVWNOP TPFN[RNKHFYÛYURIIWHFRIHLYTRIP_\]OPYIHQPHGNOKYIKFP`
XUYNZ[KLONYTRIP\]OKYTRFQ[HJUTPRIKFPR[[PIRNKHFYOKQHGXUYNZ [PIRNKaPNHP[[H[`KYKFJKTRNKaPHGNOPNHIP[RFTPHGNOPTHFYUbP[
[KLONYTH[PYRLRKFYNYPFYH[WKFNPFYKNWH[RLRKFYNKFL[PJKPFNTHFZ GH[JPaKRNKHFYG[HbNOPKJPRIIPaPI\
cKNOOPJHFKTXUJLbPFNY`NOPQHNPFNKRIKbQRTNHGVPKFL IKbKNYNOPXUYNZ[KLONYTRIPNHRGPcYKbQIPRNN[KVUNPY
HGGG[Hb NOPXUYNZ[KLONQHKFNTRFVPPYNKbRNPJUYKFL NORNR[PcKJPIWUFJP[YNHHJYUTORYYcPPNFPYYRFJYRINKZ
DQPFRINWRFRIWYKYM^JKYTUYYPJVPIHc_\dFHNOP[RJaRFZ FPYY\NOP[bH[PNPTOFKTRIJPYT[KQNKaPRNN[KVUNPYNORN
NRLPKYNORNNOPYPYTRIPYLKaPJK[PTNKHFRIKFGH[bRNKHFGH[ [PfUK[PN[RKFKFLcHUIJVPUFYUKNRVIPKFRTHFYUbP[
TORFLPRFJTRFJHYHVWNPYNKFLHFIWRYKFLIPQ[HJUTN NPYN\GTHU[YP`hdiYTRIPYYOHUIJFHNVPUYPJcKNO
^FHTHbQIKTRNPJJPYKLFPJPeQP[KbPFNYR[P[PfUK[PJ_\ N[RKFPJQRFPIYRYNOPXUJLbPFNKYRVHUNQ[HJUTNIKSPY
gHbPJK[PTNTHbQR[KYHFYORaPVPPFbRJPTHFTP[FZ RFJJKYIKSPY\
KFLNOPQP[GH[bRFTPHGhdiYTRIPYaP[YUYHNOP[bH[P ]OPPFJQHKFNYbUYNVPN[UPHQQHYKNPY\D]HHNOKFM
N[RJKNKHFRIbPNOHJYRFJFHTHFYKYNPFNRJaRFNRLPKY aP[YUYD]HHNOKTSMKYRIPLKNKbRNPPeRbQIP\jUND]HH
RQQR[PFNRNNOKYNKbP^jHcP[RFJjHWJ`kllkmnQIP[ YHU[MaP[YUYD]HHYcPPNMR[PFHNHQQHYKNPYPaPF
PNRI\òppqmrHQQP[RFJs[HIÌkllt_\dLHHJ[PaKPc NOHULONOPWbRWYOHcRFPLRNKaPTH[[PIRNKHF^RYHFP
HGNOPKYYUPYKFaHIaPJTRFVPGHUFJKFuRF][KXQPNRI\ LHPYUQNOPHNOP[LHPYJHcF_KFRQ[HJUTN\]OPYP
^kllq_\ YOHUIJVPYPQR[RNPYTRIPY\daHKJTHbQIPeRNN[KVUNPY
IKSPDT[PRbWMcOKTOR[PbRJPUQHGYPaP[RITHbZ
vwxyxz{|}|~~| VKFPJfURIKNKPY^YbHHNOFPYY`YIKQQP[WFPYY`bHUNOTHRNZ
KFLàKYTHYKNW`RFJJRK[WR[HbRTRFRIITHFN[KVUNPNH
T[PRbKFPYY_\daHKJKFOP[PFNFPLRNKaPYIKSPVKNNP[FPYY
]OP[PR[PYPaP[RITHFTP[FYRFJQKNGRIIYNHVPRcR[PHG UFIPYYNOPWR[PRTNURIIWJPYK[PJKFRQ[HJUTNYUTORY
cOPFUYKFLhdiYTRIPY\]OPUYPHGhdiYTRIPYRYYUbPY VPP[\jKNNP[FPYYKFbKISORYFHHQNKbUb\daHKJaRLUP
NORNRIINOPTHFYUbP[YUFJP[YNRFJcORNNOPRNN[KVUNP QHYKNKaPNP[bYIKSPDFRNU[RI\MdYcKNORIIYTRIPYKFTHFZ
IKYNPJHFNOPYTH[PYOPPNKY[PGP[[KFLNH\FHNOP[cH[JY` YUbP[NPYNY`RaHKJ[PJUFJRFTW\NJHPYFHNbRSPYPFYP
NOPTHFYUbP[YbUYNORaPRTHbbHFKJPRH[THFYPFZ NHORaPYPQR[RNPhdiYTRIPYGH[NOKFRFJNOKTScOPF
YUYUFJP[YNRFJKFLHGNOPRNN[KVUNPKFfUPYNKHF\]OKY NOPWR[PN[UPHQQHYKNPY\
001 23 4556789
5668

([QQ ([KLMO\P
!"#$%#& !(1(/'1,
'(& (.''(''!
&('()'! (%I'''.(.!%
''*+!%#,"#$ -,&(
!((!'(+ ('('
!%-&,((.! %I((.!(.()
'/&(&! /'((. (,
&%0(1! *+!((%
!'(( -&,('''&,R('(+
/,/'%'( '1S((
((% '&!,
(.'(('+
1%I*''
(+
234546789:8;:<=><=?@A8;:B@C;<CDE@8A & ,&((
!((&'
FG8A:=H !('(&
(%
I.&'.( # ('''('Q
(''& .(*.!(
%#'(&"( &'KT!(T,
JKLMNOP(JKLMNNP(&'+ LMNYP%I]&(
Q%0(R& R((*%S^.,'(
!S(,R*+!S,( '((,!
RS!%T&(U.( ((!,.(/(_('
KLMNVP!./'.*+! '
'Q''(%0 K`(a,LMN\P%^(Q(a
,'( KLMMMP('(((*](
'((!('.(%#( (((&&]&!!
'K(PR.& (((,!(
!S(''(R.& /(((!.
%SI'R!W &%
'S(('(R.&
!&!SKP,R!&!SK
P,(R!SKP%U 234543b=8Ac>:><deb?f8;8
('(!'&''+
(!.!R'S'
&('.''+ g('((1((
Q'(%X "#$(h"#$(
'!.(*+! 1'''*+!
('( !&,(&,
(T&(U.()/', .(((%g(
('(!( '((('
!'( ((.% .h.&,('
#.'& *+!(',&+
*(!!'(,( '( ,(]/
'( R(S %[1'&/+
'(K.I*'%,YZZOP%#&/' ],&&'!(
0123 456789
578 76
!"#$%&'(
)*+,"-.//01/23045"675
5#8!9-"#5*6%8#$7"#:"
!6);:*<=>#5:-$*--)6$*"-
&'(:*-%6? $)5?88-*#:@
5;%%-5*#$%$6:"$%-$--)5*6
A-8 )-+B6);:*CDE#5$6*
5?88-*#:F5G-7-)H$)"5
9%-5-%8-$*6IJ568-7"*
$6*57--*-$6;%"K5;%%-5*#$%
$#$:-5-#$*"-57--*-$-
9-L-9:6;9)A- $
#8!6L-8-$*+,"-4MN01
/23045"675*"-58-)*
%!"-) 5 :6$5*$*5;8A
%!"+
5 5#8!9-A%!"F"#5*6%8H6 I:*#6$*-)A 7"6 - A6L-&'(*6*"65-A-967&'(@7"#9-#%$6T

55"67$#$O#%+P<+C+Q$-:6;9)-5#9?A-8#59-)A? #$%*"-&'(L6*-5FU*6$- $)U#)-9@][[<H+,"#5:$A-
-R8#$#$%6$9?*"-8-$5+O6-R8!9-@*"--8#%"* 5#8!9-A#$68#9!6AA#9#*?*-5*%#$5*$-R!-:*-)
A-*765-%8-$*56I:6$5;8-5@6$-%6;!*"*!-I-5 L9;-6I[+=F-V;9!6!6*#6$5H+a6*"*"-_T*-5*I6*"-
5*6$%-9-L-96I*"- **#A;*- $) $6*"-%6;!*"* 8-$ $)*"-A#$68#9*-5*:$!6L#)--L#)-$:-*"*
!-I-59-55+SR8#$#$%6$9?*"-8-$I685;:"6!#$T *"-)* -5G-7-)"#%"-6967-*"$*"-8#)!6#$*@
#6$5:6;9)9-)*6 I95-#8!-55#6$*"**"-!6);:* I6 5#$%9-!6);:*+
7 5 *6$-*"-6!*#8;89-L-9+U6!96*?6;)*+ ,"-$-R*:6$5#)-*#6$#5"67*6:68!-8;9*#T
,"-5-:6$)V;-5*#6$#5@JW6X"L--$6;%"Y;5*T#%"* !9-!6);:*5*65--#I $? -)#II--$*6:965-*6
L6*-5*69-L-*"-!6);:* 5T#5ZK':6886$A-$:"T *"-#)-9+XI-:":6$5;8-"5-L9;*-) 99*"-!6)T
8 G#5568-*"#$%9#G-C[\ &'(L6*-5#$*"-:-$*- ;:*5F 7#*"#$T5;AY-:*5)-5#%$6:68!9-*-A96:GH@*"-
:*-%6?F(6*"8 $ $)BG-@][[^H+Q$-V;$*#**#L- :"#T5V;-5**#5*#:A5-)6$ :655T*A;9*#6$#5$6*
*-5*I6A-#$%5G-7-) 7 ?I68*"-Y;5*T#%"*!6#$*#5 !!6!#*- 5#*#5A5-)6$*"- 55;8!*#6$*"**"-
5#8!9-6$-T58!9-_T*-5*6I*"-!6);:*5̀8-$L-5;5 )* -I68#$)-!-$)-$*58!9-5+U#$:-*"--L9;T
*"-:-$*-!6#$*5̀L9;-6$*"-5:9-+'5#8!9-$6$T *#6$5--9*-)@5-L-99*-$*#L-5-L#9A9-+O#*b
!8-*#:*-5*#5*6:68!-*"-I-V;-$:#-56I*"65- F][[^H!6L#)-5 %66))#5:;55#6$6I*"-5-8-*"6)5
001 23 4556789
5668

!" 1!!6----27896
#!$%$&'()#%&*+! ---!-:;%"
, ",--!-.--/0- -<2=->(>?."
--!1!-2.!" @AA7B-"79C7+D1!,!1$
!,!$,,.10!-- "-%!E!,-
343,"1!-2-!$ "-2!--)*
2-(,/5$-2"5$-2",5$-2+ !2.--111!6
FGHIJKIL 6:;%1-11!, !--!/1@1"-!-P996-/ ,

!M=?!6<!$ !1-2-<!-11!,,-=1
-!2-",/5$-2(-" <-21-2-<!-11!"!
2+",5$-2(-"+"1-2 !,!-@4 @,11.-,.
-5$-25$,$-2!2.N-=2 %!E-1-!!-/$
D10!-1-2-!!2.!--2 -2:;%:=->1
--21,!6O@ -2-
0123 456789
578 76
!""#$%&'('!#"$$) bcdefghij k.&"#6#6+$+% %&"$
a$)1Y]1YV
*&&+,-)$+&!$%./0$+&. k&" l
&)&$+12.+#+&$0&0+#$3#$ m88nopq8rstrr8urvv
+#&4&$+4 &$)&$+5 & O&&+, w1' xY1] Yy
+6&+#%+"&+5,$"7 P0+7!&07$) z1w yV
$+$+$+#$387++%#6+$+&.4$ &+,&0) y1' x'1] [y
&&"+12.&#6.,+"&+
)&$+%+&0#%!%+)&$#& {|o}~urvv
O&&$3 w1y xV1z '
&%$#1 P0+7!&07$) z1' [z
O&&$ w1' xY1V [y
9:;<;<=>?@ABCD?@ACEFEGHIJ>@?KHGLM nqo8n
O&&+&) y1' x'1w Yz
P0+7!&07$) z1 w[
N &+&0&.4#0!#$.&$& !.& O&&,3 w1y xY1[ 'V
&4##"$)+&##%$ +/0+7
$& $1OPNQ% !&!$%,$$+
$.&$& &++++"&$#$"&.!$) 4&+)5,++##!0+&)#6%$++7
&7PNQR&...&S0+7$)T& &4##"!$#7 +_%)&0"$ 0""#.&1a0+
$6&."&%01O""&$++$"#%"&%+ $$ )!,%&+,%$..+6#
+.&##&,+R.&U$%5'VVWTX &."&%0,$$#%4&,&0#%""#&$+
RYTU"%$&)&0"+,!&45
!#&,5%S0+7$))&61
R'TZ#0# %&$+&+.& ´£µ¤³°¤§ÿ¬²³ ÿµ¶·¤¸ªº
¤±³ÿ¹
"!$#$6+#+.&&.)&0"+1 «¤¬ÿ®¯¯¬ÿ̄ª¤°±¦
R[TU0! &.!&47PNQ)&0".& º
PNQ)&0"%#$3,$++0! &.
!#&,7PNQ)&0".&PNQ)&0"1R&X
$$+$"&&0+PNQ)&0"\+ % &
&4##% .&$+"0"&+1T
ÿ
£¤¤ÿ®¯¯¬
R]T^#&+0#$)%$..+&+5* º º
%&"5-&"#6$+"#&&. %&" £¤¤ÿ®¯·©
´µ³°§¬ÿµ¶·¸¤³¹ º
4+0+")&.&#&+0"#
$)&61 º
£¤¤ÿ£¦§¨©
£¤¤ÿ£¦§ª

2$+"#&5"&$+&,+#) %&"%
#)")$+0+.&& %+0))++ ÿÿ
¡¢
"&%0!&%$_%$ ""&"$%$7
$&1N `"#$++&, $ O!#Y]1'%"#&% »¼½ighig¾ ^#6#6+$+% %&".&&6"&$#
$ a$)1Y]1YV12 $+"#&,+$+#) +"0&
%01a&&.+#++&, $ O!#Y]1'5&7
)&0",&.#"&%0,+&&+,%% S0+7$)+ )&$¿%+$)+#!&45!#&,&$
#) %&"&%0$& $&4##" $+ #0#)&%.

61O %&$+&RW7"&$+#4#0T
&)&0"1a&)&0"+!&4&!#&,
+&1O"&%0%4#&"$),& PNQ)&0"5$+ &++0!%.& &.
$++,++#4#$ ,4+$&1O O"&.&+0+*
PNQ) & 0
"1 O $
+" & %
0 %&"-+&.&)&0"1
,,&#))&0"+5&&.,$.#"&%0 "#&%)$+$ %&"+#
" .##$&)&6
,+&&$3%&&&$5!0$"#$+, +,&0)-+!&$)") 1& )&6*&
%$) %&"5
+##+& &$& %!31Q)%$).0$ +0))+$)"&$#$"&4,$$+%+,++1
012 31 4556789
5668

#'^ !"[!4.&#
!""# _""1""
$!" "% #`1"!!!
&" 34.&!3(
'()*++,-#."!"(*/* %5#63
!0!1 12134.&(% "1"(WU%X1
2(!"!%5 "(!)
)6"*++,-#. 0!3 -3(!
( 4.&3"% ")-#]"
"!#."! "!(
!"!!" 1%4.& "33
!"13%4.&!"3 !""#6
!!!!3 !)## !!!1
7%-# 4.&"!"5#
6!!1!1%
"! 4.&!!"%!
89:;:<=>?@ABACDE@FGHIJKGGL@GMN>? #6

!1!!
!3!")##
OPQRSHE@G "!-U%#6!
!!"!3
T "!4.&! U%3 !"" !
1U!"1 !"!#6 13#4$!()a,bb-!"%
!&"'()*++,-#V " 3!
"4.&!3(313 1"1! c!V)a,b*-
""#T1! W" ')a,b,-#V "%
31XW 1X4.& " !4.&
!#.1% 1 (#
1"!""" 6! Z,#T"
1#.Y"!"% 311"5
!!1#.!!! ""!%
!"3 13 "
!!1! !!11"3"#
4.&!("!0!1
#Z"3 !
3!1! ! defghijklmnokpkqrsnqtiutvwipktir
!1#61 xyyonkzji{
!!13%
"1!!"1"
[#6!1! T0!"13
133( 3#`((
(#\Y"1 "!! ! !Y
1 !!1 "" "1 "3
#]"1(" !"!#\1
3% #T!3"! "1(""%
1321!"1( 3!)|(#a,,}-
"3 !#]"1 3"!3!%
"!3"!!" "~Y!)Zca,b*-#.
!3"#]3""!!1 ~!3" !1
01"1!!" "[!#]"(
0123 456789
77
5575
765 10
!"#$%##&!"%'()#"*+, nopqrstuv D24)#2%#"4#%"#%24"*&F?A'0
-../0"%1"$# "#% 23#4%22#%# g3:*"##2!$))#:%#$g"
21*"2%2$))#2&5462,-.770+8"#% g3:*"##2!$#%
2#"6*92#1"$21 *$"))"*%%# gg3 *
6
:* e:%#*$"##2
#2"%3:*"##%3"%#%
"92#"4#,9:#4%2:1%2#"6*92##%"$2% g3:* #2"!"*"9*9:#3:*%#:# 1$3"$
"33**4#"%*$4"%")2%;2 *%2"9:# g%#*"6# #9:#3:*"##%"%44"2%
##1#3"#% #"#)+<%42" g3:*"*$
"*22)4=4# 2#"#"419%+>"*#: g3:*"##2!%*$
"##2
31"$*66#4:)%%"*,6#4:)%1"2 g3:*"##2%*$#g3 3%#442
4#
)#"#21"$1"$%#21"))"*%#"!% ?#>4:#E&-.GI0
%!:"*+? )2%;22#4"*)%42"422
1"%$ 21"$ "*#)4##%4"44)@
#"%4242#12%"%"4#:"*#"2#%&A"**"% OPQRQwVxxyTxyY\X`Z`zz^W\_`z
B"**,-.7/90+A%#C#"%C)4#"#%24"%" 4#
21)*%4(:1%#2&D*E""%B"4F,-..GH
B2*1"%,-../H>42#%,-..I0,232:* g%*"#36,>:#E4"#4%#C#:"*#1
C)4##1#"!1)#"%# 4#2%"4#:"* %2#) "#+?)))"#%22"#%24"%9:2
44"%4%2:1)#%2#:"#%2+'"#%2#:"@ #"2222# 4#2 4%#C#"224"#3#@
#%4"%"!"9 4#&83"2#"*+,/JJKHL% %42)%22# &>4:#E,-.770+FC"1)*,
#"*+,/JJM0+N"9#,C)%4,4%#C#,"%"###:2 %1"$"**$*6)EE"9:#3%"26#"#
"1)#"%#4%#9:#2##"4#:"*4%2:1)#% %;2*6%"#7?+B+1"$%#=%#)EE""))"*@
"%"2)4=42#:"#%+'"2221 #2 %2%4#22"%%")))"##1#4%2:1
*1#"#%2 21)*"44)#"%4#2#%211 )EE","#*"2#12#))*+>)*"2"%#%22%"
9"!"**$%#"))"42"!9%:#*E+ ):*$2%2$9"22"%")))"#%22%"4%#C#
1"$%#"*3"$2941)*#*$)""***+g%")))"#
4%#C#2 "!%4:*#:1"$!%{:%42
OPQRQOSTTUVWXYTZ[\XYZ]^W\_àSVbcd } 2%2$*6%&|"#%1"6"%':*",-..M0+
*#24#"%*$)229*#)**4%2:12"9:#
#")))"#%22 2%%#4%#C#2,#
?%C"1)* "9"!"**$%#"))"4# 2%2$24%#2#2:*92%2#!###2#%9:@
24"*% "44)#"9*#$3"2!29$>4:#E %#"#4"%"2%"26%#1"%$e:2#%2+g
&-.GI0+N!*)"24"*9"2%"###:2"% -I 2"!"*:"# ")))"#%22%/J4%@
"4#%2,419%%212#"#1%#2"9:#e:%4$ #C#2"44%2:1)"%*2#3:*"!#"#KJJ
4%2:1)#%&fg3:*"##2 !$))@ 24"*2+
#:%#$g"h0"%211#!"#%"**$*"#2#"#@ ?)))"#%22(:1%#2"#"#%"**$%
1%#2&fg3:*"##2 %*$g3 4 9#" "%":2"2#"#1%#*6f"#%
#h0#):4"1"4#%@%#%C 9"6"2#+h'24"*411%*$:22"M@)%#
"44)#"%4+'23"24"**#F?4#%i"#% 24"*"%% 1f%!h#f"*3"$2+h~##"##2
>4"*F?A'+'41)*#*2# 24)#22 2%#"24"*#"#"%2 1 f%#")))"#h#
23%%'"9*-j+K+?*#:>4:#E"2%#"# f!$")))"#+hi"#,#"##1)#2##")%#"
#9"!21#%#"*3"$21"#4:)3#"44)@ %#% # e:%4$3#34"%#1 3:*
#"%4"224"*%"#"#%"*.@)%#%424"*, "))*$ #"#:2"2#:"#%,%"" "2%+
"2#:$ #42)%%4 ##31"2:2 'e:2#%%"#%#"62# 1 "1"#C
23" )2#!4*"#%&kl ,3# 2"232"%:2"2#"#1%#2
mJ+.M%"e:2#%%"2#:$ *620+D"#" "24*:1%2+?!3 ")))"#%22e:2@
1 #F?A'24"*"!*31"%!"*:29:# #%2"%2"1)*"))*4"#%24"%9 :%%>4:#E
*2226341)"##.@)%#%424"*+g% &-..j0+g%212#:241)"%")))"#%22#
2)# #24*"#%,#24"*2"%#%422"*$ #.@)%#%424"*,A"**"%>4:#E&-..G0
%#4"%"9*+ 23#"#23#e:"*"44)#"9*#$4:*
012 31 4556789
5668

! +)&"TU$
"#$$% P&'567R<9/$
&'$ .$
$"($') &%
"*+% %&&
+',-. &"0&
/"0& +&&
+ &'$ %$&%"0!
/"-& '$ $+3
%'$ &&!
&&1$ /"0&
/" $$$
21$&! .5""%$$9
.& +&>."0
$ $
3"4'"567789 $ &$&5"$"%V'+.
W$%677=9"Q"567RR9
"0 6: $$3
+& &+.%$
& &"0$ %$'&+ "X
3+ $&+3
1$ $&$&!
";% "
$ 0' !
+ "0! %&+)
$+' &";Q"567RR9%
+$ ' +'?@?@
54'%67<=9"-$ +&"Y
$)% %!
$&' '%.
&)! "Y$%
"#!>$&$ '$!
%&! &7!";
!.$"#! &''!
'+$ %&&
?@?@ Z3$"#
/-." --5[\\89%
+ +"]+
'$$&$
ABCDCEFGGHIJKLMHJMNOH %&3'
";&!
+&
#&$ $"*+%
&' +$
+)! $+$
$"P&'$5Q"%67RRS +&
TU$P&'%67R<93! "(3%!
&." #+$Q%&
0.+) Q+#5-
0123 456789
56
1
!"#$$ &!"# &$&#, && &#
%&& ' $(&#&# &$&! ( $# &% $$&# &#$&#B
"#)#&% &*)&# , &$!"#$$$ &&#)
& & %%& )&# &$+$&$&)' $,$$ &, &$(#
$&$&&# ,& $&$&'! &#$&#, $(%#)#&#
-&$&$. $/#)#& &# &# & $ % &!
$$&#$ &!0 & & %&'$$ "#&&'&#$&# ' B
&#&## (,%&#&#& $$&&$&,, &#% && )&#
&#)$+&#&'$)&! & &$!
123432567897:;6<9=
JKLMNOPQRSTUOPT
-&# #& & ( $,
$&) % &$!> ($$ "# $$$$&%B %& & %B
?@AA % &$& '* &'$&#$&$$$', &!
&$!"#& $( $% $&#B V$ & $$$& C%&&DB
%# &% $&# & $(& &&,$ +&$ +!-
, &$ % &$!"#$ B , &'$&#$ , %&&#$$'
$( $% $&#%$, &&# && ) $&$&& $$$$&# &$
&$(&#&# + ) $ &# &,$ ) $&#$!V#
&#*&# &&## &% ( $ #)#'$ % &$$ & &$%# , $&$&#
& &# &&& $ +CB $$$&# &( +(# (&#
)D$( )!"#, ) ( $ %&$# $$
&, $&# &&&#, !-& % &&&#&$
&& % &$ &#&$ $$# $&#$&,* &
&#$ $ )(&#&& & &#'$ )&$$) + ))
&#$& &$,) &. & , %$!W $&'#(&# B
&,&!E&# &$,$$& )&#$&%&
&,$( $&# &&#AB&$ ( $& # ,&$$#!
$& ( +&# &,&$B H$&&#%,$& &
$$ ' $& &&& #, & %&'&#$)$$'&#)$
% &$!F &(&$)#&% & $&$$$&& & # )$
)#& *,&##$ %&&#$ &#*$!H & +&)$ #$
$ &$)&@G&#$&# B '&$& &#& &$&)B
&$&&#&(&$$&#$ $%%' &$ %B %% $$!!(&#
&$!H&( $#'&#$.&# &&#% &, & (+!"#&#%$$'
&#&$ % &$($#( $ #$&$&')# &$&# &
$ % &&$ &,&'&# &%$ &#&$, &(&#&#& B
) )! $ '&#& &)'! & &$
"#, $# &% $GG, &$ $$' & &$& ), &(&#
($& & &#&#'(%& &#B)&&#'+('&# &&$
&(B&% &$ *,B& $$ &$&$ %))$ &#' ( &# &
& & & * + )% &$!"# &#))$ .$&&&( ,#$B
I$&&$&( $(#&#&#, $&# & '%&# &$&# &)#&
$ % &$%&%'&#$ ,& ' &&#&&!"#$
$$ % &&#, $&# I$&$(#&#&#' )&$ %
&$!"# & $#( $&& &,&'(&# ))$$&#% $$&#$$' &&%&$
& %&!"# & $#( C &D ?& $&&& &!!
011 ]1 233456783494
68
A'K&K&'A'C&'L'C&5&B'K&012(&
X" ""F""#
!" # A'K&
& '*'0(WI0H*&
$%&'())*' "K& 'Y'M&Z&.'.&012)&/[
#""!Q""F"
"+"," # "&DC!'W0'23WI230&
!-&. !',&/ A'%&0

12(& , "
!&ZO.&#%&
,, ,, %
"""# 66 &6\$&-'6,,G
"6&NXAC!
$%&'0123-! 5"'A'&(WI*H&
"! , A/ 'C& &011H&U!+,!,,O
!,4&5
"[!
,&
'&/, UV
"#", A'C& &01, 6"'W'H0IWH&
1W&/!"+,&UV,
"&%!," A6 'G
"'3'()0I()J&
& &A,'R&())(&X"""
"' ""
',, E " ,#
&G"M,M'02'
"+" # (*WI(H2&
& !'#,, A,!' &G &M'K&/&0130& [!
# (H'(21I*
))
"&5K,D'
&
&6,$# D' &N& L'S&012(& ," '
,'-''# !"""!#
''!'"' D' &N &G"UM'HJ'0WH1I0WW(&
'' ""& L'S&012(&R
,"$%&'())*-&. !' UM'HJ "&G"
'0WW*I0WWJ'0W30&
,7"8 # D' &N &M\'.&%&0113&U#
,&/ P E "#
"9",#9! J'(*1I(H1&,!&UV,6"
""", D' &N&M\'.&%&())H&R&
, T #"K L
& $K""!-&G"M,
M'01'*H0I*H3&
D' &'K '.&/&M\'.&%&())2&X""
"Q
&UV,6"'(0'*(*I**H&
:;<;=;>?;@ D' &N&'M\'.&%&'K'K&K&L'X&
())W&C!"
"!,& 'HH'0I0*&
A'B&C&0121&D!,E D' &N&'5'D&M\'.&%&())*&
F"&G 6"","
"A6,,'H'02*I01J& ",OC!#
A'K&L&'M,'C&G&C"",'N&A&())3&. " ,,&/QR
KLMON""P G'J*'((0I(*J&
,''Q6RS6&6 D' &5&'R' &N& &6'K&6&0113&
())3 L' "D ""#
M& ",&G"M,M'00'
A'R&K&5&'T'G&6&L' &G&0122& 0H0I03*&
C!","# D'M&#G&'N' &())J&K#,
""&G"M, """ &UV,
M'*'0*JI0H1& 6"'02'13*I1JH&
A'C&G&R,'G&D&5&())0& "M D'M&#G&'N' &())J&Z["
,!&UV,6"' ,#4,# ,"
0('*W*I*WJ& &UV,6"'02'(W3I(3H&
A'K&K&01J1&C"F" D\'.&0113&/"""#
"&G"TX'00'JJI2)& '"!"
0121314516 789

D $=I
3
H$.
3
"
:9
!"# $%&'(% %&&) *
(00)<2
#

/

+
!,%'!,! >
2

-"
./ $+(001

2 $
""(%%7)'%,7
/
#3#45

2
#26 D/ H

+ 9$9%&&75
6
/&6 /
$
""(! #
B49$9
+
.9
(17'(&( :/EF
+5I
3
.
8

+9$9%&&):/#
6
= %!@'%))
;
;

D"<3 D5
5
J

8
D+

/
#42$
" (00@=

#

6
"%%%0!'%(1

+?

<
$
9=<

=/=
5%1!7('!,(
(00!:/>
#
/

9
HD/+D
$/ D*%&@@+

+?

5%7

/4-

2/
)7@'),(

//$
+"7(
<.*
A
(00&< (7)'(,%

4:#

2 9 H $
# "8 -
I:
.
/
#(

$
+" @7 -(001-
N2

"&@'"%0) /+?

5 %&
<# :%&@7/
// ),%'))%
B4:3 $C$/ $
.D 9#9"/ 8
*
:%&&@- 6
EF9
8 <2
A
#

#+?

*G3 %(%'%7%
51&@'%0&
<# :%&@1:/#

9
"M
=
9:%&&&/
J6
B4$9HEF 5I/
2 2

L

-/ B
82
=
#
/#

6
()!'(@! /
#+?

5 %0
<# := =5
53 %&@76 7%'7&

.2
9C:
8
-
I:(0015
5/#%0@7%'@77 #4
#

< " -/
<
H H<%&&@9
2#O+?

5%&

/
J#/
6 @%%'@%1

$
""%! 9#/DI
/*
C
3#B%&&(- 6
%&%'%&@ .=/
#/K
+$=%&1%"

/8

5 H#$
""@%%&'%!(
"<C*G3 $
3+85##$8
5
%&&7K

6
+-D
.
/-(007K
#/
//

2

6

>#

+?

5,
#
+?

5 (1%'(&0
%,@%,'@(@ $3$"%&@,:/K+
#
-5
+ -(00& ; D4/ # C*G3
//$8

# E-/6 $/$
M3P%&1@2#/#

L
-/
6
69
EM9F"
6
;
#/#
26
*
FB4=8/

$5
3EF
$
""((1!'(&(
$668#/"
4.# K
# IN
$/ $8
M3 P%&11/ ;
83":

*=)! ": B

/
/ 3B4.9:/EF
-//35 (&'!@ +
<2 "=
D

-83"H-
$+D2
%!,'%7%
<D
-
C
/
8C%&1)" $ =M 5
.8
:/ ==%&,,

#$
""% .2

#Q 6
%@,'%1( +8
/(0,%(',(0
D
==9
<
"
8D(00!
# H
/3 5:
9
92 =%&&,.

+?

2#

5%!!'%)
//

/ +?

D
-

8%&&(

/ 6 5)%(@'%!!

$
""@ H3
HH

H

5
C

(0,'((1 "

"5=:
9
5

D ID "/
D"
D %&&! (00@
$
-
*1),,')!
.2

2

6 H H6A
AQ
/%&&1

/
6

#/

-/
#
B4:

2
36
#
"%1)1!'@0( $
""%!(7%'(7&
012 X1 3445678945
579
!"#$$% '&9')K6H#$.4J)S+'&='E'+-&
&'(&)*('+*,''-+*''( &'((&&'=,''&,''J '8&A'
%./01$23$%$43$$ 5-
5&676'&88'9- '&9')K6HI#$$F-*+'@''=
5'9:2;;<-*==+&'=+'>?@&' &&'(9''?9B(&&'=&>=''
=''++A@(B+'C'(&(,''D ( +'&,''D (E=+4.03$;
E=+#.F.34F OU*'(E96E9!VO2;;%
'&GE5''-7',7?(5+' 5'A&''=*&&+(&&'=*'+ &
2;;0HA@(*'+&&'(&B (&&'=?+ A+'A&'&?&&H
+'+A>A'',''D (E=+#% >(&&'=*9!''=&'(&
#4F3#<4 #$F#$3F2$
'I!#$$;G?&+&='&&'(& EG@@JH2;;F5'&&
)*+*,''-+*''(%%/##1<.3.;.2.%.4 )*+*'=''+'@'&='&+*''+*
7:J'@+9!E#$.<E'+'AK' ,''D (E=+#4#0#3#0.
*++A'&&K!''=&'(&2 EGHIK5S'L2;;0
##$3#04 G?''=*'+*'='&*++A
7*97- '6#$$<L@( &AA'A @('=&+*''+*&@(+*L!''=,''
&&'=9&?(&)=+','' ++#F2#F322%
D (E=+..F3$; E@'HE+'S#$.2-&A+A''=
7)&&6-#$<<-*A&&&'=>&& &'+*'''(9'=&+
'=+'&'(E'+&&&#22<320< !''=5+J '0FF#;3F2;
7)&&6-#$$%5'>&&A+& E+!6'*I85I#$.4G?'
'=++A@(,HA'M-5J $%#<.N( '=*&+*'&='*'+&!''=&'(
H&+*O=C+ &#2<3%4
7)&&6-2;#;5'('P5'A?A=' E(SH#$.$HW+'&LB&'(G?'
+'=*A'*'+*(@&=Q& 5@''='I&-5'G#.'
&+&''=A=+A&R,'' &'(G?''=&E'+&E*A*
D (E=+2##4F3#44 AA2#30;
7)&&6-'!#$.4+'A&''=&+ E(SH'J,#$F2?+&&
*'&B&?(A+&?& *',''G2%F.34##$%
!''=&'(&##FF3#<% E(SHE,!#$F<6'+&+*'
7)&&6-67S5'S#$.23 '=&=''A=+&,''-+*''(A@
#$.0?&'&'@&&++A'&'& #$F<$3#%
'A&+*''+*!''=-'>+''(B5+ E'9'(!S?9!#$.4AA+''=*'+&
-'>+''(#$$F#3$4% &'(A'C&='*+*+K''==''T(S
7)&&6-E'AAH'I2;#;5'A&''=* J*.<F<3<40
@==+?/71&+##A'+ E'AAH'IH2;;F!&@'*&+&
'(&+*-'A'*'+&+,'' +'&&+*5*'&<#34
D (E=+2#%3#2 E''G5#$.$I&D =(!&7)+
7?9J!#$$%&''=='' G@&&'+&6&J!
++A+!''=,''++F$#0223#02% H&++':#$$.5'&&'(-&='
&HS7)&&6-#$.FQ&' E'+S?'A&A*&@S
'AK''=&&(AA4#;03##% H'*6H@#$$#6'+&A@('A
+IH#$$;-*I&&E',+'+'* '=&+@=&!''=&'(&4
@'J8/ &1 .#3..
+IH7#$.2H@&&&'(?'!' H'*7E9!2;;$!&@'H*+&B
'=,''-+*''(#<%;F3%#; S&N&IC&H&9&-J740
+SH)(,#$.#E=+& -L'5'&*'*'+9E
&&'(?'B&'?&*$A' +*==&6!J#$$F5'>&*=&*'+Q
*'+&+!''=,''++%4#.23#.F !''=&'(&#;0.#30$2
+G)!#$$4E=+AA='A'+'AK +*6!I+*A#$..9''
'LB ?(&&'=S&'(++ A=+?&'&*('+*5*
G&?AA++7''AA<#3#;2 S?'AF$##043##%0
H(5#$$FT&'=&&'(& +*,#$$#-&)*+*B@
,''D (E=+42<#32.; =''='9&-K'J)&+*
&67#$$2*'''(*'(* #$$#%23%F
&+*AA#$%$3FF +*2;;$AA>7BE((&&''A
5?:5I-#$$#&'( (&&LB7H'*!E9/G&1!&@'
G?'-+*T&+'G'5H5I'+H' H*+&BS&N&IC&H&9&-
'&9')K6H#$.;E&(+*'+?''==''A= J740- L'5'&*'*'+9E
+&!''=,''&?+(&&##%$3#4< AAF;3%<
0121314516 789
!"#$#! $#!%%#!'F "/=####-

#%#& "'
##()(*(+, +?*+,)
--.#/%####0%%%##!!! @A#!--

#!/!%#H!#/"#$# !
$#! ##1 "#%#20345$! L !!"#!E###!!& "
#!/
#!(
6#7'#%8"/9"!#:#;%%*(, *-
,%%%##!!!$#! ## @"" #:K343

":4=?))
:#<#1 "!%#!#%#20& =#$###<%#/8!#"<
4'#345$!6#!74#! #$#' L ! !"#!# ###"%####
=####!>%$""%%?*) $%!'F "/=####?)*?
>#""@6?))7"8#"##:# @"" #:K34=###&3

":4
$ #6;47!"#!!#!! "AB!"A =?))>$%:#%#$##/8#
& "
#!/
#!+* !"#!$%##"#!#"<L !
#%#C
$D'"#>-5##"< A !"#!'F "/=####
!%8#E###!/##":#<<#"! *+)(
'F "/=####+*-) MA#"2K4'#&3#! ! $#
#
#"&?)),
#!/9:" =#!5 "!8"#H!##!/<:###
99"!#:##$
3# E#"/! 8!#A!$%"#!$$/8##!#'
#!A

A#9G+>! $##1 # F "/=####?*()-
%H"#$#& "'
##,)?(*? MN?))+>$%!#;4
5#$%"#;?))%%#1C0><!I #20;C$ #!" $#!##$#MH#:! ""A
4&=A#6#!7& !<8 <C
"#!03#! 4C9%A "8#"!" #!E#8<8!#""##!
J!#.##H!C!A!
544 "4K;(
54 #1%###!=#!##+=8
2#">!A#=%%+*-
#!/
##
/$%! $4#%"!4K-B?B)+
5#%%#.&
#
9
##>4,
#!/
%%"#$#8!.AB3#"##4 "
%##%$$! ! E!" $#< O9;$&5$AD2!CD$D<N
!'F "/=####?,*?? NP4/4?))(
# ! #<%
:5L%>4;#$A;5 "? #!"#!0 !! " "! /E$#
@#/!##A #! $%!%###!# &%#!#D#! $#!& "
#!/
#!
%%##-*, -*(
:5L%>4= #=4A'D ;
?))+5#I #!##"% 0>$% ##
012345678
09
565542
549
2655
!"#$%&#&%%'$ "''$&()*!$('$(!('*+%&$%$','-"&(*

%$$', '!.*)',!%'$&"..(!"$('"')(+%*%$%$/%!('$&*!$('(0"*%0*.
!('*+%&1*%$(''"&%&%1*&% 2($ 3.."')%4#%&%'!%/5%'%&".&*.% 0(&1*%6
$(''"&%)%,'"')1*%$('!('$&*!$('"&%#&%%'$%)/7"&(* 1*%$('0(&+"$
*!" ",&%%8)",&%% !".% "')(#%'6%')%)1*%$(' "&%)!* %)/
9:;:<=>?@A>B=CDEFG?G:HGIJC:;:<=>?@A>B=CDEFGFKHFH>>:H<F=E=@GDL:BG?G<:GGG=JH@C
C:;:<=>?@A>B=CDEFGFKHFH>>:H<F=HGDMNE?:@F@DLO:B=ME=@GDL:BGH@CHEK?:;:E=LL:BE?H<
GDEE:GGOHG:C=@G>:E?NEODG?@:GGEB?F:B?H?G;:BIC?MNED<FP
QR$('%"')R)%.STUUVW
09
45
4 rs//s -()7",*%*%$(' /////// gY
rs// u%!30(&-%&."#
XYZX [\]^_`\a]bc\de^]\f`\a]b /// ghi "')u(+#.%$%'% ////////// gY
XYZj `\a]b[\]a^ke\]lmâ^]n rs//V (($x%")$%%#(')%'$///// gY
o^p\q\ ///////////////// gYX rs// -()+2,*$"')(*2.%
rs/T/r t*&#(%(0$%%$ ///////// gYX *%$(' ///////////// g
rs/T/T u('*+%&v()%. ////////// gYj rs// %u"&%0*.'(&)',t&%%'$($
rs/T/w u%'$&".x(!"$('%$ //////// gYg .$%&'"$-% ///////////// g
rs/T/y z(+%{ %%$Sz{W /////// gYh rs//rU%"&%(0z".( "')z(&' ///// g
XYZg |al}\a]e^]~da]be^]dp\ rs//rrt&%6$%$/////////////// g
a\l~`\ ///////////////// gYY XYZ \q\dld\a^]k[aa^]n
rs/w/r "3 "')%$%,' //////// gYY b\\p\]n]~ f\]¡¢]~\~d\a^] /// g
rs/w/T R"+#.%R%"')R$&"$!"$(' //// gYY rs/V/r R"$0"!$('///////////// g
rs/w/w %$%,' //////////// gY rs/V/T x3%&$S,&%%8",&%%WR!".% //// g
XYZh ]\a]baa\l~[\a\ /////// gY rs/V/w #%'6£')%)*%$(' /////// g
XYZ e^]lda^] //////////////// g
rs/y/r u((',,%'!%u(++*'!"$('
"')%$R#%!!"$(' //////// gY ff\]~a¤Xk[pfl\`\[f\a¥a^][\\ //// g
rs/y/T '!)%'!%u($"')%!&*$+%'$/// gYi ff\]~a¤jk[pfl\[\\]a]bd\a^]]a\//// gX
rs/y/w R(+%#("')('$ ///// g ff\]~a¤gk[pfl\|^~dd\a^]]a\ //// gh
rs/y/y R$%# '%$',$%%"&! \\\]\ //////////////////// g
R*##.%& ////////////// g
XYZY d\a^]]a\\ab] //////////// gj
rs/s/r #% (0'$%&-% ///////// gj
rs/s/T *%$(''"&%.(&)%& 78¦7§5
96¨54
©5609
ª534
(0*%$(' //////////// gj
rs/s/w '$%&-%', //////////// gg 54

XYZ dl\^`dpê^]da]b
d\a^] ///////////////// gh )( (+"'#&()*!$0".'$%!(+#%$$-%+"&6
rs//r 5%'%&".t&'!#.% ////////// gh 3%$#."!%«(+'+%'%#&()*!$0".*&%('% $&"$6
rs//T &%-$/////////////// gh %,$('*&%$"$$%!('*+%&#%&!%-%$&(*,
rs//w { %t."'x"',*",% ///////// gh
rs//y !!% 2.$(0$%'0(&+"$('//// gY $%%'%"')&%#%"$%)%4#%&%'!%$%!"&"!$%&$!
z//x".% z/z%+"''¯:@G=BI°;H<DHF?=@=M±==C(()R!%'!%%4$R%&% ¬®
rU/rUUV²V6r6yyr6y6s³rs´R#&',%&R!%'!%µ*'% v%)"xxuTUrU

GT ĐÁNH GIÁ CẢM QUAN THỰC PHẨM PDF

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

GT ĐÁNH GIÁ CẢM QUAN THỰC PHẨM PDF

Uploaded by

Copyright:

Available Formats

01234567

  !"#$$%&'("( (%)$#(*

$)--,'"%(:&(2,+,&#-'$)-$66,"&(2, %#,-)("&,'&' &,'&'5:$%&("(:$)&'+4'&1,"#',)"$%,64887& ()'4%,

 !"#$%&&%'%( )"!*+&,-./0"$%&"1-./%&#...+ %.,&2 !%.!%

ÿ !"#ÿ$ÿ%&'&&(ÿ )

fVg̀QR ,""') 9"!*+'&,'

g A)('(!$4#/(%&$ -(H"%('!/#/8('#$4#/ #"

RSTUVUW B.*(.'/.((.*(+=>.<0=+00+/,5',==( 1+('%0+/().'.1(,5030*+I6,(((.*(,*1+('

MNOPQPR ;#*$ $'' #%:STUVWXYTS+ ##*$%:^]fgVWXYTS +  # )

ghijkjl OBH30.O]^O./9/79->09C6/8C>4.8Pmnopqr\nmsY30. O6/88YAG>9/9C39OBH E-1/83?/99/68/=3639->0>59389/6

 !"#$%"&'(")*)+% ,$#"##"!-!- $5+#+#%)"#%$#)3.6.17,#!.%"+#"'),$#+$5+#

D#)+#%J&%0%M(%+!&&%0)#!&.#)+#%($+DN)( )& bcdcbefghijkjlimnkopmfnqihrh

 !" #$$%&'%"'$'()*%""(#%'!+)$, #$$%&'%"%5'%"**)!$#)!'5'%)$$%-)2'%",'$

Z[\]^]_` 9'& $a$%%$('!$%&%($% @(defg%#fgghi"$'(.:D$%#.,;<E12@(

- $$ " "./0&

E $".  #  !*

E $".  #  !7 7)77

5   > "

98  ";!9  8@$9 $

8,,  " 97**3& !"

.-8!9  ?7BB=/ !F 9 9,! 

.- ?-3ND   K:8!! %$

#  '9 1) $9%

%9%'2! ? '!

E7BBB9  9  !999 9"9 =(!  

99%9 6*9!  59! 8 &%

?7BB9 !&! 2'9 1 

#< ;  !  >'99%99!9  1

6. &6-:! %'3?: 7BBB+7B=

Quality Control and Shelf-Life (Stability) Testing

Abstract Two routine functions of a sensory department may be quality control

Contents 17.7.1 Desired Features of a Sensory

Fig. 17.1 X-bar and R charts X-bar Chart for Thickness

Mean Thickness Rating

17.3.5 Descriptive Analysis Descriptive analysis requires extensive panel train-

(P − 1/k) − (1/2 N) these three pieces of data, three comparisons can be

Completely Very Somewhat

Fig. 17.2 A sample ballot for Apple Juice QC Ballot

too sour about right too sweet

Off aroma _X_ ___ ___ ___ ___

Off Flavor _X_ ___ ___ ___ ___

Checklist: (circle any defects)

Vinegar-like Butyric Lactic acid Painty/solvent Fusel Oil

Sauerkraut-like Other fermented Bitter Astringent Musty

Table 17.2 Ten guidelines for sensory quality testing

Table 17.3 Guidelines for participation in sensory assessments

17.6.1 Standardized Commodities

Consumer Ratings An early method for wine quality assessment was

Rancid opinion. Second, whether wine quality can actually

Table 17.5 Example of 20-point wine scoring scheme

Table 17.6 Bayesian incidence chart

Part 4. Acidity test

Good performance: All four correct.

 Again, EA can give some idea of the susceptibility

 !"#$%&! !&'"'#()##'"!&!"(#"*&)

  !"!#$%&' <%$$)&'%!%&&H'!&%')$2'&'GD<+H'

#!.*:$#$3&&3&G"!":&&/&3"? J3$#"#$%&&&"# !"(!"&# !"

 ! "# !""&"(,!""!!"

()-$!%(C'1"$"(%!"'#-3#"-$+-331-% H%'#"#"#$"%&'#J '-&"1%('""-33(<"1

 !"#$%!&" '&#!"($#)"*!"(+#)!"()",-,&"!%

#!&#!! #)-#-)"%$!"(-&?#&'-2 1!-)",6,/0)"6! 1),#&&$'&#%&16

  !"# $MK&$%7

 !"#$ ABCDEFGHIJKEF

!"#$$%&'("( (%)$#(*

!"#$%&&%'%( )"!*+&,-./0"$%&"1-./%&#...+ %.,&2 !%.!%

ÿ !"#ÿ$ÿ%&'&&(ÿ )

fVg̀QR ,""') 9"!*+'&,'

g A)('(!$4#/(%&$ -(H"%('!/#/8('#$4#/ #"

RSTUVUW B.(.'/.((.(+=>.<0=+00+/,5',==( 1+('%0+/().'.1(,5030+I6,(((.(,*1+('

MNOPQPR ;#$ $'' #%:STUVWXYTS+ ##$%:^]fgVWXYTS + # )

!"#$%"&'(")*)+% ,$#"##"!-!- $5+#+#%)"#%$#)3.6.17,#!.%"+#"'),$#+$5+#

D#)+#%J&%0%M(%+!&&%0)#!&.#)+#%($+DN)( )& bcdcbefghijkjlimnkopmfnqihrh

!" #$$%&'%"'$'()*%""(#%'!+)$, #$$%&'%"%5'%"**)!$#)!'5'%)$$%-)2'%",'$

Z[\]^]_` 9'& $a$%%$('!$%&%($% @(defg%#fgghi"$'(.:D$%#.,;<E12@(

-$$""./0&

E $".#!*

E $".#!7 7)77

5>"

98 ";!9 8@$9$

8,,"97**3& !"

.-8!9?7BB=/!F9 9,!

.-?-3ND K:8!! %$

# '91)$9%

%9%'2! ?'!

E7BBB99!999 9"9=(!

99%96*9! 59! 8&%

?7BB9!&! 2'91

#<;! >'99%99!9 1

6.&6-:!%'3?: 7BBB+7B=

Off aroma _X_ _ _ _ _

Off Flavor _X_ _ _ _ _

Again, EA can give some idea of the susceptibility

!"#$%&! !&'"'#()##'"!&!"(#"*&)

!"!#$%&' <%$$)&'%!%&&H'!&%')$2'&'GD<+H'

#!.*:$#$3&&3&G"!":&&/&3"? J3$#"#$%&&&"# !"(!"&# !"

! "# !""&"(,!""!!"

()-$!%(C'1"$"(%!"'#-3#"-$+-331-% H%'#"#"#$"%&'#J '-&"1%('""-33(<"1

!"#$%!&" '&#!"($#)"*!"(+#)!"()",-,&"!%

#!&#!! #)-#-)"%$!"(-&?#&'-2 1!-)",6,/0)"6! 1),#&&$'&#%&16

!"# $MK&$%7

!"#$ ABCDEFGHIJKEF

! "! #!$%%!%&%%!$ '""!&

G$ &$5 # D#$ ,D#$' ''H \]^^_^

0/ 8,18" , -,9" &99$

$ 99 9

!"#!"$! !%!&'(!)%"#!"*!&'+&

vwxyz{yz ET#($ klmnÿpqrslmÿ klmnÿpqrslmÿ

!"#$"%&%"''(' !")"#*+$',),- $1++"'.))',%5))1 '$,-(,)+,).,$'++

01+"#()!.+,+5+,6G.$)5!!. G 1+%)+,-!.5+(>%5)"#))1

!"#$%&''%($$#$)%*(%")($+,, +,!'0",)+)(!.()(0,)(!1!0"!!))0+-/)678)(!,

2 ' ( N6(W (a6M.CDG+0$

(6M 6XY6

HHIG3 N$6P1 M 6BCDOU

M! O%K: J

!"#$ ( @%$#L#"? (#$

:;#;<=>892 $(#%? @%/13 #?(#%" 2L#"

EFFG)$H?H=>>9F2 4(#

#E<G$%2:HC=>II "" $B"#?7===-==7=

#E<G%2:HC$#<A P$=>IDL$$#??(?((A1

#O$MH2=>>7C A$"(# )#$2 O#$78-8

"#C(%/

? HH33*F(#*R"AC2

TTUVUVWXYZ[\]^XZ_`a^bZ\[^ 1"8"!*)**%%&!(p@k"1$ ('

(( ' #

# ' '

7! )!#

!A$(-&#!02!0#4!N0A((#! A)!C:!D<E=DEE

.72889))):+ )+ !

)%# '#)'$(%"#*(!($#%-".%'#+ '# (#> $("8''%'"# ? (!"+

95 434944889