You are on page 1of 16

Seies

Tme Senes

IntroduenoD
One Of the imporio moSt
tasKs boore EcongmiSt8
Businessmen those IS
day to ma Ke estHmaHe fer the Putune
e.g A bus1nesS man interested
gales yedr 2023
inding
in
hs brely 0

a
long term ploming in. 2028 0r
the year 2035 80 tht he Could
ddjust "is produeton
or
accordgy
inadequcte
dvoid
producHon to
the

meet
passibiby
the dema nd.
Of e ther unsold Stcks

Smlorly dn economist IS inerested


estimating the

lvoly populnon the coming yedr So that


Can be ¡ned With proper plonning
regard t0
fod supply, jobs for
the
people ete.

very neressar to esttmdte feltlre foY the

bus1ness poin eCcnomic poit of View.


Jo es+ mcte pmdrt the future 1 the irst
Step is
informaton from the past Ths the fTSt
gathe ing
trend the futtire
8tep.
this connecHon ) SHatisical daa which aTe collected,cbd
obserrod TRCOTded dt SUesSiVe interyals of Hme.
tim Such
dasd generally rererred ds Tme Series
()
Thus, when we obgerve numerical daa at diterent
point of time the set Of
cbservatHons known

eTime Series the Set 0f dasa wh1ch is collec+ed


a regular interval 0f Hme.

Different deeiniHons Time Series Ore.

Series set of Statistical observatHons.


dTrAnged in chrono logica ordor. order of OcouTance.

* Tme Series consists of StatiStHcal data wbich


collecied recorded 8 observed SUCosSi ve increments

0f me.
per
e8 populaton, capral income price producHon for

the last 5 |0, 15, 20 year So rme other tlme


feriod.
wouldl be called Hme Series.
the senes emerg
The term Time Series not applcable to Business or

Eonomics only Cdn to all other phenomena


dre related to Hme
eg' aeidents in
day a parica ar
the tempercttture
Hme
period the yamaton
pettent
CerteLin
period no- of marmiage toking ploce
duving This d dbout Hme
duri ng
Certein
period ete. Seneg.

Jime Senes , what we do on te

fture. Ths
bag1s of past clata, we proaie TIme
the

of hme Series Analysis Senies 1S diererent


purpose
8 Time Senes drtrerent
(2)
Past
l1me Series nedns, the StatisHcal dota which JS drranged

acerrding the
t0

help
tIme

of Hme
8 Time Sories
Can
prodrt
medns

Puture.

year Sales of TefrgeraD TS n thoudod


1446 40
1941 42
14 48
1944
41ale s deoreased ten Inereased.
43
2001
48
2002
65
2003 42

We observe the ab0Ve senies We fnd fhat gereraly the


salos have incraosed but for 2YearS|9g-2000 decreased.
The Setsicran +ries to onalyse the effect VamouS

forCes.
Of ty pes VanaHons changes Can be

4 ypes
) Seoulor mO vemen t8:~
Changes that have ocourod result
the daa t0 inerease 0 r
general tendency deoreayet s

long term Tegulor movoments- eg populaton ehange, tæhro kgcl


progress:
Seasonal Vanatons : change tht have taken
placo dme
period
weather
Of 12 mo nths
conditlons festivals
resultof changein
etc. eg iceeream Sale
climate, besaa

yolical VamiaHons: Changes that hgve 7esult


taxen
ploce
booms 8 depresions. Such changes dre classt
clasSt fred
Pred funder He
These due
cyclical natons.
VamatonS vamattons de te
boom, depreSion
Tecesion in the Economy
(3)
(4) Iiregulor | FrraG VanatHons

chonggs that Tesult sud


have te ken place
ferces which Can not be prodicted rue loods, eorthquaves,
amines (lack oe food long period time. in a large
that
mang people )
Time Soiea
Analyas
() I bolps in undlersia neing ast be havour.
(e) I helps in planning operaHong.
, (3) I holps in accempush menrs
(4) II
evaluang
CUrrent

which we can fiod


fucil res comparison. sing
erpected pereermance/ trend
to drow Conclus1ons. in nestt 5 yers
perpermance.
(5) Ii helps pred1cg Pture.
2 uing which we can nd
CUrrent aecemphshment

Componenrs fatern Movemens Elements Of Tinme Series

) Secular Irend - long rerm movement eg popadaHon change


term hange tesbnolog1ca progres
) change
Seaso nal Vananon- penoduc mÙvement8 which ocours due to in.
OD,whicb occur
e-g.climate , weaher condHons regularlydunng
everyapenod
yea of |2 moniE
3) Cyelhcal Vamiaton;elated to oonnnmis 8 buinegs
long term moremet hch
repedt
CCurs due er oscletory up-down 8 all bubin9S
4) Iregular, Vaicon
ErraHc, Aee,denal
ISe then decline, r1se hen,y Sale of Ice cream
decl1 ne.
Random - do not repeat eg business cyole
a deote patern eg Aloods, eatthquare, Stri ves Wars
Suddon
change
in domand
Relanonslp the Com ponent8. Coro na
very inregulor &unpredeiable
ßr very () TradHonal classcal +1me onalygs - There is
Series onalyss
rap1d bes
tech
o mabipdcaoe elaronslup these 4 Componens
progres

Another thougla that T+S +C+I


Medsurement 0fOf Trend.
The Vamous methods that be fer
deermining rqend
ae.
Freehand or Graplic me tod
O) Semi- dverage metod
(3) Moving overage method
Method of Least Se ua re.

Graphc Meitod 3
S1mplest method of study1ng rend.
s- ) Plo+ the time senes on
Plo graph.
Esearmine the direcHon of the trerd
(8) Draw d St raiglt Iine which wll bes+ Ar to the dase.
Ex Fit a trend ne t0 the falloing dewa the freehand
method.
year Produoton of Steel
Iq45 29 Trend by Freehandt method.
20
1446 22

0q47 24
25
|948 21

23 23
9000
Trend
95
ne
200| 23
2002
26
1996 l497 A98 I999 200 200|02 2008
2003 25
Yeans

(5)
2 Mehod of Semi dverage8 -
Steps:- ) D1vicle he data wo equal pors.
eo Suppose you are given 18 yeans of doua
from |968 to 9003,. divicle it equal
from 1986 I0 |494 from 2003.
Case
yeans Live. 9.13, 1 ee ) Omit the mddle
year then divide it in te 2 equal
1S Ivn tor from |985 to 20 03

then wo equal be |985 te 194 8 hen


from |945 t0 2003. middle omiHed.
the year 1q44
2) Find the d verage Of each part Thus 2 poins
3) Each Pins is plated the
mdpoint 0f class
4) Then
poims we ge Straught Jine wbch is
the trend line.
reqë red
5) The line Can be escHended downwards
upNords
predict future values.
Ex: Fit a trend Iie to the folbwing dae the me thod of Sem -averag.
year
(thoysand
Sales.
units)
Areroge
|02 Thene 1s a
|998 trend
1944 14 120
Tmsing
2000 Omited
2001 |08
)5 Acal daa
2002 116 |12

2003 I12
Trend Ine
|05

1996 2001 2002 2008

(6)
3 Methed
Mong rage
whle
period
appyng ths method select
for moving dverage such
8uch as 3- yeory movig overage
55-
yeorty moving average, 8- yory moving
dverage.
Data a, b, c, die, f,g, h
The
8-yeorty moving average be determined by
d+ bsc btc+d C+d+e d+etf
3 3 3
The
5-yearly moving overage con be determined by
d+ b+c+ dte btc+d+ e+f c+ d+etf+o
5 5 5

Ex Caleulde B- yeary moving overage s trend vaue .for the


Polbwig da+a
value tota!
( Trend Valuss )
8 yeorly moving 3 yery mo ving average

|492 18 18]3 = 6
I943 21

24
I944 94)3= 8
I995 8 31 3)3 = l0-33
1946 J4 g5 353 - I1-66
J2+
|947 13
Acrual daq
8

Trend line

1441 1992 I993 1944 |445 1996 444


(7)
2010 2009 2008 2007 9006 2005 2004 2003 9002 2001 20ho
deua Aal J0+
line. Trend
20
25
95 2010
20 2004
I4 87/5- 81 9008
16
l4-6 73/5= 13 12 2007
2-6 63/5= 63 LI4 9006
I|65/5= 55 9005
11·4 57/5= 57 Q004
0 50)5 60 2003
8"4 425= 42 -14 9002
900|
arerage meving yeory 6 2000
toja! moving yarly 5 value year
da.Aollowing
values trend obrain the for
2 age ave moving yeorly 5-
aleuae Ex
Ex. Caleulate the 8-yearly moving ave rages the producon
Pigures given below draw a trend line
yean prodcHon y 3 yeorly reral 3 yeory moving overage
1945 15
2| 66 22
I946
87 29
1991 30
|08 36
1998 36
j24 41-33
19949 42
138 46
46
152 50-67
2001 50 56-33
2002 56 l69
63
9003 63 189
2004 207
2005 226 75-33
74
2006 82 246 82
2007 267 | 84
287l| q5"67
2008
2004 l02
of the number oF Students
Ex Construct 5- yeory movi ng averages
college shown below:
studying
yeam No- of 8tuden ts 5 yeorty toral 6 yearly moving averoge
1948 332
817
1444 1800 360
9000 357 374·6
1873
9001 342
402 1966 393.2
2002
2003 405 2036 A07-2
g049 409-8
2004 410
2005 427 2085 41
405
2009
9007 438

(9)
Calculae 55yeardy 3 1 yolymoving averages foT te
Commercial 8 Industrial falure
failure in
following dua.
dara. of the
country dhing 1944 t0 2004

Yerr No-of frulure 5ycarly 65 yearly yearly


moving totaf movi ng
average moving
moving
tor average
1944 29
1945 26
129 25-8
I996 28 2)-9
118 236 I53
1941 32 20-0
104 208 140
20
1998 123 17-6
|2 86 15"4
1999 I08
12.6
2000 12 63 l2.4
T|-2 87
200| 56
81
2002 55
81
2003 13 57
|·8 18
2004 59
2005 14 54 |18
63
2006 12 64 q-8
2007 39
2008 3
9004

Even feriod 0f Moving dverage I the moving average


IS dn even
period moving dverage. 4 yearly 6yearty
the moving torel 8 Nhich are placed dt the
moving average
centre of the Hme from wbch th fall
computed
between wo time
periodg. This placement inconv Ènient S1nce

the mo dverane Wouldl n0t Coinide with the original Hme


poriod We thereforO, sychronise movigg averages origina) dasa.
Ths proCOSS is called centering 8 always onsiStS 0f ta king
period moving overoge Of the moving averages.

(10)
There dnother me thod 0f cenrering the noving dverages. I
We. aTe caleulasing 4 yeurly moving averngo 4-yearly
ta ve

totals these totals We


ogun tel ke 2-yeury totals
8 divide these totals
bybu 8.
e year Value 4 yerly centered
moving,
2000
foral( moving aVerge overage
200 | 87
362 q0.5
9002 l05
34 43-57
9003 >q0.75
J00
352 88
2004 82
296 65.7
2005 65 57·5
230 ’49o
9006 49 42.0
l68 84-75
9004 34

2008 20

9009

Ex Asume 4-yearly eyele 8 caleulate the +rend the

method of moving averages ron the Following daya.


production- 4yearly
tote). mov1ng
'averdpeoving
4yeorle 4 yeary
'averdaemoving
enered.
9000 464
900| 5)5
1464 491 --445o
2002 518 2002 5005
467
--503 625
9003 506.75
202 5||: 625
2004 502
2066 516-5
529- 5
2005 540 - 542.5
557 5530
2006 2254 563- 5
2007 57| -572.5:
23 26 58|·5
2008 586

2004 612
The method Of Least Squa res
This method is most used prarte. T is
mathe matical method trend line is
fitted to the daHa
WIth
help a

such a mannr that t he


following 2
Condittons Satisfied.

)2y-y) =o ie. the Sum of deviations 0 the actual values


Of the ompused values of y is o
2
2) (y-ye) is least ie the Sum Of Squa res 0f the deviatons
0f the acual computed values is Jeast from tis line
bence the name mathod of least squares
The ine obrai ned b this method know n as the line
of best fit.

The method Of least Sq ucares may be used either eit c


St. line irend on parabol1c trend.
The st Iine trend is ropresented the equaHon
e a t bx
Where usedto designate the trend vaues to
iS

the acteia Y valuag.


desinguish
them from IS the
y itercept or he
Compused trend figure f the Vanable when X= 0, b represents
the slope of the trend line

I order to determine the valus Of the Constens

the
tollowigg 2 normal equns be solved.
Zy= Na+ bbIx Where N= no of years months
2xy = ax +bx2 time

(12)
Whenever Ne St line trond by te leost squares
method , thre things shouwd be speifcd
() Which year Was selected as the oTigin ?
(9) What is the Of time Tepregented by x? T iit hals yean,
year
One five years ?
(3)In what kind units is y being meagured
9
Is ir it

procucton tonneS salos in rupees, price rupees ,


employment in thousa nds of

We Can meagu re the vamable x from any point of Hme

in oTigìn such as the Arst year But the calelctons ae

nhen the
much
smplnid in Hme is taken, as

the origin beccuu se that Case 2X= 0

X= 0) Nomal equns

2y= Na
2xy = b2x
Uog whch we Can the vaues of a b

2
N

should be noted thdt. OF odd no- 0f yearS 1f


the midadle year hen
from
the deviatons. dYe ta ken

in the daHa gven.


Sx=0,
However
prowdel there gap In Case

of even
years also 2x=0 If the deviattons dre ta ren from the
m1dpoint bet ween the two middl yea rs.(3)
eg. the years dre 2004, 2005, 9006, 2001, 2008 , 200g . then
we take
devidttons from the middle
year 2006- 5 . Thus
the deviaHons NoLld be -2.5) -}5, -05 0·5) 5) 2.5. or
1

the nous yearS o Sx= 0.


vanous

Both in odd as well ds 0n


no- of yeaTS, we can
the simple procedure Of
derermi ning the vdlues of the Constn8

Below are given the figures producton (mllian tonnes) of


Sugar fociOTY
2003 2004 2005 2006 9007 2008 2009
froducon 80 40 4283
( m. tonne ) 44 42.

() it a St line trend to these Aigu res


iu) Plot these figures On
graph 8 show the trend. Jine
N=1 630
xy = 56
Ye= q0+ 2x
Yc= 84
X=-2 Jc = 86
X= Ye= 88
X=0 45+
Ye 90
0+
X=| Yc = 42 Trend ne
85*
X= 2
Ye = 94 80
X= 3 75
lctual data.
Y
2003 2004 2005 9006 2007 2008 09

(l4)
If instead af middle yer as oTigIn, we tare fITSt yean asarigin
N7 y: 630 2xy = lq46 Zx': q) IYc= 630 (o

630 = 7a+ 91b


1446= 9)a+ b
) a 84 b= 2
Thus he eq uaHon Ye- 84+2xX.
Ongin 2003 X unit8
snits
produemon million Honnes
X=0 Ye= 84
Note : The di ference the two eq uatonS because Of he
d1sterences in ornigin. n he iTSty Case 2006 Was as origin
taken
whereas the seoo nd Ca se 2003 Wag taken
origin. However,
trend yalues are the same
Ex. Fit a St. line trend Por the following
serieg. Estimate the vaue
for 2012 .
year 2001 2002 20032004 2005 2006 2007
Producion 60 72 15 65 80 85 95
0f Steel
( mtonnes) a: 6 b= 4-857

YE t6+ 4-857X Y9012 = 6 +| 4-857(8)


Ex Fit a st. lioe trend by the method =of l14- 86
edst Sq ua res- to the
followi data. what would be the predied eaT nings fer the
2009 9
sing year
2000 200| 2002 2003 2004 2006 9007
2005
40
Earnings 38 65 12 69 60 87 q5
a= 6575 ongin = 2003-5
b xy 3-667 65:15 t 3-667X a= 6Ó. 75

be 7333 .
Bor 2004 X=
X= 5-5
Y= 055 + 3-667 () = l06-081
(5)
X
X
88803+·38 X
03 45 ·38 b=
563 4|· 51b d+ l59-
803 88a=
183 40 +b 7a 623=
803 88
423 87
I baZx+ Zxy=
043 86- bIx +Na2y=
83.283
362 =Yc2 204 Onigyn:
51ZX 1592xy= Zx=| 623 Zy- N=. tonnes) m.(
48 85 44 88
ProducHon
2009 2006 2005 9004 2003 9002 2000 year
acHory Sugi
a OF
s) m1llion (producHon
in Figures
of given Below B,
the

You might also like