Professional Documents
Culture Documents
North-Holland
Susan Pourciau
1. Introduction
Total cites
654 Deletions:
Financial institutions. regulated companiesh
(160)
‘Routine’ executive change’
(267)
Mi scellaneous“
73
“For instance. if the President’s letter of ABC Company stated. ‘Mr. X. retired CEO and chairman
keyword
regulated companies. Another 154 vvere deleted for a
variety of reasons. includ- ing bankruptcy. foreign
incorporation, missing data, and the resignation of
more than one person from the same position over a
three-year period.
Also excluded from the final sample are 267
companies which I identified as ‘routine’ executive
changes. ‘Routine’ executive changes include: 1)
those changes which were described in the company’s
annual report or in the Wall Street Journnl as
retirements, and 2) resignations when the executive
was reported to ha1.e remained on the company’s board
of directors or continued to serve the company in some
other capacity (e.g.. as a consultant). As discussed
previously, the incentives created by routine and
nonroutine executive turnover are very different. The
theory suggests that nonroutine executive changes pro-
vide greater incentives and opportunities for earnings
management than does routine turnov’er. Therefore, to
construct a powerful test of the earnings manage- ment
hypothesis. I examine the cases in which earnings
management techniques are most likely. i.e., the
nonroutine executive changes. Vancil’s (1987) character-
ization of routine executive turnover suggests that
retirements are generally well-planned. Further. the
retention of the former executive on the board of
directors or as an advisor is evidence of a routine
executive change. For the purposes of sample selection,
companies fitting this description are classified as
routine; any executive changes not classified as routine
are considered nonrou- tine and constitute the final
sample of 73 firms.
A profile of the sample is presented in table 2.
The sample companies represent a fairly wide range
of size and performance (panel A) as well as industry
(panel C). Panel B of table 2 presents information
regarding the years of the turnovers and the companies’
exchange listings. Compact Disclosure pro- vides the
dates of the executive changes which were verified,
when possible, in the Iti111 Srreer Jourd. As shown in
panel B of table 2, the resigning manager’s last year was
1986 or 1987 for over 90 percent of the firms in the
sample and approximately 75 percent of the sample
companies are OTC firms.
For the purposes of this study, the resigning
manager’s last year of tenure is identified as the latest
year during which he/she had been in the management
position through the year as well as throughout the
three months following fiscal year-end. For instance, a
manager of a 12/31 year-end company who
resigned on 12 31,/87 had his/her final year of tenure in
1986, since the executive did not maintain the position
through the first quarter of 1988. This approach is an
attempt to identify the last year during which the
executive had control of the annual financial statements.
A manager resigning as of 12/31/87 would not likely
have had input into the determination of accruals and
other discretionary accounting decisions affecting the
financial statements that are issued during the first
quarter of 1988. For this company, the year of executive
change is identified as 1987. The year of executive
change is the first year the resigning executive does not
appear to have control over the year-end financial
statements. It should be noted that the empirical
Table Z
Profile of sample of 73 firms changing top executives over the period
1985-1988. Panel ;\ presents descriptive statistics on market value.
total assets. and net income (loss): panel B summarizes
exchange listings and years; panel C analyzes industry membership.
8: Erc~htrnyrlisriqs
Last year of Number of
Exchange Sumber of
resigning manager companies listing companies
1985 2
NYSE 7
1986 25
ASE II
1987 43
OTC 25
1988 2
Total 2
Total 73
Panrl c: Itlt/lrsrr~
Number of Sumber of
Z-digit SIC code Industry companies Z-digit SIC code, Industry companies
I Agriculture I
12 Coal I
I3 Oil and gas 5
I7 Special construction I
23 Textile-apparel 1
27 Publishing and printing 2
6
28 Chemical products
33 Steel-related I
35 Mfg.. & office II
machines. equipment
36 Electric and electronic 4
equipment
37 Autos. airplanes. etc. 2
38 Instruments and 8
photograph)
39 Toys and leisure I
50 Durable goods 4
51Wholesalers I
58 Restaurants 1
59 Miscellaneous retailers 3
71 Personal services I
73 Business service 3
agencies
75 Auto rental 1
78 Entertainment I
79 Recreation I
80 Health care services 5
87 Engineering. mgmt. 3
services
Total 2
‘Market value of equity, total assets, and net income (loss) are
all reported in thousands of dollars.
bMeasures reflect values for the last full year of the resigning
manager’s tenure as executive.
Susm
327
4. Empirical tests
Means and medians for return on equity and securtty market return
measures for years surrounding executive resignations for 73 firms
parenthesest.”
Starket-adjusted
Return on equityb security returni
Number
~~.~____
Yea? of firms Mean hledian hleun !vledian
0.1160 - 0.1351
(0.2880) (0.68781 t0.3580) (0.3155)
7 61 - 0.22s5 - 0.1333
- 0.12’8 - 0.3415
t00023t t0.0013) (0.3438, t0.0008)
equity,- ,.
accrualsl.
4.2. Results
“SW
0.15
0.1
0.05
-0.05
-0.1
Unexpected Accruals
Fig. 1. Median unexpected accruals and earnings for the period
beginning five years before and ending two years after the executive
change (year T) for 73 firms changing top executives over the
period 19851988.
The unexpected components of earnings and accruals are defined
as the first difference of the variable, after scaling by sales [e.g., UE,
= (E, S,) - (E,_, S,_ ,)I. The year of executive change, year T, is the
first year the incoming executive appears to have control over
the annual financial statements. Control over annual financial
statements is evidenced by the executive being in his/her
executive position at the end of the first quarter
following year-end.
Sum Pourciuu.
-0.02 --
-0.03 --
-0.04 --
-0.05 --
4.06 *-
-0.07 J I
Fig. 2. Trimmed means of special items for the period beginning five
years before and ending two years after the executive change (year
7-) for 73 firms changing fop executives over the period The two
1985-1988.
extreme observations for each year are deleted due to the effect of
outliers. The year of executive change, year K is the first year the
incoming executive appears to have control over the annual financial
the executive
being in his ‘her executive position at the end of the first
ILtrr T - I’
Predicted sign +
Mean 0.169
(p-\alueJd (0.264)
Median - 0.022
(p-\alueF (O.SYh)
,t 59
Yrw 7.“
Predicted sign
Mean - 0.230
(p-value+ (0.3091
Median - 0.046’
(p-value)d (0.020)
,I 64
YCLI).r + I’
Predicted sign +
Mean 0.653
(p-\alue)d (0.1211
Median 0.064’
@-valueId (0.001)
,I 53
+ ‘,
- 0.206 - 0.057
(0.641) (0.079)
- 0.066 0
10.927) (0.057)
43 66
0.870 - 0.066’
(0.773) (0.001)
- 0.075’ 0’
(0.019) (0.001)
50 71
+
I.SJO’ - 0.796 - 0.012
(0.010) (0.1691 10.1461
0.17” - 0.030 0
(0.002) (02821 10.1091
46 16 54
“The unexpected components of eat-nings. accruals, and cash Hous are defined as the first
*The deletion of the two extreme values of special items for each
year minimizes the effect of unusual values. The years for which
Each of these extreme values was more than six standard deviations
away from the untrimmed mean. It should also be noted that medians
for all the years were equal to zero. The number of observations ranged
of 50.
References
Albrecht. W.S.. L.L. Lookabill. and J.C. McKeown. 1977. The time
series properties of annual
earnings. Journal of Accounting Research 15, 226-244.
10. 3-36.
Dechow. P.M., 1992, Accounting earnings and cash flows as measures