|
||||||||||
|
Table: Empirical studies on
ownership structure and performance[1]
A B C D E F G H I J K L M N O P Q R S T U V W X Y Z |
||||||||||
|
Author(s) &Journal |
Sample
& Period[2]
|
Ownership
|
Performance
variable(s) |
Other variable(s):
Controls
& dependents[5] |
Statistical
methods |
Main
results |
Preferred
explanation |
|||
|
47
large 1959-64. |
1)
MC £10% single block of voting control. OC
³10%. Dummy
for change in control from 1929 to 1963. |
Return
on equity. |
1)
Monopoly power by barriers to entry or concentration. 2) Size by sales or
total assets. 3) Industry growth. |
OLS
regression. |
OC
firms are insignificantly more profitable than MC firms. The ownership dummy
is significant and positive. Significant
controls: Barriers to entry, sales, assets, and growth. |
The
incentive alignment argument. |
||||
|
76
US, management buyouts (MBOs) 1980-86. |
1)
% pre- and post-buyout ownership by all buyout managers not including
directors. 2)
$ pre- and post-buyout ownership by all buyout managers not including
directors. |
Market
adjusted stock return in the ranges: 1) From two month before buyout to
buyout date. 2) From buyout date to first post buyout valuation date. 3) From
two month before buyout to first post-buyout valuation date. |
Pre-
and post-buyout leverage by book value of total debt to book value of total
assets. |
No
regressions. Primarily a descriptive study. However, in this study the changes
are so profound that the demand for statistics is less. |
Kaplan
obtains a post-buyout market valuation
for 25 of the 76 buyouts. For these firms the three measures of market
adjusted stock return increased respectively by 38%, 42%, and 96%. Managerial
ownership increases from 9% to 31%, but the value of the new ownership stake
is less than the old stake. Other results: Leverage increases from 21% to
86%. |
A
combination of the incentive argument, Jensen's [1986] cash flow argument and
a monitoring argument. |
||||
|
72
small and medium sized 1986-88. |
1)
Equity ownership by the board of directors. 2) MC £X% of equity ownership by directors, XÎ[10, 20,…, 90]. OC ³X% of equity ownership by directors. 3) Equity ownership by the board of directors in the ranges [0-68%], and [68-100%]. |
Return
on assets. |
1)
Remuneration of directors divided by assets. 2) Dummy for perceived need for
management systems. 3) Size by assets. 4) Leverage. 5) Dummy for secured debt
or not. 6) Growth in assets. 7) Industry. 8) Firm’s age. 9) Directorships in
other firms or not. |
OLS
regression and the Halbert-White technique to correct for heteroscedasticity.
Test for roof-shaped relation by squaring board ownership and use piecewise
linear regression. |
Performance
is significantly increasing with board ownership and this relation is significantly
roof-shaped. This is also confirmed using piecewise linear regression. Performance
also increases if directors are represented in other firm’s boards.
Significant controls: Need for management systems and director remuneration. |
A
combined entrenchment and incentive argument. |
||||
|
Same
as Morck et al. [1988]: 371
of the largest 1980. |
Combined
shareholding by members of the board in the ranges: [0%, 5%], [5%, 25%], and
[25%, 100%]. Data
from Corporate Data Exchange |
1)
Tobin’s Q by market value of stock, preferred stock and debt to replacement
cost of plant and inventories. 2) Return on assets. 3) Change in Tobin’s Q.
4) Change in return on assets. |
1)
Size by replacement cost of assets. 2) R&D costs to size. 3) Advertising
to size. 4) Long-term debt to size. 5) Firm growth as average growth in sales
1977-80. 6) Stock performance as market adjusted abnormal return. |
OLS
regression. Use piecewise linear regression. Use sub-samples of high and low
R&D firms to check directly for Q bias due to R&D. Finally, he uses
lagged performance variables to test for causality. |
Starts
reproducing the model of Morck et al. [1988]. When that model is run on a
sub-sample of low R&D firms [a better way to correct for Q bias due to
R&D) profitability is only significantly increasing for board ownership
in the [0, 5%] range. Same result is produced running a full sample less outliers
[Q >4,5] and including two more controls: Firm growth and stock performance.
In this model all controls [1-6] become significant and the adjusted R
reaches 55%. Finally, Kole runs a series of lagged OLS regressions indicating
that ownership is endogenous. |
The
incentive argument versus the reward argument. |
||||
|
The
200 largest non-financial |
MC
£10% single block of voting control. OC
³10%. |
|
|
|
Finds
that 44% of the firms in the 1929 sample were management controlled compared
to 85% in the 1963 sample. |
|
||||
|
187
of the 500 largest non-financial |
MC
£10% single block of voting control. OC
³10%. |
Return
on equity. |
1)
Size of firm by assets. 2) Industry concentration. 3) General econ. state. 4)
Industry econ. state. 5) Risk by equity/assets. |
OLS
regression. Corrects for heteroscedasticity by weighted regressions. |
OC
firms are significantly (weak) more profitable than MC firms. Significant controls:
General state and industry state. |
The
incentive alignment argument. |
||||
|
470
large industrial, 1983-85. |
1)
Five ownership concentration indices: Herfindal index and combined holding of
largest 1, 5, 10, and 20 shareholders. 2)
Six control type indices for MC or OC: OC ³5%, 10%, and 20% of
cohesive stock ownership or OC if largest cohesive stockholding has 90%, 95%
or 99% chance of winning a majority vote. MC
otherwise. |
1)
Historic market value / ordinary share capital. 2) Return on sales 3) Return
on equity. |
1)
Size by log of sales. 2) Product diversification 3) Export intensity of
sales. 4) Capital / labor ratio. 5) Age of firm. 6) Beta risk. 7) Standard
deviation of return. 8) Industry. Other dependents: 1)
Sales growth. 2) Asset growth. 3) Salary of highest paid director. |
Multivariate
regression. Also
apply simultaneous equations but without significant results. Multiple regression
analysis. Each performance variable is regressed individually. A
specification search is made over pairs of ownership structure variables
(concentration and type) that report the specification yielding the highest
significance. |
OC
firms are significantly (weak) more ‘profitable’ than MC firms with regard to
return on equity, return on sales, growth of sales and growth of net assets.
Rules based on OC ³5%, 10%, and 20% were never
significant. More concentration caused significantly less performance in
terms of historic market value / ordinary share capital and return on sales.
Significant controls: Size, export intensity of sales, beta risk and standard
risk. |
With
regard to control type the incentive argument is used and with regard to concentration
an incentive argument is coupled with an entrenchment argument |
||||
|
Levin and Levin [1982], Review of Business and Economic Research |
111
large non-financial 62
non-diversified and 49 diversified. 1967-76. |
Not
altogether clear but probably this: MC
£10% single block of common stock and not FC.
OC
³10% and not FC. FC
³10% of cohesive ownership by financial
institutions and ³10% of ownership by
non-financial, or ³5% and £10% of ownership by non-financial, or financial
institution is a leading creditor and £10% of ownership by
non-financial. Observed
1967-69. |
1)
Return on equity. 2) Return on stocks. Observed
1967-76. |
1) Monopoly power by barriers to entry. 2) Product differentiation. 3 Industry growth. 4) Firm size by assets. 5) Big foreign business or not. 6) Big business with government or not. 7) Corporate diversification. Other dependents: 1)
Standard deviation of earnings. 2) Dispersion of earnings. 3) Growth of
sales. 4) Difference between growth of sales and return on equity. 5) Capital
structure. 6) Dividends to earnings. |
Covariance
analysis. The sample is classified in order to control for corporate diversification.
Firms with substantial production in unrelated areas are classified as diversified. |
With
regard to return on equity and return on stocks MC, OC and FC firms have
significantly different returns if they are non-diversified. Otherwise not.
The direction of the performance difference is not reported and neither is
the significance or direction of the control variables. |
None. |
|||
|
Lewellen, Loderer and Rosenfeld [1985], Journal of Accounting and Economics |
191
successful 1963-81. |
1)
% ownership of bidder by bidder's officers and directors. 2) % ownership of
bidder by bidder's highest paid executive. 3) % ownership of bidder by two
highest paid executives. |
Cumulative
abnormal return, CAR by the bidding firm over the interval (A-5, R0) where A
is the announcement date of the acquisition activity and R is the merger
approval date. |
None.
|
Event
study. Initially Z-tests are applied to test for significance of CAR in the
total sample. Finally, they run univariate OLS regressions for bidder CAR on
the various ownership variables. |
Bidder
CAR increases significantly with bidder management's ownership of the bidder
company for all the three definitions of ownership. |
The
incentive alignment argument. |
|||
|
1978-88. |
1) Inside ownership by managers and directors prior to the acquisition announcement. 2)
Ownership by institutional owners. 3)
Ownership by outside 5% blockholders. |
1)
Tobin’s Q prior to announcement by market value of equity plus book value of
long and short term debt to book value of assets. 2) Six-day's cumulative
abnormal return, CAR, by the bidding firm over the interval (AD-5, AD0) where
AD is the announcement date. |
1)
Size by log of sales or log of market value of firms stocks. 2) Dummy is one
if the medium of payment in the acquisition is 100% common stock. 3) Standard
deviation of the stock return. 4) Variance of the stock return. 5) Industry (two-digit SIC). |
Event-study. OLS
regression. Test for roof-shaped relation by including the squared insider
ownership. Two-stage
least squares regression. |
In
two separate OLS regressions with market value of equity as the only control
variable Tobin’s Q and CAR are significantly increasing with inside ownership.
The 2SLS regression reveals that inside ownership increases significantly
with CAR, and CAR decreases significantly with inside ownership. Further,
inside ownership decreases significantly with Tobin’s Q, and Tobin’s Q decreases
insignificantly with inside ownership. All controls are significant. For both
OLS and 2SLS there are none significant non-linear effects. |
The
incentive argument versus the insider-investment argument. |
||||
|
|
|
|
|
|
During
the five years leading up to bankruptcy, the stockholdings of officers and
directors in NYSE and AMEX bankrupt firms are no different from those
observed in non-bankrupt firms. |
|
||||
|
|
|
|
|
|
Find
that managers make abnormal returns when trading in their firm's stock. |
The
insider-investment argument. |
||||
|
- Copyright 1997-2008, ViamInvest. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Legal notice. |
||||||||||
