|
||||||||||
|
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 1997-2010, ViamInvest. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Legal notice. |
||||||||||
