Empirical studies on ownership and performance: CD

Table: Empirical studies on ownership structure and performance [1]

Introduction: Want to find the empirical study by Demsetz and Lehn [1985]? Just click D below and move down alphabetically on the resulting web page. Note that this page is updated when new papers emerge. Also, a few studies have blank cells. This is temporary. They will eventually be completed.

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

variables(s)[3] [4]

Performance variable(s)

Other variable(s): Controls & dependents[5]

Statistical methods

Main results

Preferred explanation

Chen, Hexter and Hu [1993], Managerial and Decisions Economics

Fortune 500.

1976 & 1980 & 1984.

Tobin’s Q by

Find that Tobin's Q increases for management ownership in the range [0-7%] and decreases in the range [7-12%]. For the 1976 sample it continues to fall in the [12-100%] range, but increases for the 1980 and 1984 sample.

Cho [1998], Journal of Financial Economics

326 of 500 large US firms (Fortune 500).

1991.

1) Inside ownership as share ownership by officers and directors of the board. 2) Insider ownership in the ranges: [0%, 7%], [7%, 38%], and [28%, 100%].

Data from Value Line Investment Survey.

Tobin’s Q.

1) Corporate investment by 1.1) capital expenditure and 1.2) R&D expenditure. 2) Size by log of replacement cost of assets or market value of equity. 3) Financial leverage by market value of long term debt to replacement cost of assets. 4) Industry (two-digit). 5) Liquidity by cash flow to replacement cost of capital. 6) Volatility by standard deviation of profit rates 1986-91.

OLS regression. Test for non-monotonic relation by piecewise linear regression and fix the breakpoints by a grid search technique that maximizes significance.

Two (and tree) stage least squares regression. Estimate three equations with ownership, performance, and investment as the dependent variables.

In two separate OLS regressions Tobin’s Q and capital expenditure is significantly increasing for inside ownership in the [0,7%] range and significantly decreasing in the [7%, 38%] range. The 2SLS regression reveals that inside ownership increases significantly with Tobin’s Q, and Tobin’s Q increases insignificantly with inside ownership. Performance increases significantly by capital expenditure and also so the other way around. Significant controls: Market value of equity, and liquidity.

Argues that inside ownership determines investment, which in turn determines performance, which in turn determines inside ownership.

Cho uses the insider-reward argument to explain why performance determines ownership.

Cotter and Zenner [1994], Journal of Financial Economics

Finds that the likelihood of a successful tender offer increases with managerial ownership and that successful tender offers are associated with significant abnormal returns.

The takeover premium argument by Stulz [1988].

Cubbin and Leech [1986], Managerial and Decision Economics

43 large industrial UK firms.

1969-74.

1) Dummy for external control that is control by non-directors and managers. 2) Degree of control of largest shareholder.

Average profit rate.

1) Size by net assets. 2) Diversification. 3) Beta risk. 4) Profit salary trade-off for internally controlled firms. 5) Asset growth rate. 6) Internal assets growth rate. 7) Industry sub-group average profit rate.

OLS and 2SLS regressions. Dependent variables are growth and profits. Checks for simultaneous effect of internal vs external control and degree of control by including interaction terms.

It is more a study on the relation between growth and profits than on ownership and performance. Positive relation between asset growth and profits but none for ownership and profits. Significant controls: 1) Industry sub-group average profit. 2) Beta risk.

The natural selection argument.

Curcio [1994], Discussion Paper, Centre for Economic Performance, London School of Economics.

389 quoted, UK, manufacturing firms.

1972-86.

1) Combined equity or voting ownership by members of the board in the ranges: [0-5%], [5-25%], and [25-100%].

2) Equity or voting ownership by the board of directors. 3) Disparity between board stock and voting ownership.

Observed 1981.

1) Tobin’s Q.

2) Total factor productivity growth or real value added by employee remuneration plus interest payments, depreciation, amortization and profits.

No controls are used in regressions of different measures of ownership on Tobin’s Q. Regressions on factor productivity growth include the following controls: 1) Log of average hours worked. 2) Log of capital stock. 3) Log of average hours worked. 4) Time. 5) Industry. 6) Market share and its growth. 7) Market share by 5 largest firms. 8) Import penetration. 9) Leverage and its growth. 10) Small firm dummy. 11) Union density.

OLS regression and a heteroscedasticity consistent technique. Test for roof-shaped relation by including the squared insider ownership and by using piecewise linear regression.

Profitability is significantly decreasing with board ownership in the [25-100%] range with regard to Tobin’s Q.

Profitability is significantly decreasing with the disparity between equity and voting ownership both with regard to Tobin’s Q and productivity growth.

Significant controls: Average hours worked, import penetration and leverage.

The incentive alignment argument with regard to managerial stock ownership, and a combined entrenchment and incentive argument with regard to voting and share dispersion.

Dann and DeAngelo [1983], Journal of Financial Economics

# stock repurchase and standstill agreements.

Find evidence in support of the entrenchment argument.

DeAngelo and Rise [1983], Journal of Financial Economics

# adoptions of anti-takeover amendments.

Find week evidence in support of entrenchment.

Demsetz [1986], American Economic Review

56 large US firms. 28 with high insider ownership. 28 firms with low insider ownership.

1980.

Insider ownership by directors and officers.

Insider trading volume.

Insider trading involvement as insider trading volume to insider ownership.

None.

Descriptive.

Insider trading involvement is 7 times higher for firms with high insider ownership.

The insider-investment argument.

Demsetz and Lehn [1985], Journal of Political Economy

511 large US firms, including utilities and financials.

1976-80.

1) Log of an Herfindal index. 2) Log of combined holding by 5 largest shareholders. 3) Log of holding by 20 largest shareholders. 4) Holdings by 5 largest families and individuals. 5) Holdings by 5 largest institutional investors.

Data from Corporate Data Exchange and other sources.

1) Return on equity. 2) Standard error of market model regressing firm return on market return.

1) Firm size by market value of equity 2) Standard deviation of stock return. 3) Standard deviation of accounting return on equity. 4) Industry dummies for utilities, financials and media. 5) Capital expenditure / total sales. 6) Advertising / total sales. 7) Research & development / total sales.

OLS regression.

Performance by accounting return is insignificantly decreasing with ownership by 5 or 20 largest shareholders or the Herfindal index. Ownership by 5 or 20 largest shareholders (or Herfindal or ownership by family and individuals or institutional investors) increases significantly by standard error of market return. Significant controls: Market value and all measures of industry and standard deviation.

The natural selection argument.

Denis and Denis [1994], Journal of Corporate Finance

72 US firms with above 50% insider ownership by managers and directors. Data from Value Line.

1985.

1) Majority ownership- >50% insider ownership by managers and directors. 2) Institutional ownership. 3) Dummy for outside blockholder ownership. 4) Dummy for family or founder involvement in management or board of directors.

1) Return on equity. 2) Return on assets. 3) Operating income to assets. 4) Tobin's Q. 5) Market to book ratio.

1) Fraction of outside board members. 2) Board size 3) Dividend yields. 4) Debt asset ratio. 5) Number of public securities offerings. 6) Size by market value of equity. 7) Variance of stock returns. 8) R&D to sales. 9) Dummy for dual class shares.

Standard t-tests are applied to test for differences between the main sample and an industry paired control sample.

No difference in performance between majority controlled firms and other firms. The likelihood of majority control increases significantly with family or founder involvement Indeed, 80% of majority controlled firms have substantial family or founder involvement. Majority controlled firms have significantly: 1) less outside directors. 2) less outside blockholdings. 3) less institutional shareholdings. 4) pay less dividends. 5) Dual class shares. Other controls are insignificant.

The entrenchment argument and the 'natural selection, argument.

Denis and Serrano [1996], Journal of Financial Economics

Find evidence that active monitoring by outside blockholders helps to increases shareholder value in the two years after an unsuccessful contest.

[1] Some of the studies have investigated other issues as well, such as, the relation between ownership structure and the risk of the firm’s performance.

[2] The reported period typically refers to the maximum period that a particular study applies. Often the performance variables are collected over the entire period, whereas the ownership variables and control variables are collected at one year in the investigated period. All studies use publicly traded firms (unless otherwise described), because they are easier to get information about.

[3] Abbreviations: Management control (MC); Ownership control (OC); Owner managed (OM); External control (EC); Strong owner control (SOC); Weak owner control (WOC); All owner control (AOC); Financial control (FC); Majority held (MH); Diffusely held (DH).

[4] The ownership variable is typically measured as concentration of ownership on a particular set of owners, e.g. ownership by managers or institutional investors.

[5] This colon includes 1) independent control variables, 2) dependent variables that are not performance or ownership variables, and 3) variables used for sample classification.