Pacific-Basin Finance Journal 13 (2005) 431 – 449

www.elsevier.com/locate/econbase

Ownership concentration, firm performance,

and dividend policy in Hong Kong
Zhilan Chen, Yan-Leung CheungT, Aris Stouraitis, Anita W.S. Wong
Department of Economics and Finance, City University of HongKong,
83 Tat Chee Avenue, Kowloon Tong, Hong Kong
Received 25 July 2002; accepted 6 December 2004
Available online 23 March 2005

Abstract

We analyze a sample of 412 publicly listed Hong Kong firms during 1995–1998 in order to
answer three questions. Does concentrated family ownership affect firm operating performance
and value? Does it affect dividend policy? What is the impact of corporate governance on
performance, value, and dividend payouts? Our results do not show a positive relationship
between family ownership and return on assets, return on equity or the market-to-book ratio. In
addition, we find a negative relationship between CEO duality and performance (where CEO
duality is much more likely in family-controlled firms). We also find little relationship between
family ownership and dividend policy. Only for small firms there is a significant negative
relationship between dividend payouts and family ownership up to 10% of the company’s stock
and a positive relationship for family ownership between 10 and 35%. Dividend payouts in small
firms also show little sensitivity to performance. Finally, the composition of the board of
directors (proportion of independent non-executive directors, outsider-dominated board, presence
of audit committees) has little impact on firm performance and dividend policy, particularly for
small market capitalization firms. Our results for Hong Kong are in line with both Demsetz and
Lehn (1985) [Demsetz, H., Lehn, K., 1985. The structure of corporate ownership: causes and
consequences. Journal of Political Economy 93, 1155–1177] and Himmelberg et al. (1999)
[Himmelberg, C.P., Hubbard, R.G., Palia, D., 1999. Understanding the determinants of
managerial ownership and the link between ownership and performance. Journal of Financial

T Corresponding author. Tel.: +852 2788 8960; fax: +852 2788 8806.
E-mail address: efsteven@cityu.edu.hk (Y.-L. Cheung).

0927-538X/$ - see front matter D 2005 Elsevier B.V. All rights reserved.
doi:10.1016/j.pacfin.2004.12.001
432 Z. Chen et al. / Pacific-Basin Finance Journal 13 (2005) 431–449

Economics 53. 353–384], who show that concentrated ownership is not associated with better
operating performance or higher firm valuation.
D 2005 Elsevier B.V. All rights reserved.

JEL classification: G32; G34; G35

Keywords: Ownership structure; Firm performance; Corporate governance; Dividend policy

1. Introduction

Does ownership concentration affect firm performance and value? The finance
literature has been trying to answer this question since Demsetz and Lehn (1985) found
no significant relationship between ownership concentration and the firm’s return on
equity. Other early studies in the U.S. highlighted a positive relationship between
ownership concentration and firm value (Tobin’s Q) for low levels of ownership, which
can be attributed to the alignment of managerial incentives with shareholder interests.
They also found a negative relationship at higher levels of ownership, which can be
attributed to managerial entrenchment, since managerial shareholdings confer to manage-
ment, among other benefits, protection against hostile takeovers (Morck et al., 1988;
McConnell and Servaes, 1990; Hermalin and Weisbach, 1991).
On the other hand, in Japan, where firms are subject to monitoring from banks and
takeovers are rare, the positive relationship between managerial ownership and firm value
has been shown to be monotonic and holding for all levels of ownership (Morck et al.,
2000). Similar evidence have been obtained by Hiraki et al. (2003), who found that
managerial ownership is monotonically and positively related to the value of Japanese
manufacturing companies, and by Chen et al. (2003), who showed that as ownership
increases, there is greater alignment of managerial interests with those of stockholders for
a sample of large Japanese firms.
However, there is little evidence on the relationship between ownership concentration
and performance in South East Asian countries, despite the fact that many economies in the
region are characterized by considerable family ownership of listed corporations (Claessens
et al., 2000). Standards of corporate governance and investor protection are also lower in
the region compared to the U.S. or Japan (La Porta et al., 1998), which adds a potentially
interesting dimension to the relationship between ownership concentration and firm value
or performance. In countries with poor investor protection, controlling shareholders may
have the opportunity to expropriate minority shareholders.
Furthermore, Himmelberg et al. (1999) have cast doubt on previous findings by
suggesting that the observed empirical relationships between ownership and performance
may be the result of unobservable firm heterogeneity, which may affect both ownership
concentration and firm value. They show that regressions of Tobin’s Q on ownership
concentration may be misspecified, because some unobserved determinants of Tobin’s Q
are also determinants of ownership concentration. These unobserved exogenous firm
characteristics may induce a spurious relationship between Tobin’s Q and ownership.
Consequently, they find no relationship between ownership concentration and firm value
after estimating firm fixed effects.
 Z. Chen et al. / Pacific-Basin Finance Journal 13 (2005) 431–449 433

Our article analyzes a sample of 412 publicly listed Hong Kong firms during 1995–
1998 in order to investigate three questions. Does concentrated family ownership affect
firm operating performance and value? Does it affect dividend policy? What is the impact
of corporate governance (CEO duality, the composition of the board of directors, audit
committees) on performance, value, and dividend payouts in family-controlled firms?
Hong Kong is an appropriate market for conducting our study because it is characterized
by widespread family control of publicly listed corporations, while at the same time having
a common law legal system and corporate governance influenced by recent developments
in the UK (Cadbury, 1992).
We report three main findings. First, our empirical analysis does not show a positive
relationship between family ownership and proxies for firm performance (return on
assets, return on equity, and the market-to-book ratio). Therefore, our results for Hong
Kong are in line with both Demsetz and Lehn (1985) and Himmelberg et al. (1999), who
show that concentrated ownership is not associated with better operating performance or
higher firm valuation. We also find a negative relationship between performance and
CEO duality (defined as the same person holding the positions of company CEO and
chairman of the board of directors, which is much more likely in family-controlled firms).
The latter result is consistent with managerial entrenchment in companies that combine
the positions of CEO and chairman of the board. Second, we find a significant positive
relationship between family ownership between 10 and 35% of total company shares
outstanding and the dividend yield for small market capitalization firms. These firms also
exhibit low sensitivity of dividend payouts to performance. This evidence suggests that
controlling shareholders may be viewing dividends as way to extract resources out of the
firms they control because dividends make up a disproportionately large part of the
income they derive from these firms. Finally, the composition of the board of directors
(proportion of independent non-executive directors, presence of audit committees) has
little impact on firm performance and dividend policy, which suggests weak corporate
governance mechanisms.
The rest of the paper is organized as follows. Section 2 describes the data, the variables,
and our estimation methodology. Section 3 presents a descriptive analysis of the sample.
Section 4 presents our multivariate analysis of the relationship between family ownership,
firm performance, and dividend policy. Section 5 concludes.

2. Data, variables, and methodology

Our initial sample consists of all companies listed on the Stock Exchange of Hong
Kong (SEHK) in 1995, which we follow through 1998. After deleting firms with
missing data, the final sample includes a balanced panel of 412 firms (representing
approximately two-thirds of all firms listed in 1998), i.e., 1648 firm-years. All data were
obtained from company annual reports and from Company Analysis provided by the
Financial Times.
Firm performance is measured by three different variables. These are return on assets
(ROA; net profit divided by total assets), return on equity (ROE; net profit divided by
shareholders’ equity), and market-to-book ratio (market value of equity divided by book
434 Z. Chen et al. / Pacific-Basin Finance Journal 13 (2005) 431–449

value of equity). Payout policy proxies are the firm’s dividend payout ratio (total
dividends divided by net profit), and dividend yield (dividend per share divided by share
price).
Ownership concentration is measured as the fraction of total company shares
outstanding held by the controlling family. To the extent that there are family
members or friends who hold shares that we are unable to trace, our variables may
underestimate controlling shareholdings. Motivated by Morck et al. (1988), to allow
for non-linearity in the overall relationship between firm performance and ownership
concentration, we define the following three variables (assuming that the actual family
ownership fraction is m)

Family Ownership ½0; 0:10 ¼ m ðif m b 0:10Þ; 0:10 ðif m z 0:10Þ

Family Ownership ð0:10; 0:35 ¼ 0 ðif m b 0:10Þ;

m  0:10 ðif 0:10 V m b 0:35Þ; 0:25 ðif m z 0:35Þ

Family Ownership ð0:35; 1:00Þ ¼ 0 ðif m b 0:35Þ; m  0:35 ðif m z 0:35Þ

Our choice of cut-off ownership levels in this piecewise linear specification is

motivated by characteristics of the Hong Kong market, namely the minimum ownership
level that necessitates disclosure of substantial shareholdings (10%), and the level that
triggers a mandatory general offer (35%)—although firms regularly obtain a waiver from
the latter requirement. An examination of the frequency distribution of family ownership
in the sample (reported in Fig. 1) also shows a disproportionately large number of firms
with family ownership just under 10%. We have also replicated all regressions by using
cut-offs at 10, 35, and 50%; and the 5 and 25% cut-off points employed by Morck et al.
(1988). Our results are unchanged when we use different cut-offs.
Our board of directors proxies are CEO duality (dummy variable equal to one when
the CEO also serves as chairman of the board of directors), the natural logarithm of the
number of directors on the board, the fraction of independent non-executive directors, a
proxy for outsider-dominated board (dummy variable equal to one if the fraction of
independent non-executive directors on the board exceeds 50%), and the presence of an
audit committee (dummy variable equal to one if such a committee exists). A priori,
we expect that firms with smaller boards of directors, larger proportion of independent
non-executive directors, audit committees, and separation of CEO and chairman
positions will have higher value and better performance. The characteristics of these
boards of directors may also be expected to limit the manipulation of dividend policy
in ways that expropriate minority shareholders. Finally, our regressions include as
control variables additional firm characteristics that may affect performance and
dividend policy, namely the debt-to-assets ratio (long-term debt divided by total assets),
annual sales growth, and firm size (natural logarithm of deflated total assets).
Himmelberg et al. (1999) show that regressions of Tobin’s Q on ownership
concentration may be misspecified, because some unobserved determinants of Tobin’s
Q are also determinants of ownership concentration. These unobserved exogenous firm
characteristics may induce a spurious relationship between Tobin’s Q and ownership. For
 Z. Chen et al. / Pacific-Basin Finance Journal 13 (2005) 431–449 435

200

Frequency (number of firm-years) 180

160

140

120

100

80

60

40

20

0
0% 10% 19% 29% 38% 48% 57% 67% 76% 86%
Family Ownership (percentage)

Fig. 1. Distribution of family ownership in Hong Kong (1995–1998). The figure shows the frequency of
observations (number of firm-years) associated with different percentages of family ownership, for a sample of
412 publicly listed Hong Kong firms during 1995–1998 (in total 1648 firm-years). Family ownership is the
ownership fraction of the controlling family, aggregating the shareholdings of directors with the same surname.

example, firms operating unrecorded intangible assets may provide larger amounts of
stock to their managers because intangible assets may be subject to more managerial
discretion. Unrecorded intangible intensive firms will also report low book value of
assets. Therefore, the inclusion of such firms in empirical analysis may induce a
spurious positive correlation between Tobin’s Q and ownership concentration. The use
of fixed effects estimation can solve this problem, assuming that the unobserved firm
heterogeneity that induces the spurious correlation is a firm fixed effect. Following their
approach, in our empirical analysis, we report results for three different types of
regressions: using all firm-year observations pooled, estimating industry fixed effects,
and estimating firm fixed effects. All regressions incorporate year dummy variables.
Reported p-values in the tables are based on White (1980) heteroskedasticity consistent
standard errors.
Our empirical results are reported under two caveats. First, although the use of firm
fixed effects represents a methodological improvement in theory, it may nevertheless fail
to capture any relationship between ownership and performance if there is little within-
firm variation in ownership over time. This is actually the case in our sample. The
variation is even smaller than the one reported for the U.S. by Zhou (2001), because
concentrated shareholdings appear stable over time. The within-firm, year-to-year
absolute percentage change in family ownership is less than 10% for more than three-
quarters of our Hong Kong sample (compared to 47% of Zhou’s U.S. sample).
Therefore, a large proportion of Hong Kong firms shows very little variation in
ownership over time, and the results of the firm fixed effects specifications should be
interpreted in this light. Second, Chen and Steiner (1999) have suggested that in an
436 Z. Chen et al. / Pacific-Basin Finance Journal 13 (2005) 431–449

agency framework, ownership concentration, capital structure, and dividend policy may
be endogenously determined. Similar arguments of an endogeneity of ownership
structure have been made by Cho (1998). If this is actually the case, then in an empirical
model like ours, ordinary least squares parameters may be biased and inconsistent.

3. Ownership concentration and firm performance in Hong Kong: Descriptive

analysis

Descriptive statistics about the sample are reported in Table 1. In panel (A), we observe
that the majority of Hong Kong firms are closely held: 75.6% of the firms have a family
controlling at least 10% of voting rights, and 61.1% have a family controlling at least 30%
of voting rights.
In panel (B), we observe that the boards of Hong Kong firms have 9.4 directors on
average. Following the publication of the Cadbury committee report on corporate
governance in the UK (Cadbury, 1992), the listing regulations of the SEHK stipulated the
mandatory introduction of at least two independent non-executive directors on all boards
from 1995 (they must hold less than 1% of the total issued share capital of the listed
company, they must have no past or present financial or other interests in the business of
the listed company or its subsidiaries, and they must be free from any relationship that
could interfere with the exercise of their independent judgment). On average, there are
only 2.8 independent non-executive directors on our sample’s boards, representing 30% of
board membership. Even this number probably overstates the proportion of really
independent directors, since the corporate sector in Hong Kong is small and it may be
difficult to recruit true outsiders. A priori, the small number of independent directors raises
doubts on whether boards are sufficiently strong and independent as to have an impact on
firm performance and to perform adequate monitoring functions. Also, in 52% of all firms
in our sample, there is CEO duality, i.e., the CEO is also chairman of the board of directors
(in figures not shown in the table, out of all firms with CEO duality, 76.5% are firms where
a family controls at least 10% of voting rights).
Audit, remuneration, and nomination committees are not common in the sample. Only
in 1998 did the code of best practice of the SEHK’s listing rules include guidelines for the
recommended introduction of audit committees. Just 1.7% of the companies in our sample
had audit committees in 1995, and 6.3% had audit committees in 1998. Remuneration
committees are not mandatory. Only four companies in our sample had remuneration
committees. Nomination committees have not been introduced yet.
Descriptive statistics about the remaining variables are reported in panel (C) of Table 1.
In our sample, the mean market-to-book ratio was 1.4. This compares with a median
Tobin’s Q ratio of 1.0 reported by La Porta et al. (2002). The mean ROA in our sample
was 1.6%, mean ROE was 4.2%, the dividend payout ratio 47.2% and the dividend yield
4.8%. La Porta et al. (2000) report a median dividends to earnings ratio of 45.9% for a
sample of large Hong Kong firms in 1996.
An examination of correlations between the variables (not reported in tables for brevity)
does not show any large correlations. The largest correlation coefficients are between total
assets and the number of directors on the board (0.35), total assets and the debt-to-assets
 Z. Chen et al. / Pacific-Basin Finance Journal 13 (2005) 431–449 437

Table 1
Descriptive statistics for sample firms
Panel (A)—Ownership concentration
Family ownership [0, 10%)
Firm-years 402
Proportion of sample (24.4%)
Family ownership [10%, 20%)
Firm-years 110
Proportion of sample (6.7%)
Family ownership [20%, 30%)
Firm-years 129
Proportion of sample (7.8%)
Family ownership [30%, 100%)
Firm-years 1,007
Proportion of sample (61.1%)
Family ownership (mean percentage) 35.9% of total shares outstanding

Panel (B)—Board of directors

Total board size
Mean number of directors per board 9.4 directors
Standard deviation [3.3]
Executive directors
Mean number of directors per board 5.6
Proportion of board size (59.6%)
Standard deviation [2.6]
Independent non-executive directors
Mean number of directors per board 2.8
Proportion of board size (29.8%)
Standard deviation [1.5]
CEO duality
Firm-years 850
Proportion of sample (52%)
Audit committee
Firm-years 51
Proportion of sample (3%)
Remuneration committee
Firm-years 11
Proportion of sample (1%)

Panel (C)—Operating performance ratios

Market-to-book ratio
Mean 1.42
Standard deviation [3.65]
Return on assets (ROA)
Mean 1.6%
Standard deviation [24%]
Return on equity (ROE)
Mean 4.2%
Standard deviation [200%]
Long-term debt to total assets
Mean 9.1%
Standard deviation [10.65%]
(continued on next page)
438 Z. Chen et al. / Pacific-Basin Finance Journal 13 (2005) 431–449

Table 1 (continued)
Panel (C)—Operating performance ratios
Dividend payout ratio
Mean 47.2%
Standard deviation [27.9%]
Dividend yield
Mean 4.8%
Standard deviation [7.9%]
The table reports descriptive statistics for a sample of 412 publicly listed Hong Kong firms during 1995–1998
(1648 firm-years). Panel (A) reports data on ownership concentration, panel (B) on board of directors composition
and panel (C) reports operating performance and dividend payout ratios. Family ownership is the ownership
fraction of the controlling family, aggregating the shareholdings of directors with the same surname. CEO duality
is a dummy variable equal to one when the CEO also serves as chairman of the board of directors. Audit
committee and Remuneration committee are dummy variables equal to one if such a committee exists. Market-to-
book ratio is the market value of equity divided by book value of equity. Return on assets is net profit divided by
total assets. Return on equity is net profit divided by shareholders’ equity. Long-term debt-to-assets ratio is long-
term debt divided by total assets. Dividend payout ratio is total dividends divided by net profit. Dividend yield is
dividend per share divided by share price. Throughout the period of the study, the Hong Kong dollar was pegged
to the US$ at the rate of HK$7.8 = US$1.

ratio (0.26), total assets and the presence of an audit committee (0.19), CEO duality and
the number of directors on the board (0.18).

4. Multivariate analysis

This section reports our multivariate analysis of the relationship between ownership
concentration, corporate governance, firm performance, and dividend payouts in Hong
Kong.

4.1. Ownership concentration, corporate governance, and firm performance in Hong

Kong

Table 2 reports results of regressions of return on assets (ROA), return on equity

(ROE) and market-to-book ratio on ownership concentration and corporate governance
variables. Panel (A) reports results for the whole sample. We also recognize that the
impact of ownership structure and corporate governance on performance may be
different for small and large firms. There is less publicly disclosed information about
small firms and less scrutiny by the financial press and analysts, which may provide
controlling shareholders with more opportunities for potential expropriation of minority
shareholders. In addition, the political cost hypothesis (Watts and Zimmerman, 1978)
suggests that larger firms may be less likely to expropriate. Therefore, we also divide
our sample in three equal sub-samples based on stock market capitalization. Panel (B)
reports results for the sub-sample of small market capitalization firms, and panel (C) for
large market capitalization firms (results for medium market capitalization firms are not
reported in order to economize on space).
 Z. Chen et al. / Pacific-Basin Finance Journal 13 (2005) 431–449 439

The results for the whole sample reported in panel (A) show a weak relationship
between ownership concentration and firm performance. There is a negative relationship
between ownership concentration below 10% and the three proxies of firm performance,
but only the coefficients for market-to-book are statistically significant at conventional
levels in columns (7) and (8) (the coefficients of ROA are marginally not significant in
columns (1) and (2)). On the other hand, there is a positive relationship between family
ownership in the 10 to 35% range and all three measures of performance. However, only
the coefficient in the pooled regression of ROA (column (1)) is marginally statistically
significant. Finally, there are mixed results for ownership over 35%, with the coefficient of
ROE in column (6) negative and statistically significant at the 4% level.
To make our results directly comparable with Morck et al. (1988), we have replicated
the regressions reported in Table 2 after using family ownership cut-off levels at 5 and
25% to define our family ownership variables. To economize on space, these
specifications are not reported in the table but will be discussed in the text. Using the
alternative cut-offs, we find no statistically significant relationship between ownership
concentration below 5% and any of the three proxies of firm performance. There is a
uniformly negative relationship between family ownership in the 5 to 25% range and all
three measures of performance. However, the result is statistically significant at the 5%
level only for the market-to-book ratio (in the pooled and industry fixed effects
regressions), and it is marginally not statistically significant at the 10% level for ROA
(in the pooled and industry fixed effects regressions), and for ROE (in the regressions of
industry and firm fixed effects).
Overall, the results on ROA and ROE are in line with the findings by Demsetz and
Lehn (1985), who also find no positive relationship between ownership concentration and
firm operating performance. On the other hand, the results on the market-to-book ratio can
be interpreted as weak evidence consistent with the existence of managerial entrenchment
and poor alignment between managerial incentives and shareholders’ interests for levels of
family ownership up to 10 or 25%. The results in the 5 to 25% family ownership range are
in line with what has been observed for U.S. firms by Morck et al. (1988), and McConnell
and Servaes (1990). In line with Himmelberg et al. (1999), however, the relationship
between ownership concentration and firm performance disappears after estimating firm
fixed effects.
With respect to the corporate governance variables in Table 2, there is a negative
relationship between CEO duality and the market-to-book ratio, which is statistically
significant in all three specifications (at the 1% level in columns (7) and (8), and at the 7%
level in column (9)). It is particularly interesting that the relationship holds even after
controlling for industry and firm fixed effects. As we observe in panels (B) and (C), this
result is driven by large market capitalization firms. There is also a negative relationship
between CEO duality and the market-to-book ratio, which is statistically significant at the
5% level in the pooled and industry fixed effects regressions, when we use the alternative
cut-offs for family ownership of 5 and 25% (not reported in the table). In line with the
results reported in Table 2, the results using the alternative family ownership cut-offs are
driven mainly by large market capitalization firms. Assuming that CEOs who are also
chairmen of the board of directors can exercise more control over their firms, and may be
less likely to be replaced or to have their decisions challenged by the board of directors,
 440
Table 2

Z. Chen et al. / Pacific-Basin Finance Journal 13 (2005) 431–449

Regressions of performance measures
Dependent variable Return on assets (ROA) Return on equity (ROE) Market-to-book ratio
Estimated specification ROA = f (family ownership, CEO duality, board of directors ROE = f (family ownership, CEO duality, board of Market-to-book ratio = F (family ownership, CEO duality,
composition, firm size, leverage, sales growth, year dummy directors composition, firm size, leverage, sales growth, board of directors composition, firm size, leverage, sales
variables) year dummy variables) growth, year dummy variables)

Estimation method Pooled Industry fixed Firm fixed Pooled Industry fixed Firm fixed Pooled Industry Firm fixed
sample (1) effects (2) effects (3) sample (4) effects (5) effects (6) sample (7) fixed effects (8) effects (9)
Panel (A)—Whole sample
Family ownership [0.00, 0.10] 0.480 (0.12) 0.475 (0.13) 0.177 (0.84) 1.246 (0.74) 1.498 (0.69) 6.853 (0.47) 7.192 (0.07)* 7.699 (0.06)* 7.421 (0.55)
Family ownership (0.10, 0.35] 0.259 (0.10)* 0.264 (0.13) 0.308 (0.48) 0.998 (0.58) 1.021 (0.59) 2.092 (0.67) 1.998 (0.25) 2.271 (0.19) 0.175 (0.95)
Family ownership (0.35, 1.00) 0.070 (0.15) 0.064 (0.26) 0.046 (0.79) 0.372 (0.40) 0.530 (0.32) 3.017 (0.04)** 0.499 (0.73) 0.545( 0.72) 3.704 (0.14)
CEO duality 0.000 (0.96) 0.002 (0.87) 0.001 (0.98) 0.102 (0.30) 0.128 (0.26) 0.529 (0.18) 0.499 (0.01)*** 0.508 (0.01)*** 1.084 (0.07)*
Ln (number of directors) 0.031 (0.23) 0.027 (0.29) 0.036 (0.69) 0.547 (0.15) 0.535 (0.19) 1.555 (0.12) 0.558 (0.02)** 0.582 (0.01)*** 1.203 (0.25)
% Independent non-executive 0.036 (0.61) 0.035 (0.62) 0.015 (0.92) 0.965 (0.30) 1.084 (0.32) 2.458 (0.20) 0.148 (0.82) 0.046 (0.95) 3.820 (0.04)**
directors
Outsider-dominated board 0.011 (0.70) 0.001 (0.96) 0.015 (0.73) 0.080 (0.56) 0.156 (0.26) 0.060 (0.87) 0.092 (0.77) 0.074 (0.83) 0.591 (0.32)
Audit committee 0.024 (0.12) 0.023 (0.14) 0.063 (0.01)*** 0.076 (0.48) 0.142 (0.29) 0.131 (0.60) 0.121 (0.66) 0.120 (0.62) 0.080 (0.73)
Ln (total assets) 0.032 (0.00)*** 0.038 (0.00)*** 0.186 (0.00)*** 0.008 (0.88) 0.050 (0.31) 0.579 (0.14) 0.184 (0.05)** 0.203 (0.07)* 1.190 (0.05)**
Debt-to-assets ratio 0.185 (0.00)*** 0.248 (0.00)*** 0.143 (0.28) 0.685 (0.04)** 0.742 (0.05)** 1.710 (0.18) 0.131 (0.83) 0.176 (0.80) 1.646 (0.43)
Sales growth 0.004 (0.11) 0.006 (0.07)* 0.001 (0.81) 0.009 (0.41) 0.018 (0.23) 0.001 (0.98) 0.004 (0.90) 0.012 (0.75) 0.040 (0.24)
No of observations (firm-years) 1647 1647 1647 1647 1647 1647 1647 1647 1647
Adjusted R 2 0.08 0.09 0.17 0.00 0.01 0.05 0.01 0.01 0.36

Panel (B)—Small market capitalization firms

Family ownership [0.00, 0.10] 0.584 (0.32) 0.614 (0.30) 0.469 (0.82) 1.180 (0.90) 4.281 (0.61) 8.378 (0.66) 4.377 (0.52) 2.469 (0.60) 11.616 (0.46)
Family ownership (0.10, 0.35] 0.331 (0.30) 0.395 (0.29) 0.738 (0.40) 1.443 (0.76) 1.859 (0.66) 0.921 (0.92) 2.300 (0.45) 1.795 (0.43) 4.140 (0.58)
Family ownership (0.35, 1.00) 0.128 (0.22) 0.077 (0.51) 0.258 (0.50) 0.178 (0.85) 1.011 (0.24) 3.574 (0.23) 0.312 (0.62) 0.253 (0.67) 2.668 (0.30)
CEO duality 0.003 (0.91) 0.002 (0.95) 0.169 (0.26) 0.322 (0.27) 0.364 (0.26) 1.550 (0.28) 0.203 (0.20) 0.228 (0.14) 1.264 (0.31)
Ln (number of directors) 0.055 (0.44) 0.013 (0.88) 0.040 (0.84) 1.950 (0.13) 1.614 (0.25) 1.450 (0.40) 0.214 (0.60) 0.149 (0.45) 0.453 (0.42)
% Independent non-executive 0.107 (0.58) 0.073 (0.71) 0.514 (0.22) 3.415 (0.29) 4.298 (0.26) 4.700 (0.20) 0.296 (0.59) 0.688 (0.08)* 1.677 (0.14)
directors
Outsider-dominated board 0.017 (0.83) 0.022 (0.76) 0.018 (0.90) 0.593 (0.36) 0.518 (0.37) 1.271 (0.26) 0.002 (0.99) 0.083 (0.64) 0.030 (0.95)
Audit committee 0.063 (0.13) 0.025 (0.65) 0.027 (0.70) 0.117 (0.72) 0.082 (0.82) 0.076 (0.83) 0.068 (0.59) 0.073 (0.58) 0.120 (0.58)
Ln (total assets) 0.049 (0.04)** 0.051 (0.06)* 0.277 (0.03)** 0.009 (0.97) 0.203 (0.20) 0.660 (0.34) 0.330 (0.06)* 0.458 (0.00)*** 0.357 (0.39)
Debt-to-assets ratio 0.091 (0.53) 0.152 (0.31) 0.025 (0.94) 0.805 (0.29) 1.907 (0.09)* 8.297 (0.08)* 1.282 (0.00)*** 2.061 (0.01)*** 4.304 (0.27)
Sales growth 0.003 (0.33) 0.004 (0.37) 0.003 (0.67) 0.010 (0.51) 0.040 (0.24) 0.040 (0.43) 0.001 (0.93) 0.009 (0.57) 0.004 (0.93)
No of observations (firm-years) 548 548 548 548 548 548 548 548 548
Adjusted R 2 0.04 0.03 0.30 0.01 0.04 0.40 0.03 0.17 0.10

Panel (C)—Large market capitalization firms

Family ownership [0.00, 0.10] 0.008 (0.97) 0.041 (0.86) 0.099 (0.75) 0.644 (0.34) 0.578 (0.36) 2.003 (0.24) 6.582 (0.28) 7.875 (0.22) 39.123 (0.15)
Family ownership (0.10, 0.35] 0.007 (0.94) 0.033 (0.70) 0.024 (0.79) 0.393 (0.19) 0.385 (0.19) 0.467 (0.27) 3.079 (0.46) 5.572 (0.12) 3.804 (0.55)
Family ownership (0.35, 1.00) 0.070 (0.13) 0.063 (0.22) 0.029 (0.53) 0.027 (0.73) 0.040 (0.64) 0.195 (0.11) 0.669 (0.87) 2.292 (0.44) 10.287 (0.23)

Z. Chen et al. / Pacific-Basin Finance Journal 13 (2005) 431–449

CEO duality 0.004 (0.67) 0.005 (0.59) 0.008 (0.48) 0.022 (0.31) 0.016 (0.43) 0.023 (0.33) 1.358 (0.02)** 0.856 (0.02)** 0.299 (0.67)
Ln (number of directors) 0.003 (0.81) 0.014 (0.48) 0.051 (0.03)** 0.030 (0.39) 0.037 (0.43) 0.138 (0.05)** 1.129 (0.00)*** 1.422 (0.03)** 4.239 (0.30)
% Independent non-executive 0.028 (0.49) 0.003 (0.96) 0.074 (0.45) 0.016 (0.87) 0.105 (0.46) 0.410 (0.15) 1.430 (0.32) 1.195 (0.55) 3.799 (0.15)
directors
Outsider-dominated board 0.006 (0.74) 0.009 (0.67) 0.014 (0.60) 0.005 (0.91) 0.017 (0.73) 0.101 (0.23) 0.999 (0.14) 0.533 (0.52) 0.272 (0.79)
Audit committee 0.014 (0.41) 0.003 (0.83) 0.005 (0.78) 0.035 (0.21) 0.004 (0.87) 0.024 (0.55) 0.320 (0.60) 0.175 (0.77) 1.153 (0.23)
Ln (total assets) 0.003 (0.47) 0.005 (0.50) 0.015 (0.63) 0.002 (0.86) 0.003 (0.87) 0.074 (0.30) 0.985 (0.00)*** 1.203 (0.00)*** 3.006 (0.09)*
Debt-to-assets ratio 0.197 (0.01)*** 0.298 (0.01)*** 0.353 (0.06)* 0.104 (0.63) 0.059 (0.84) 0.307 (0.54) 3.574 (0.00)*** 2.807 (0.05)** 2.298 (0.35)
Sales growth 0.004 (0.35) 0.003 (0.50) 0.001 (0.84) 0.009 (0.31) 0.006 (0.50) 0.002 (0.73) 0.163 (0.50) 0.088 (0.65) 0.120 (0.58)
No of observations (firm-years) 549 549 549 549 549 549 549 549 549
Adjusted R 2 0.07 0.14 0.46 0.00 0.09 0.42 0.07 0.20 0.62
The table reports results of regressions of firm performance variables on family ownership concentration, after controlling for corporate governance variables. Panel (A) reports results for the whole sample. Sample firms are also
ranked based on stock market capitalization and divided in three equal sub-samples. Panel (B) reports results for the sub-sample of firms with the lowest stock market capitalization, and panel (C) reports results for the sub-sample
with the largest market capitalization. Return on assets is net profit divided by total assets. Return on equity is net profit divided by shareholdersT equity. Market-to-book ratio is the market value of equity divided by book value
of equity. Family ownership is the ownership fraction of the controlling family, aggregating the shareholdings of directors with the same surname. The Family ownership variables are constructed as follows. Assuming that the
ownership fraction of the controlling family is m, where m = 0, . . ., 1, then
Family Ownership ½0; 0:10 ¼ m ðif m b 0:10Þ; 0:10 ðif m z 0:10Þ
Family Ownership ð0:10; 0:35 ¼ 0 ðif m b 0:10Þ; m  0:10 ðif 0:10 V m b 0:35Þ; 0:25 ðif m z 0:35Þ
Family Ownership ð0:35; 1:00Þ ¼ 0 ðif m b 0:35Þ; m  0:35 ðif m z 0:35Þ
CEO duality is a dummy variable equal to one when the CEO also serves as chairman of the board of directors. Ln (number of directors) is the natural logarithm of the number of directors on the board. % Independent non-
executive directors is the fraction of independent non-executive directors on the board. Outsider-dominated board is a dummy variable equal to one if the fraction of independent non-executive directors on the board exceeds
50%. Audit committee is a dummy variable equal to one if such a committee exists. Debt-to-assets ratio is long-term debt divided by total assets. Sales growth is annual growth in sales. Ln (total assets) is the natural logarithm of
the firm’s deflated total assets. The table reports three different types of regressions: using all firm-year observations pooled (pooled sample), estimating fixed effects at the industry level (industry fixed effects), and estimating
fixed effects at the firm level (firm fixed effects). All regressions include an intercept and year dummy variables. Intercepts, year dummy and fixed effects coefficients are not reported. Adjusted R 2 reflects the inclusion of fixed
effects. Reported p-values in parentheses are based on White (1980) heteroskedasticity consistent standard errors.
***, **, and * denote significance at the 1, 5, and 10 percent level respectively in two-tailed tests.

441
442 Z. Chen et al. / Pacific-Basin Finance Journal 13 (2005) 431–449

this result suggests that managerial entrenchment in the form of CEO duality is associated
with lower firm value.
On the other hand, there is a positive relationship between the presence of an audit
committee and ROA (the coefficients in columns (1) and (2) are marginally not significant
at conventional levels but the coefficient in column (3) is significant at the 1% level).
When we use the alternative family ownership cut-offs of 5 and 25%, the relationship
between the presence of an audit committee and ROA is statistically significant in all
specifications (not reported in the table). Therefore, it appears that firms with an audit
committee on their boards have better operating performance. The interpretation of these
results is offered under the caveat that audit committees were not mandatory during the
period under study. Therefore, it is not clear whether it is the monitoring role of the audit
committee that drives this result or whether there is self-selection bias at work here (i.e.,
that bbetterQ firms choose to appoint audit committees on their boards).
Many studies (see for example, Yermack, 1996; Hermalin and Weisbach, 1988)
recognize that the size and composition of the board of directors are important
determinants of the effectiveness of board monitoring over company management.
However, the remaining corporate governance variables have little explanatory power in
our analysis. There is no consistent relationship between the proportion of independent
non-executive directors on the board and firm performance, even when independent
directors constitute the majority on the board. Similar results are obtained when we use the
alternative family ownership cut-offs (not reported in the table). This result casts doubt on
the effectiveness of independent non-executive directors as a monitoring mechanism in
Hong Kong, and may be due to the difficulty of recruiting truly independent board
members in a small corporate market.
Overall, the evidence reported in the current section shows that (i) there is no
statistically significant positive relationship between family ownership concentration and
firm operating performance or market-to-book at any level of ownership, (ii) there is
evidence of poor alignment between managerial incentives and shareholder interests at low
levels of family ownership, and evidence of managerial entrenchment at higher levels of
family ownership, (iii) there is further evidence of managerial entrenchment in companies
with dual CEOs, (iv) the presence of an audit committee is positively related to firm
operating performance but the relationship may be the result of self-selection bias, and (v)
there is no significant relationship between the presence of independent non-executive
directors and operating performance or firm value. Overall, therefore, corporate
governance mechanisms appear weak in Hong Kong companies.

4.2. Dividend payout policies in Hong Kong

The economic interpretation of the coefficients of dividend payout regressions in

countries where outside investors may be expropriated is subject to controversy in the
literature for two reasons. First, there is disagreement in the academic literature on the type
of equilibrium relationship that is being observed. On the one hand, Faccio et al. (2001)
assume that the observed phenomena represent an equilibrium where expropriation does
not take place, and therefore the empirical relationships reflect the resolution of agency
conflicts. This implies that investors anticipate potential expropriation and obtain higher
 Z. Chen et al. / Pacific-Basin Finance Journal 13 (2005) 431–449 443

payouts from firms with the highest agency conflicts, such as Western European and East
Asian firms tightly affiliated to business groups and, within groups, firms with wider
divergence of control and cash flow rights. On the other hand, La Porta et al. (2000)
assume that the observed empirical relationships reflect an equilibrium where expropria-
tion may be present. They show that firms in countries with poor legal protection of
minority shareholders make lower dividend payouts because investors have no legal
avenues to force higher payouts from firms.
Second, there is also disagreement between the academic literature and market
participants on the implications of dividend payouts in firms with concentrated ownership.
Although the academic literature tends to associate large payouts with the resolution of
agency conflicts, the financial press has viewed large dividend payouts by closely held
companies in countries with weak legal protection of outside investors as potentially sub-
optimal behaviour, that is, as extraction of resources from the company by the controlling
shareholders. For example, when the controlling shareholders of the Russian oil firm
Sibneft sold and bought back their stake in the firm in 2001, the deal was criticized on
corporate governance grounds because, among other things, bthe shareholders bought
back their stake just in time to collect a juicy dividend of around $150 m—far more than
the paltry $26 m the company says it made in the share shuffle.Q1
Consequently, the economic interpretation of some of the coefficients of regressions of
dividend payouts by Hong Kong firms may be subject to controversy. Table 3 reports the
results of regressions of the dividend payout ratio and of dividend yield on family
ownership. In this regression model, we include the market-to-book ratio and ROA as
control variables for firm performance. The regressions of the dividend payout ratio use
only observations with positive net profit (1444 firm-years). The regressions of the
dividend yield use all observations.
The results for the whole sample in panel (A) (and for the sub-sample of large market
capitalization firms in panel (C)) show no statistically significant relationship between
family ownership and dividend payouts. Faccio et al. (2001), who perform regressions of a
variety of dividend payout ratios on shareholders’ ownership-to-control-rights ratio, also
fail to find any statistically significant results for Hong Kong. In contrast, the results of the
analysis of small market capitalization firms in panel (B) show a negative relationship
between dividend yield and family ownership up to 10% of the company’s stock and a
positive relationship for ownership in the 10 to 35% range (all coefficients are statistically
significant at the 10% level in the pooled and industry fixed effects regressions in columns
(4) and (5)).
The positive relationship between dividend payouts and controlling family ownership
that we observe for small firms is consistent with the conjecture that investors demand
higher dividend payouts from the companies with potentially the highest risk of
expropriation of minority shareholders. However, our evidence is also consistent with
an alternative conjecture, namely that the larger payouts represent extraction of company
resources by the controlling family shareholders who may be receiving large amounts of

1
bBait switch, swallow, gulp: Russia’s new business class still has some bad old habitsQ, The Economist, 3
November 2001, p. 75.
 444
Table 3
Regressions of dividend payout ratios
Dependent variable Dividend payout ratio Dividend yield
Estimated specification Dividend payout ratio=f (family ownership, CEO duality, Dividend yield=f (family ownership, CEO duality, board of
board of directors composition, firm size, leverage, sales directors composition, firm size, leverage, sales growth,
growth, market-to-book ratio, ROA, year dummy variables) market-to-book ratio, ROA, year dummy variables)

Z. Chen et al. / Pacific-Basin Finance Journal 13 (2005) 431–449

Estimation method Pooled sample Industry fixed Firm fixed Pooled sample Industry fixed Firm fixed
(1) effects (2) effects (3) (4) effects (5) effects (6)
Panel (A)—Whole sample
Family ownership [0.00, 0.10] 1.267 (0.62) 0.275 (0.90) 3.938 (0.40) 0.035 (0.63) 0.010 (0.89) 0.150 (0.48)
Family ownership (0.10, 0.350] 0.386 (0.72) 0.570 (0.59) 0.340 (0.79) 0.022 (0.48) 0.011 (0.74) 0.088 (0.27)
Family ownership (0.35, 1.00) 0.391 (0.67) 0.099 (0.92) 0.077 (0.94) 0.010 (0.53) 0.008 (0.63) 0.007 (0.77)
CEO duality 0.042 (0.72) 0.063 (0.58) 0.022 (0.81) 0.001 (0.86) 0.002 (0.53) 0.005 (0.38)
Ln (number of directors) 0.078 (0.80) 0.110 (0.71) 0.107 (0.88) 0.009 (0.20) 0.007 (0.39) 0.016 (0.54)
% Independent non-executive directors 0.725 (0.47) 0.857 (0.42) 0.750 (0.62) 0.026 (0.17) 0.025 (0.21) 0.001 (0.98)
Outsider dominated board 0.436 (0.30) 0.547 (0.22) 0.387 (0.63) 0.015 (0.06)* 0.014 (0.06)* 0.004 (0.85)
Audit committee 0.298 (0.03)** 0.204 (0.13) 0.192 (0.35) 0.014 (0.01)*** 0.012 (0.04)** 0.021 (0.07)*
Ln (total assets) 0.040 (0.39) 0.025 (0.60) 0.162 (0.37) 0.003 (0.11) 0.003 (0.32) 0.034 (0.08)*
Debt-to-assets ratio 0.514 (0.39) 0.366 (0.59) 2.916 (0.04)** 0.029 (0.47) 0.026 (0.55) 0.119 (0.14)
Sales growth 0.082 (0.02)** 0.082 (0.05)** 0.075 (0.02)** 0.003 (0.00)*** 0.003 (0.00)*** 0.001 (0.17)
Market-to-book ratio 0.077 (0.43) 0.086 (0.39) 0.010 (0.48) 0.001 (0.06)* 0.001 (0.10)* 0.001 (0.13)
Return on assets 0.177 (0.18) 0.107 (0.44) 0.000 (1.00) 0.039 (0.02)** 0.037 (0.03)** 0.026 (0.09)*
No of observations (firm-years) 1444 1444 1444 1645 1645 1645
Adjusted R 2 0.00 0.01 0.17 0.04 0.03 0.21

Panel (B)—Small market capitalization firms

Family ownership [0.00, 0.10] 0.085 (0.97) 0.521 (0.83) 4.093 (0.78) 0.197 (0.07)* 0.229 (0.04)** 0.097 (0.57)
Family ownership (0.10, 0.35] 0.227 (0.75) 0.998 (0.36) 3.130 (0.28) 0.105 (0.01)*** 0.100 (0.02)** 0.076 (0.24)
Family ownership (0.35, 1.00) 2.087 (0.14) 1.845 (0.26) 0.439 (0.88) 0.007 (0.77) 0.014 (0.59) 0.005 (0.92)
CEO duality 0.021 (0.89) 0.042 (0.80) 0.359 (0.28) 0.002 (0.60) 0.002 (0.73) 0.024 (0.19)
Ln (number of directors) 0.677 (0.47) 0.703 (0.44) 0.984 (0.57) 0.005 (0.67) 0.001 (0.97) 0.010 (0.57)
% Independent non-executive directors 2.505 (0.07)* 2.045 (0.06)* 6.338 (0.10)* 0.035 (0.30) 0.041 (0.27) 0.101 (0.01)***
Outsider dominated board 0.489 (0.43) 0.381 (0.51) 1.222 (0.28) 0.014 (0.45) 0.011 (0.57) 0.005 (0.80)
Audit committee 0.766 (0.06)* 0.549 (0.13) 0.397 (0.421) 0.001 (0.94) 0.002 (0.88) 0.008 (0.65)
Ln (total assets) 0.011 (0.92) 0.016 (0.90) 0.132 (0.756) 0.004 (0.25) 0.004 (0.37) 0.001 (0.95)
Debt-to-assets ratio 0.228 (0.74) 0.212 (0.71) 1.740 (0.23) 0.014 (0.51) 0.015 (0.52) 0.016 (0.78)
Sales growth 0.039 (0.08)* 0.041 (0.24) 0.029 (0.67) 0.002 (0.01)*** 0.002 (0.00)*** 0.002 (0.09)*
Market-to-book ratio 0.022 (0.51) 0.018 (0.53) 0.036 (0.36) 0.002 (0.16) 0.002 (0.05)** 0.001 (0.42)
Return on assets 0.255 (0.07)* 0.261 (0.06)* 0.086 (0.65) 0.015 (0.12) 0.013 (0.19) 0.008 (0.39)
No of observations (firm-years) 445 445 445 548 548 548
Adjusted R 2 0.00 0.02 0.17 0.04 0.05 0.28

Panel (C)—Large market capitalization firms

Family ownership [0.00, 0.10] 0.132 (0.95) 0.118 (0.96) 9.137 (0.44) 0.055 (0.51) 0.109 (0.15) 0.135 (0.27)
Family ownership (0.10, 0.35] 1.323 (0.56) 1.575 (0.50) 1.943 (0.55) 0.004 (0.92) 0.022 (0.58) 0.055 (0.22)
Family ownership (0.35, 1.00) 1.299 (0.54) 1.585 (0.49) 0.101 (0.92) 0.020 (0.41) 0.025 (0.35) 0.011 (0.65)

Z. Chen et al. / Pacific-Basin Finance Journal 13 (2005) 431–449

CEO duality 0.093 (0.46) 0.079 (0.65) 0.014 (0.93) 0.002 (0.66) 0.000 (1.00) 0.002 (0.69)
Ln (number of directors) 0.352 (0.18) 0.104 (0.75) 2.995 (0.26) 0.011 (0.07)* 0.007 (0.31) 0.021 (0.32)
% Independent non-executive directors 1.104 (0.55) 1.465 (0.51) 0.323 (0.91) 0.065 (0.00)*** 0.064 (0.00)*** 0.026 (0.55)
Outsider dominated board 0.476 (0.54) 0.582 (0.49) 1.306 (0.61) 0.031 (0.00)*** 0.031 (0.00)*** 0.027 (0.06)*
Audit committee 0.222 (0.24) 0.253 (0.28) 0.035 (0.92) 0.013 (0.01)*** 0.006 (0.22) 0.012 (0.29)
Ln (total assets) 0.033 (0.64) 0.218 (0.23) 0.009 (0.98) 0.003 (0.06)* 0.001 (0.79) 0.015 (0.27)
Debt-to-assets ratio 1.539 (0.23) 2.780 (0.19) 3.462 (0.30) 0.001 (0.96) 0.003 (0.91) 0.037 (0.43)
Sales growth 0.221 (0.26) 0.317 (0.20) 0.183 (0.18) 0.005 (0.18) 0.006 (0.14) 0.002 (0.47)
Market-to-book ratio 0.124 (0.42) 0.177 (0.31) 0.027 (0.45) 0.000 (0.73) 0.000 (0.84) 0.001 (0.32)
Return on assets 0.793 (0.43) 0.673 (0.52) 1.954 (0.37) 0.172 (0.07)* 0.196 (0.06)* 0.221 (0.16)
No of observations (firm-years) 518 518 518 549 549 549
Adjusted R 2 0.02 0.01 0.32 0.14 0.17 0.37
The table reports results of regressions of payout policy proxies on controlling family ownership, after controlling for firm performance and corporate governance variables. Panel (A) reports
results for the whole sample. Sample firms are also ranked based on stock market capitalization and divided in three equal sub-samples. Panel (B) reports results for the sub-sample of firms with
the lowest stock market capitalization, and panel (C) reports results for the sub-sample with the largest market capitalization. Dividend payout ratio is total dividends divided by net profit for
firms with positive net profit. Dividend yield is dividend per share divided by share price. Family ownership is the ownership fraction of the controlling family, aggregating the shareholdings of
directors with the same surname. The Family ownership variables are constructed as follows. Assuming that the ownership fraction held by the controlling family is m, where m = 0, . . ., 1, then

Family Ownership ½0; 0:10 ¼ m ðif m b 0:10Þ; 0:10 ðif m z 0:10Þ

Family Ownership ð0:10; 0:35 ¼ 0 ðif m b 0:10Þ; m  0:10 ðif 0:10 V m b 0:35Þ; 0:25 ðif m z 0:35Þ
Family Ownership ð0:35; 1:00Þ ¼ 0 ðif m b 0:35Þ; m  0:35 ðif m z 0:35Þ

CEO duality is a dummy variable equal to one when the CEO also serves as chairman of the board of directors. Ln (number of directors) is the natural logarithm of the number of directors on the
board. % Independent non-executive directors is the fraction of independent non-executive directors on the board. Outsider-dominated board is a dummy variable equal to one if the fraction of
independent non-executive directors on the board exceeds 50%. Audit committee is a dummy variable equal to one if such a committee exists. Market-to-book ratio is the market value of equity
divided by book value of equity. Return on assets is net profit divided by total assets. Debt-to-assets ratio is long-term debt divided by total assets. Sales growth is annual growth in sales. Ln
(total assets) is the natural logarithm of the firm’s deflated total assets. The table reports three different types of regressions: using all firm-year observations pooled (pooled sample), estimating
fixed effects at the industry level (industry fixed effects), and estimating fixed effects at the firm level (firm fixed effects). All regressions include an intercept and year dummy variables. Intercepts,
year dummy and fixed effects coefficients are not reported. Adjusted R 2 reflects the inclusion of fixed effects. Reported p-values in parentheses are based on White (1980) heteroskedasticity
consistent standard errors.

445
***, **, and * denote significance at the 1, 5, and 10 percent level respectively in two-tailed tests.
446 Z. Chen et al. / Pacific-Basin Finance Journal 13 (2005) 431–449

dividend income. To get a measure of the level of dividend income received by controlling
shareholders, we collect data on executive compensation and dividend income for the
firms in our sample (not reported in tables). We observe that the average CEO who was not
simultaneously chairman of the board received annually almost HK$13 million as
dividend income (in constant 1994 Hong Kong dollars) and only HK$3.7 million in the
form of cash emoluments. These figures represent a ratio of dividend income to cash
emoluments of almost 4:1 (4 dollars in dividends for every dollar received as cash
emoluments). Similarly, the average dual CEO received HK$49 million in dividend
income and HK$3.5 million in cash salary (ratio of dividend income to cash emoluments
of 14:1), and the average chairman who was not CEO received dividend income of HK$42
million and cash salary of HK$2.4 million (ratio of 18:1). Therefore, top managers with
significant shareholdings may care more about their dividend income than their cash
salary.
The impact of the different corporate governance variables on dividend payouts is also
ambiguous. The presence of an audit committee has a significant negative impact on
dividend yield in columns (4)–(6), and on dividend payout in column (1). The table shows
a negative coefficient for the audit committee dummy variable in all specifications. In the
regressions of dividend yield, the coefficients are all statistically significant at the 10%
level even after estimating firm fixed effects. This implies that firms with an audit
committee have a lower dividend yield. However, this effect is weak in the analysis of
small and large market capitalization firms (in panels (B) and (C)).
The remaining corporate governance variables are not significant in explaining
dividend policy, with the exception of the dummy variable indicating that the majority
of board members are independent non-executive directors, whose coefficient is positive
and statistically significant at the 10% level in the pooled and industry fixed effects
regressions of dividend yield (columns (4) and (5)). This suggests that firms with a
majority of board members as independent non-executive directors have a higher dividend
yield. This result is driven by large market capitalization firms in panel (C). However,
there is also a positive relationship between the proportion of independent non-executive
directors on the board and the dividend payout ratio or dividend yield for small market
capitalization firms in panel (B).
Interestingly, there are mixed results on the relationship between dividend payouts and
firm performance. In panel (A), there is a negative relationship between market-to-book
and dividend yield, statistically significant at the 10% level in the pooled and industry
fixed effects regressions in columns (4) and (5) (and at the 13% level in the firm fixed
effects regression in column (6)). In contrast, there is a positive relationship between ROA
and dividend yield, which is statistically significant at the 10% level in all three
regressions. This result is mainly driven by large market capitalization firms in panel (C).
In panel (B), there is no relationship between dividend payouts and firm performance for
small market capitalization firms. Therefore, in line with expectations, it appears that in
large firms dividend policy may be more sensitive to performance compared to small
firms.2

2
The results when using the alternative cut-off ownership levels of 5 and 25% (not reported in the table) are
similar to the ones reported in Table 3.
 Z. Chen et al. / Pacific-Basin Finance Journal 13 (2005) 431–449 447

Overall, our main results in this section show (i) that there is a negative relationship
between dividend yield and family ownership up to 10% of the company’s stock and a
positive relationship for ownership in the 10 to 35% range for small market capitalization
firms, and (ii) that dividend policy is not sensitive to firm performance in small firms but is
more sensitive to performance in large firms. These results suggest that families with
controlling shareholdings, especially in small firms that are subject to less scrutiny by
investors, may be using dividend payout policy as a means to extract resources out of the
firms they control. Alternatively, our results are also consistent with the conjecture that
outside investors anticipate potential expropriation by the controlling shareholders and
demand higher payouts from firms with potentially the largest agency conflicts.
Unfortunately, it is difficult to distinguish empirically between the two explanations.

5. Conclusions

Our article contributes to the literature on the relationship between ownership

concentration and firm performance or value (e.g., Demsetz and Lehn, 1985; Morck et
al., 1988; McConnell and Servaes, 1990; Hermalin and Weisbach, 1991; Himmelberg et
al., 1999), and to a growing body of literature on international corporate governance (e.g.,
La Porta et al., 1999, 2000, 2002; Claessens et al., 2000; Faccio et al., 2001).
We analyze a sample of 412 publicly listed Hong Kong firms during 1995–1998 and
try to answer three questions. First, and most importantly, does concentrated family
ownership affect firm operating performance and value? Our empirical analysis does not
show a positive relationship between family ownership and return on assets, return on
equity, or the market-to-book ratio. Therefore, our results for Hong Kong are in line with
both Demsetz and Lehn (1985) and Himmelberg et al. (1999), who show that
concentrated ownership is not associated with better operating performance or higher
firm valuation. In fact, the only statistically significant relationship is negative, suggesting
the presence of agency costs at low and moderate levels of family ownership. In addition,
we also find a negative relationship between CEO duality and performance (where CEO
duality is much more likely in family-controlled firms), which is evidence consistent with
managerial entrenchment in companies that combine the positions of CEO and chairman
of the board.
Second, does concentrated family ownership affect dividend policy? We find little
relationship between family ownership and dividend policy. Only for small firms there is a
significant negative relationship between dividend payouts and family ownership of up to
10% of the company’s stock and a positive relationship for family ownership between 10
and 35%. In addition, dividend payout policy is not sensitive to firm performance in small
firms. These findings may suggest that dividend payouts are potentially used by
controlling shareholders in smaller Hong Kong companies as a way of extracting
resources out of the firms they control. We report evidence which shows that owners–
managers may care more about their dividend income compared to their cash salary, since
on average their cash salary is much lower than dividend income. However, it may also be
the case that investors anticipate the potential expropriation by the controlling shareholders
and demand higher payouts from firms with potentially the largest agency conflicts.
448 Z. Chen et al. / Pacific-Basin Finance Journal 13 (2005) 431–449

And third, what is the impact of corporate governance on performance, value, and
dividend payouts in family-controlled firms? Our evidence shows that the composition of
the board of directors (proportion of independent non-executive directors, outsider-
dominated board) has little impact on firm performance and dividend policy, particularly
in small market capitalization firms. On the other hand, the presence of an audit committee
is positively related to firm operating performance. However, this result must be
interpreted with caution because only a small fraction of firms in our sample have audit
committees on their boards (which were not mandatory during the period under study) and
the relationship may be the result of self-selection bias.
One policy implication that can be derived from this analysis is that corporate
governance in Hong Kong needs to be strengthened. In addition, more effort is needed in
order to ensure the true independence of non-executive directors and that they are able to
perform an adequate monitoring function. The recent introduction of institutional
investors, in the form of large private funds managing pension contributions under the
Mandatory Provident Fund, could potentially help in raising the standards of corporate
governance. Our findings have relevance for other East Asian countries too, which are
characterized by even lower standards of corporate governance and investor protection
compared to Hong Kong.

Acknowledgements

We would like to thank Ghon Rhee (the editor), an anonymous referee, and participants
at the 2001 annual meeting of the Multinational Finance Society in Garda, and the 2001
Australasian Finance and Banking Conference in Sydney for their helpful comments on
earlier drafts. We also thank Helen Tse and Yinqing Zhao who provided research
assistance.

References

Cadbury, A., 1992. Report of the Committee on the Financial Aspects of Corporate Governance. Gee and Co.
Ltd., London, United Kingdom.
Chen, C.R., Steiner, T.L., 1999. Managerial ownership and agency conflicts: A nonlinear simultaneous equation
analysis of managerial ownership, risk taking, debt policy, and dividend policy. The Financial Review 34,
119 – 136.
Chen, C.R., Guo, W., Mande, V., 2003. Managerial ownership and firm valuation: Evidence from Japanese firms.
Pacific-Basin Finance Journal 11, 267 – 283.
Cho, M.H., 1998. Ownership structure, investment, and the corporate value: An empirical analysis. Journal of
Financial Economics 47, 103 – 121.
Claessens, S., Djankov, S., Lang, L.H.P., 2000. The separation of ownership and control in East Asian
corporations. Journal of Financial Economics 58, 81 – 112.
Demsetz, H., Lehn, K., 1985. The structure of corporate ownership: Causes and consequences. Journal of Political
Economy 93, 1155 – 1177.
Faccio, M., Lang, L.H.P., Young, L., 2001. Dividends and expropriation. American Economic Review 91,
54 – 78.
Hermalin, B., Weisbach, M., 1988. The determinants of board composition. Rand Journal of Economics 19,
589 – 606.
 Z. Chen et al. / Pacific-Basin Finance Journal 13 (2005) 431–449 449

Hermalin, B.E., Weisbach, M.S., 1991. The effects of board composition and direct incentives on firm
performance. Financial Management, 101 – 112 (Winter).
Himmelberg, C.P., Hubbard, R.G., Palia, D., 1999. Understanding the determinants of managerial ownership and
the link between ownership and performance. Journal of Financial Economics 53, 353 – 384.
Hiraki, T., Inoue, H., Ito, A., Kuroki, F., Masuda, H., 2003. Corporate governance and firm value in Japan:
Evidence from 1985 to 1998. Pacific-Basin Finance Journal 11, 239 – 265.
La Porta, R., Lopez-de-Silanes, F., Shleifer, A., Vishny, R.W., 1998. Law and finance. Journal of Political
Economy 106, 1113 – 1155.
La Porta, R., Lopez-de-Silanes, F., Shleifer, A., 1999. Corporate ownership around the world. Journal of Finance
54, 471 – 517.
La Porta, R., Lopez-de-Silanes, F., Shleifer, A., Vishny, R.W., 2000. Agency problems and dividend policies
around the world. Journal of Finance 55, 1 – 33.
La Porta, R., Lopez-de-Silanes, F., Shleifer, A., Vishny, R.W., 2002. Investor protection and corporate valuation.
Journal of Finance 57, 1147 – 1170.
McConnell, J., Servaes, H., 1990. Additional evidence on equity ownership and corporate value. Journal of
Financial Economics 27, 595 – 612.
Morck, R., Shleifer, A., Vishny, R.W., 1988. Management ownership and market valuation: An empirical
analysis. Journal of Financial Economics 20, 293 – 315.
Morck, R., Nakamura, M., Shivdasani, A., 2000. Banks, ownership structure, and firm value in Japan. Journal of
Business 73, 539 – 567.
Watts, R.L., Zimmerman, J.L., 1978. Towards a positive theory of the determination of accounting standards.
Accounting Review 53, 112 – 134.
White, H., 1980. A heteroskedasticity-consistent covariance matrix estimator and a direct test for
heteroskedasticity. Econometrica 48, 817 – 838.
Yermack, D., 1996. Higher market valuation of companies with a small board of directors. Journal of Financial
Economics 40, 185 – 211.
Zhou, X., 2001. Understanding the determinants of managerial ownership and the link between ownership and
performance: Comment. Journal of Financial Economics 62, 559 – 571.