190 /304
投資機會和投資者保護機制對控股股東派息的影響
190
the year.
Leverage
is measured by total liabilities divided by total assets at the end of the
year. Following La Porta et al. (2000), we also include three country-level variables to
control for cross-country differences: difference in economies (
Ln(GNP)
), difference in
legal reserves (
LRes
), and difference in dividend tax advantages (
Dta
).
We estimate the regression coefficients with each of the four dividends payout rates.
Table 4 displays the regression results for each of the four industry-adjusted dividend
ratios (
IADvd
divided by Sales, Cash flow, the Market value of equity, and Earnings).
Each of the three legal institutional measures is used in Panels A through C. The
coefficients of
Diverge
are positively and significantly associated with
IADvd/CFO
and
IADvd/MV
in Panel B, but none of them is positive and significant in Panels A or C. The
coefficients of
G*Diverge
, which is the interaction of
Diverge
and investment
opportunities (G), are positively and significantly associated with most of the dividend
payout ratios (Panels A through C), which suggests that divergent firms pay higher
dividends even though they have good investment opportunities. This is consistent with
the work of Faccio et al. (2001), who argue that divergent firms pay more dividends to
mask expropriation, irrespective of their investment opportunities. The coefficients of
Legal (
Right
and
Disclosure
) are positively and significantly associated with each of the
four dividend payout ratios in Panel A (Panels B and C), which indicates that firms that
operate in countries with better investor protection pay higher dividends. Most of the
coefficients of the interaction between
G
and
Legal
(
Right
and
Disclosure
) are negative
and significant in each panel, which implies that in countries with better investor
protection, faster growing firms pay lower dividends. These results are consistent with
those of La Porta et al. (2000), which suggests that legally protected minority shareholders
are willing to wait for their dividends because of higher reinvestment rates.
The main objective of our study is to examine the governance role of legal
institutions in determining the dividend policies of controlling shareholders with a
large control divergence, with a focus on the interactions among Diverge, legal
institutions, and investment opportunities. The coefficients of
G*Diverge*Legal
and
G*Diverge*Disclosure
are negatively and significantly associated with all four dividend
payout ratios in Panels A and C. The coefficients of
G*Diverge*Right
are negatively and
significantly associated with
IADvd/Sales
and
IADvd/E
in Panel B. These results suggest
that the level of dividends that are paid out by firms with a high level of control