385 / 414
臺大管理論叢
第
27
卷第
1
期
385
Table 6 Logistic Regression for the Occurrence of Equity-based Compensation for
Audit Committee
Variable
Coefficient
Z-statistic
FREE_TOBIN
-0.001***
(-3.26)
OVERLAP_AUCOMP
1.956***
(6.00)
OTHER_EXE
-0.405***
(-2.68)
INGOV
-0.450***
(-4.14)
EXGOV
0.060***
(2.87)
CASHCOMP
-10.414***
(-17.29)
BOARDSIZE
-0.008**
(-2.02)
LNSALES
0.062
(0.37)
LIQ_CONSTRAINTS
6.936*
(1.45)
STOCK_RETURN
-6.556***
(-3.17)
R&D
27.905
(0.49)
Intercept
6.235***
(5.42)
Industry Indicator Included
Yes
Pseudo-R
2
25%
Number of Observations
5,259
Note: The table presents model coefficients, with the following indicators of significance of the
Z-statistic: *: significant at the 10% level; **: significant at the 5% level; ***: significant at the 1%
level. Variables are as defined in Table 2.
Most of the other control variables also reveal a significant association with the
dependent variable. The results of the corporate governance variables (
INGOV
and
EXGOV
)
are mixed.
INGOV
shows a significantly negative association (Coefficient = -0.450
Z-statistics = -4.14), suggesting that better internal governance structure leads firms to give
their audit committees cash retainers instead of stocks and options. On the other hand, the
results of
EXGOV
(Coefficient = 0.06, Z-statistics = 2.87) indicate that a better external
governance structure makes firms more likely to compensate audit committee members with
stocks and options. Because our external governance measure is constructed using
institutional ownership and analyst following, the results might suggest that these external
monitors, including institutional investors and analysts, prefer equity-based compensation for
audit committee members because stocks and options better align committee members’
interests with those of investors.
BOARDSIZE
and
STOCK RETURN
are significant and
negatively associated with equity-based compensation for audit committee members
(Coefficient = -0.008 and -6.556, Z-statistics = -2.02 and, -3.17, respectively). Not
surprisingly,
CASHCOMP
shows a significantly negative (Coefficient = -10.414, Z-statistics