Variables Predicted
sign SCAR16 SCAR12 SCAR12_May
Notes: The pvalues are reported in parentheses. *, ** and *** indicate statistical significance at the 10%, 5%, and 1% level, respectively in a two-tailed test.
SCAR16 is standardized cumulative abnormal returns from January of fiscal year t to April of fiscal year t +1; SCAR12 is standardized cumulative abnormal returns from January of fiscal year t to December; SCAR12_May is standardized cumulative abnormal returns from May of fiscal year t to April of fiscal year t+1; DO is a dummy variable that equals 1 if the firm purchases D&O insurance, and 0 otherwise; UE is unexpected earnings that is measured by income from continuous operations in year t minus the income from continuing operations in year t-1 and then scaled by the market value of equity; AGE is the number of years a firm has been listed;BIG4 is a dummy variable that equals 1 if auditor is a Big 4accounting firm, and 0 otherwise; GROWTH is growth opportunities that is measured by the sum of the market value of equity and the book value of debt, and then scales by total assets;
PERSIST is earnings persistence that is measured by the reciprocal of PE ratio; BETA is systematic risk that is measured by market model using the data of 24-60 monthly stock returns; SIZE is natural log of total assets; LEV is total liabilities divided by total assets; ROA is net income after tax divided by total assets; RESTATE is a dummy variable that equals 1 when the firm incurs financial statement restatements, and 0 otherwise; GDR is a dummy variable that equals 1 when the firm issues Global Depositary Receipts (GDR), and 0 otherwise; ECB is a dummy variable that equals 1 when the firm issues Euro-Convertible Bond (ECB), and 0 otherwise; INDUSTRY is a dummy variable that equals 1 when the firm is in high-tech industries, and 0 otherwise;STOCK is natural log of the number of shareholders; INDDIR is ratio of independent directors on the board;CONTROL is the ratio of ultimate controller on the board; CONTROLOWN is the percentage of shares held by ultimate controller;
MGTOWN is the percentage of shares held by managers; BONUS is natural log of the remunerations of directors and officers.
χ2
χ2
Panel A of Table 4 shows that the coefficients on UE (unexpected earnings) are significant and positive (p values are all smaller than 5%), consistent with prior studies. The coefficients on UE×DO are significant and negative (-2.147, -1.843, -1.630 respectively, p values are all smaller than 1%). In other words, these results imply that stock investors believing the purchases of D&O insurance is not conducive to earnings quality, and then the ERC will be weakened. The empirical results tend to support the argument that stock investors believing the purchase of D&O insurance for directors and officers liability will transfer part of potential litigation risk to the insurance firm. It may weaken directors and officers required attention to ensure a reliable financial report, as it is not conducive to earnings quality, and their earnings information on responses to stock prices will thereby be reduced. In addition, although the coefficients on DO is not an issue to be discussed in this study, and we find coefficients on DO are significant and negative, indicating firms with D&O insurance have the lower cumulative abnormal returns than those without D&O insurance22.
Panel B of Table 4 reports the results of D&O insurance choice model, and we find that firms with lower return of assets (ROA), higher debt ratio (LEV), higher growth opportunities (GROWTH), the electronic industry (INDUSTRY), the greater number of shareholders (STOCK), greater size (SIZE), lower ratio of controller (CONTROL) and the percentage of shares held by controlling shareholders (CONTROLOWN), and higher remunerations of directors (BONUS) are more likely to purchase D&O insurance. The empirical results demonstrate the determinants of demandfor D&O insurance are related to litigation risk, firm size, and corporate governance factors. Our results are generally consistent with prior studies (Core, 1997; O’Sullivan, 2002; Chung and Wynn, 2008; Zou et al., 2008;
Chen and Pang, 2008).
4.3 Additional Tests
4.3.1 The Effect of D&O Insurance Coverage on the ERC
22 We infer that firms with D&O insurance essentially have a higher risk (they would take the initiative to purchase D&O insurance), and on average, have lower cumulative abnormal returns than those without D&O insurance.
Wynn (2008) suggests that when the opportunistic incentives and opportunities for directors and officers are greater, D&O insurance coverage is greater for transferring legal liability arising from opportunistic behaviors. In other words, D&O insurance coverage can more accurately capture moral hazard problem of directors and officers. We further explore the effect of D&O insurance coverage on the ERC for firms with D&O insurance. We divide total D&O coverage into normal insurance coverage (NormalDO) and abnormal insurance coverage (AbnDO). Following prior studies, we construct D&O insurance coverage demand model (equation (3))23 to estimate normal D&O insurance coverage and abnormal D&O insurance coverage (that is the error term of equation (3)). Table 5 presents the estimated results for the determinants of D&O insurance coverage model. We find that the adjusted R2 of D&O insurance coverage demand model is 80.5%, and most of coefficients of D&O insurance demand variables (ROA、LEV、
GROWTH、STOCK、SIZE、CONTROL) are significant and consistent with the
predicted direction. The results reveal that the D&O insurance coverage model constructed in this study can reasonably estimate normal D&O insurance coverage and abnormal D&O insurance coverage.Table 6 summaries the results of the effect of total insurance coverage (DO_Amt), normal insurance coverage (NORMAL_DO ) and abnormal insurance coverage (AbnDO) on the ERC. We find that most of coefficients on UE×DO_Amt,
UE×NORMAL_DO and UE×AbnDO are negative and significant at 10 percent
level, except that when dependent variable is SCAR12_May (only one-tailed test is significant). This indicates that regardless of total insurance coverage, either normal or abnormal insurance coverage will reduce the ERC. In other words, for firms with D&O insurance, higher D&O insurance coverage have the lower ERC and firms with higher abnormal D&O insurance coverage will further deteriorate the ERC.DO_Amt is measured by total D&O insurance coverage scaled by book value of equity, the other variables in Equation (3) are the same as Equation (1).