254 /342
美國產險業
CEO
更迭與再保險需求
254
reported that some boards of directors became more intense in replacing managers after SOX
(e.g., Kaplan and Minton, 2012; Kim, Robles, Cho, Lee, and Kim, 2008). Consequently, in
the post-SOX environment, managers of insurers are more risk averse and thus purchase
more reinsurance. These changes imply that an increase in the non-routine (forced) CEO
turnover is associated with increasing reinsurance demand to reduce corporate risk after
SOX. Thus, we expect the sign of the interaction term between non-routine (forced) CEO
turnover and the Sarbanes-Oxley Act to be positive.
On the other hand, Ho et al. (2013) find that insurers do not significantly change their
risk-taking behavior post-SOX with one exception, that is, insurers used less leverage. Since
reinsurance demand is related to risk-taking behavior, the result of Ho et al. (2013) implies
that the sign of interaction term between routine (voluntary) or non-routine (forced) CEO
turnover and SOX on reinsurance demand cannot be predicted. Based on the above
discussions, we provide the following null hypothesis:
Hypothesis 5: There is no interaction effect between CEO turnover and the Sarbanes-
Oxley Act on reinsurance demand.
3. Data and Methodology
This section discusses data collection and methodology.
3.1 Data
Our sample consists of 252 U.S. property casualty insurance companies and 2,772 firm-
years during the period from 2000 through 2010. The total premiums of our sample is 80%
of total industry premiums in 2000. We hand collect detailed information on insurers’
corporate governance variables including CEO turnover (routine and non-routine CEO
turnover), CEO/Chairperson duality, board size, the percentage of independent directors on
the board and auditors from A.M. Best’s Insurance Report (Property-casualty) from 1999
through 2011. Specifically, we follow previous finance literature (e.g., Borokhovich et al.,
1996; Campbell et al., 2011; Huson et al., 2001; Huson et al., 2004) to collect forced CEO
turnover or voluntary CEO turnover data. We use Google, companies’ websites or other
websites to collect the relevant information. Based on the collected information, we define
whether the CEO turnovers are forced or voluntary CEO turnovers. Organizational structure
and other financial data are obtained from the National Association of Insurance
Commissioners (NAIC) InfoPro database for the period 1996-2010. We measure standard
deviation of loss ratios by using five-year rolling data to proxy underwriting risk. For