265 /342
臺大管理論叢
第
27
卷第
2S
期
265
demand. CEO/Chairperson duality is negatively related to change in total reinsurance ratio
and affiliated reinsurance ratio in Models of ∆Reins_ratio and ∆Reins_aff_ratio, suggesting
that the insurers with CEO/Chairperson duality tend to decrease total reinsurance ratio. The
decreasing total reinsurance ratio is driven by reinsurance ratio from affiliated reinsurers. A
possible reason is that the CEO/Chairperson of the board duality focuses more on insurers’
performance than riskiness. The coefficient of board size is not significantly related to
change in reinsurance ratio. The change in board size does not affect change in reinsurance
ratio. The change in percentage of independent directors on the board is positively and
significantly related to change in reinsurance from non-affiliated reinsurers. This finding can
be attributed to independent directors, and can serve as external monitors to ascertain that
reinsurance decisions transfer risk better to non-affiliated reinsurers, and reduce expected
insolvency risk. The coefficient of Big 4 auditor is not significantly related to change in
reinsurance.
For control variables,
Ln
(
na
)
i,t
is a proxy for firm size. We find change in firm size is
negatively and significantly related to change in total reinsurance ratio and affiliated
reinsurance ratio, indicating larger insurers have lower reinsurance demand. This result is
consistent with the results of previous studies (Mayers and Smith, 1990; Hoyt and Khang,
2000; Garven and Lamm-Tennant, 2003; Weiss and Chung, 2004; Cole and McCullough,
2006; Powell and Sommer, 2007; Garven et al., 2014; Wang et al., 2008; Ho and Lai, 2014;
Ho, 2016). The change in Herfindahl index is not significantly related to change in
reinsurance demand in all Models. Changes in Geographic Herfindahl index in the Models of
∆Reins_ratio and ∆Reins_aff_ratio are negative and significant, suggesting insurers with
higher geographic Herfindahl concentration tend to lower total reinsurance ratio and
affiliated reinsurance ratio. This result is consistent with previous findings of Mayers and
Smith (1990), Cole and McCullough (2006), Powell and Sommer (2007), and Garven et al.,
(2014) that insurers are more geographically concentrated resulting in lower reinsurance
demand. The empirical result shows that change in leverage is positively and significantly
related to change in reinsurance ratio from non-affiliated reinsurers. This result shows that
the positive relation between reinsurance demand and leverage is consistent with the findings
of Hoerger et al. (1990), Adams (1996), Garven and Lamm-Tennant (2003), Shortridge and
Avila (2004), Powell and Sommer (2007), Adams et al. (2008), and Shiu (2011). We find that
change in underwriting risk is negatively related to change in total reinsurance ratio,
reinsurance ratio from affiliated reinsurers, and reinsurance ratio from non-affiliated
reinsurers, implying that insurers with higher underwriting risk are more likely to lower