Effects of business diversification on asset risk-taking: Evidence from the U.S. property-liability insurance industry

Decomposing diversification effect: evidence from the U.S. property-liability insurance industry

“…To the extent that diversification reduces underwriting risk (from coinsurance benefits), this theory suggests that diversified insurers will increase investment risk. Che and Liebenberg (2017) provide evidence consistent with this theory, as they find that diversified insurers invest in riskier assets and that this relation holds in an event study setting. However, while their analysis suggests a positive relation between diversification and investment risk, it remains an empirical question whether diversifiers are able to realize the higher expected return associated with riskier investments.…”

“…We measure reinsurance activity with REINSURANCE_RATIO, the ratio of premiums ceded to non-affiliated firms divided by the sum of direct premiums written and reinsurance assumed from non-affiliates. Che and Liebenberg (2017) find evidence that affiliated insurance companies hold riskier assets. Therefore, we expect affiliated firms to have higher investment returns than unaffiliated firms.…”

The effect of predation risk on cash holdings: Empirical evidence from the U.S. property‐liability insurance industry

“…While insurers report direct and net premiums written at the business line level, insurers do not report net premiums written at the state/territorial level. Given the lack of available data, we follow prior literature and calculate the LINES_DIV variable using net premiums written and the GEO_DIV variable using direct premiums written (e.g., Che & Liebenberg, 2017;Pottier & Sommer, 2006;Xie et al, 2017). 23 Given similarities between the MS_WCONC variable and our two diversification proxies, we check both correlations across the variables and the variance inflation factors (VIFs).…”

“…Che and Liebenberg () report that approximately 72% of the firms in their sample write multiple lines of business while the remaining 28% write only a single line of business. Similarly, Berry‐Stölzle et al () report that the mean and the median values of geographic diversification are 0.208 and 0.000, respectively, for single‐line insurers in their sample, implying that there are both insurers that operate in multiple states and insurers that focus on only one state.…”

“…While insurers report direct and net premiums written at the business line level, insurers do not report net premiums written at the state/territorial level. Given the lack of available data, we follow prior literature and calculate the LINES_DIV variable using net premiums written and the GEO_DIV variable using direct premiums written (e.g., Che & Liebenberg, ; Pottier & Sommer, ; Xie et al, ).…”