Risk-averse insurer capped-risk sensitive lending during the COVID-19 pandemic

ABSTRACT

This paper develops a contingent claim model of a risk-averse life insurer’s equity with various borrowing-firm credit risk features. The insurer’s lending function with various financial technology involvements creates the need to model equity as a capped/naked call option in insurer-borrowing firms. As a result, the insurer benefits from the capped-risk lending strategy yielding a higher interest margin. However, either the severe novel coronavirus (COVID-19) pandemic or the substantial risk aversion deteriorates policyholder protection. In addition, stringent insurer capital regulation reduces the insurer’s interest margin, thus increasing policyholder protection and contributing to insurance stability but discouraging insurer financial technology involvements.

KEYWORDS:

• Capped call option
• risk aversion
• Covid-19
• financial technology
• capital regulation

Acknowledgments

The authors would like to thank two anonymous referees for their helpful comments and suggestions. The usual disclaimer applies.

Notes

1. See Dermine and Lajeri (2001).

2. We model a very simple structural-break approach to demonstrate the COVID-19 pandemic. Specifically, the structural breaks are quadratic forms, decrescent to the asset return and increscent to the asset volatility. The quadratic changes might be possible from the vaccination gradually popularized on which we are silent.

3. See Appendix.

4. See American Council of Life Insurers, Fact Book, available at https://www.acli.com/posting/rp19-010.

5. There are 9 tables presented in our paper. We only focus on testing the robustness of Table 1 at a case of various levels of liquid-asset interest rate. We also conduct the robustness test using other parameters in some tables. The same pattern as previously applies.

6. Control variables for the robustness test in our model are choiceable. Our analysis is limited to the variable of the liquid-asset interest rate. We remain silent on other control-variable paraments.

7. Our results are not applicable to the case of $\mathrm{\partial }R/\mathrm{\partial }\alpha =$-21.1755.

Additional information

Funding

This work was supported by the National Natural Science Foundation of China under Grant 71603217 (Shi Chen).