Insurer’s Risk-Taking Behavior in India: Does Board Matter?

Abstract

This study investigates the board’s role in shaping insurers’ risk-taking using data of 49 Indian insurance firms from 2014 to 2021. We captured the insurer’s risk using four proxy measures: underwriting, investment, insolvency and total risks. To address potential endogeneity concerns in the relationship with board attributes, we employ a two-step Generalized Method of Moments (GMM) approach. Robustness checks are conducted using the panel quantile estimation approach. The empirical analysis reveals that life insurers face greater underwriting risk than non-life insurers. A larger board with duality appears to mitigate insurers’ risk, particularly in the case of life insurers. The presence of independent directors on insurer boards, along with an optimal frequency of board meetings, effectively mitigates risk among Indian insurers. The study concludes that dual regulations erode regulatory discipline among public insurers, leading to an inadequate governance structure.

Keywords:

• Board governance
• Risk-taking
• India
• Insurance
• Panel quantile
• Dynamic panel estimator

JEL CLASSIFICATIONS:

• G22
• G32
• G34
• C23

Acknowledgements

The authors thank to the editor and anonymous reviewers for their valuable comments and suggestions, which helped to improve the quality of the paper substantially. Any errors remain the authors’ responsibility.

Disclosure statement

Authors report no potential conflict of interest.

Data availability statement

The data that support the findings of this study are available on request from the corresponding author.

Notes

1 It is worth noting here that, in the present study, we referred to “risk-taking” and “risk-taking behavior” oppositely. Risk is a fundamental driving force in financial institutions, particularly insurance firms. So, the purpose of board governance is to control excessive risk-taking by optimizing its level to avoid a firm’s failure since a minimum level of risk is always involved in insurance business activities. We consider a negative relationship between a particular board attribute and risk-taking when it signifies that a higher intensity or the presence of a particular board attribute improves risk-taking behavior. In other words, a negative relationship between board governance and risk-taking indicates that the board effectively controls or optimizes excessive risk. However, the insolvency risk is proxied by the Z-score, which is a direct measure of soundness and an inverse measure of risk. Therefore, we expect a positive coefficient between board attribute and insolvency risk to improve risk-taking behavior.

2 As of March 2022, the solvency rates of United Insurance, Oriental Insurance, and National Insurance stood at 0.51, 0.15, and 0.63, respectively (IRDAI 2022). Over the past three years, the Government of India has injected a total of 17,450 crores into these three state-owned general insurance companies, with contributions of 2,500 crores in 2019-20, 9,950 crores in 2020-21, and 5,000 crores in 2021-22 (Economic Times, 2023). Available at: https://economictimes.indiatimes.com/industry/banking/finance/insure/finmin-to-take-a-call-on-rs-3000-crore-fund-infusion-based-on-performance-of-psu-general-insurers/articleshow/100128628.cms.

3 In 2021, the insurance penetration rate in India was 4.2 percent, and the density level was US$ 78; however, the world average insurance penetration rate was 7.40 percent, and the density level was US$ 809 (IRDAI 2021a).

4 Several studies have used the rolling window of three (Elango et al. 2008; Cummins et al. 2017), five (Ho et al. 2012; Podder et al. 2013; Han and Jung 2023), or more years (Eling and Marek 2014). In the present study, we have adopted the three-year rolling window based on our sample years. We performed the robustness check by taking the change in risk relative to the previous year. Further, for the computation of the Z-score and risk-adjusted performance, we used the rolling window of two years. The results are shown in the Appendix Table A5, and these are broadly consistent with our main finding