Empirical Study on Macroeconomic and Firm Specific Variables’ Effect on Profitability of Non-life Insurance Industry in Bangladesh

Abstract

This study tries to show the causation among non-life insurance firms’ profitability measured with ROE, ROA, EPS and firm-specific variables such as underwriting risk, reinsurance dependence, solvency margin, leverage, liquidity risk, growth of premium, length of the company, tangibility of firm’s assets in addition to macroeconomic variables such as GDP growth, inflation and improvement of stock marketplace of the insurance business of Bangladesh. This study also focuses on seven insurance companies amid 2011-2020 considering the econometric models followed numerous diagnostic checks along with the Model specification bias test, test of heteroskedasticity, cross-sectional dependence check, followed by autocorrelation test and unit root test at the validity of the models decided on for this examine. The study found that out of all independent economic factors, only the stock market development is significantly and positively affecting the profitability measured with EPS (Earnings per share). The insurance-specific variables such as leverage and size are significantly positively and negatively affecting the profitability measured both with ROE and ROA of insurance companies, respectively. They are significantly positively affected by both liquidity and tangibility of assets too. In addition, Solvency margin and premium growth are found to have significant negative impact on return on assets which contradicts the authors’ expectations. Moreover, underwriting risk is found to have a significant positive impact on EPS and ROE under the GLS and Fixed-effect method respectively. The one-step system GMM approach reveals that only the one-year lagged ratio of ROE is statistically significant among the three lag values used in the GMM approach and ROE is also significantly positively affected by underwriting risk and GDP growth. In contrast, leverage and size are found to have a significant inverse relationship respectively, with ROE in this approach, showing that tangibility of assets and leverage has a significant negative and positive impact on earnings per share. The chi-square values of the three above models are jointly statistically significant in explaining the variation in the respective dependent variable. In addition, it is noteworthy that no significant impact is found on return on assets from any of the three Macroeconomic factors considered for the study. Also, no significant relationship is found between the three measures of profitability and reinsurance dependence along with the inflation in the economy. However, other than the EPS, all the variables are jointly statistically significant in explaining 79.56% and 83.42% variation in ROE and ROA respectively under the Pooled OLS method.