The analysis of capital structure for property-liability insurers: a quantile regression approach

By using a two-stage quantile regression approach (2SQR), this study demonstrates how the insurer's leverage is determined across various quantiles. The evidence shows that the influence of the business concentration and marketing channel at the lower leverage quantiles is opposite to that at the higher leverage quantiles, which proposes that the mean effects of the two-stage ordinary least squares method are insufficient to capture the effects of business strategies on the insurer's capital structure determination. Moreover, the 2SQR evidence also shows that the magnitude of the impacts for some determinations varies among the different leverage quantiles. In sum, the evidence suggests that these two competing approaches should be viewed as complementary functions when discussing the insurer's capital structure.