An Empirical Study on the Solvency Prediction of Value at Risk and Risk-Based Capital

Assuring insurance companies' solvencies has always been the focal point of insurance regulation. Among solvency regulation methods, risk based capital (RBC) represents the leading capital adequacy requirement in the insurance industry while value at risk (VaR) has been adopted by more and more regulators in other financial industries. The purpose of this research is to empirically compare the effectiveness of RBC and VaR in the solvency predictions of insurance companies. Using the solvency record of property-casualty insurers in the United States from 1995 to 1998, we examine the Type I and Type II errors of VaR and RBC in classifying insurers. RBC figures are from the annual statements and VaR is estimated from our simulation model that is also based on insurers' financial statements. We find that VaR is better than RBC in spotting unhealthy insurers and correctly classifying insurers. VaR has lower type I error rates and higher hit ratios. On the other hand, RBC has hardly any early warning abilities and very low hit ratios. The capital requirement set up by RBC is too low to detect problematic insurers in advance. Regulators have to raise RBC requirements many times to achieve acceptable early warning abilities. Although VaR is superior to RBC, the performance of VaR obtained from our model is not satisfactory. The type II error rates are high and hit ratios are low. We attribute the unsatisfying outcome to the lack of data in estimating loss functions.