Band 4, Heft 1

We compare two prominent approaches to capital allocation in insurance firms. The financial theory approach includes Merton and Perold (1993) and Myers and Read (2001). The cooperative game theory approach utilizes concepts such as the Shapley value and the Aumann-Shapley value. We argue that, when an entire division is added or when the effect of a decision is discrete, the Shapley value approach provides an improvement over the Merton and Perold approach in that it properly accounts for the order in which divisions are added, and resoles the unallocated capital problem. When the effect of a decision is continuous, we show that the Auman-Shapley value approach not only provides game theoretic support for, but also conceptually extends, the Myers and Read approach.

This study examines empirically how risk management, financial management, and capital management are related to each other in the property/liability insurance industry, thereby reflecting interactions in managerial decisions such as the choice of derivatives and reinsurance use, the allocation of asset risks, the determination of underwriting activities and liability risks, and the adequacy of capital levels. This research contributes to the literature by adopting structural equation modeling to capture not only the dynamic interactions among capital management, financial management, and risk management, but also the effects of internal integrated factors or external regulatory factors on these managerial decisions. Among the empirical findings of this study, it is shown that risk management, capital management and financial management should be assessed under an integrated framework as emphasized in the concept of enterprise risk management. In addition, regulatory requirements embedded in the insurance industry might play an essential role in the integration of managerial decisions.

Using a sample of Malaysian non-life insurers over the period 2000-2007, I develop and test a model that explains the expense ratio as a function of output level, line-of-business diversification and other variables. Using various proxies for scale and line-of-business diversification, the empirical results show the existence of an interactive and non-linear relationship with cost. The findings suggest that there is no single minimum efficient scale level for all insurers; rather, it is contingent on the insurer's degree of line-of-business diversification and vice versa. I also find that costs for newly consolidated insurers and motor insurance are relatively higher and fail to find evidence that foreign-owned insurers or foreign joint venture insurers are more cost efficient.

In the U.S., a majority of the non-elderly population obtains health insurance coverage through an employer. State governments enacted regulations that prohibit the use of a variety of underwriting criteria, ostensibly to make insurance more affordable to those who would have otherwise been denied coverage or charged higher rates. This study tests whether regulations that restrict the use of disability status as an underwriting criterion for small businesses have unintended consequences in the labor market. The findings suggest that in states where disability is prohibited as an underwriting criterion, both disabled and able-bodied workers of small firms earn more than their counterparts not subject to such restrictions, albeit due to different causes.

Using an actuarial model, we examine the cost of delay in mortgage/credit loan payments. It is assumed that the default arrival process follows the Poisson process and the loss sizes are assumed to be independent and an identical truncated exponential. We also assume that the delay between default occurrence and partially (or fully) recovered payment is an independent identical truncated exponential random variable. For the recovery rate random variable, we simply use its expectation. Using the relationship between the shot noise process and accumulated/discounted aggregate losses process and applying the piecewise deterministic Markov processes theory, we obtain the explicit expressions for the expected value of losses and the expected value of part (or whole) of the loan recovered with the delay. Based on these moments, we define and predict the cost of delay in a mortgage/credit loan portfolio and their numerical examples are provided.

These comments are a reaction to a panel discussion on Catastrophe Risk Management and Insurance. They point out that some measures intended to provide quick relief and remediation after catastrophic events may have the long term effect of increasing aggregate exposure to such events. In some cases, these adverse long-term effects can be mitigated.