Dynamic financial analysis of property-liability insurance companies

This research applies state-of-the-art financial techniques to the dynamic financial modeling of property-liability (P-L) insurance companies. Dynamic financial analysis (DFA)--a relatively new concept in the P-L industry--attempts to analyze the underwriting and financial operations of an insurance company as an integrated whole. The objective of this dissertation is twofold: first, to provide a framework for current and future DFA models for the P-L insurance industry; second, to use these models to test various hypotheses concerning valuation, corporate decision-making, and optimal regulatory policy. Two types of DFA models are developed. The first is termed "conventional" since it builds on existing financial models in the life and P-L insurance industries. The single-period conventional model developed here incorporates insurance-specific factors, in a stochastic framework, and provides probability distributions of outcomes of key financial variables. The second type of model uses continuous-time financial techniques, which provide a more sophisticated approach to dealing with certain financial parameters. This alternative multi-period model can take many different forms, depending upon the particular specification. Each period of this model consists of a continuous-time segment between annual regulatory audits, and also allows for discrete jumps and instantaneous corporate and/or closure decisions. Parallel development of these two types of models will allow for an evaluation of the benefit of moving from a conventional to a more financially sophisticated alternative model, and will aid in the selection of appropriate parameters for future conventional models.