This dissertation introduces a game theoretic modeling framework and a series of models to examine the interactions between the key stakeholders (property owners, insurers, reinsurers and government) of a natural catastrophe insurance market, which possesses a complicated structure and faces many challenges from the natural catastrophe loss. Specifically, we integrate (1) a utility-based homeowner decision model; (2) a stochastic optimization model to optimize reinsurance decision by the primary insurer(s); (3) a heuristic government intervention model to reduce uninsured losses through price support for insurance purchase and acquisition; and (4) a state-of-the-art regional catastrophe loss estimation model, all within the framework of a static Cournot-Nash noncooperative game assuming perfect information. We allow the number of primary insurers to increase from one (monopoly) to many (oligopoly) within the Cournot-Nash framework, and examines the impacts of competition on market performance from each stakeholder's perspective. An automatic Response-Surface and Trust-Region algorithm is developed to solve the models for real, regional applications. A case study for residential wood frame buildings in Eastern North Carolina is presented. The case study suggests that: (a) private insurance market competition is an efficient mechanism to reduce uninsured loss, which should be facilitated by government; (b) more competition challenges insurers but benefits homeowners, and there exists a balance between insurer profitability and insurance penetration; (c) acquisition, price support and encouraging insurers to keep catastrophe reserve can all improve market performance and reduce uninsured loss; and (d) catastrophe reserves should be encouraged, which not only help insurers to avoid insolvency, but could also limit competition if imposed as barrier of entry, thus improve their profitability.
展开▼