| This dissertation introduces a new modeling framework to better understand and help improve regional natural disaster risk management in the United States, including the interactions among key stakeholders--primary insurer, homeowner, reinsurer, and government; and between the two important risk management mechanisms of insurance and retrofit. The framework centers on a new game theoretic model of the interactions between homeowners and insurers. A stochastic programming optimization represents the insurer's decisions about what premiums to charge and how much risk to transfer to reinsurers and in what way. The optimization interacts with a utility-based model of each homeowner's decision of whether to buy insurance and/or retrofit, and if so, which specific retrofit strategy. Reinsurer and government roles are represented as inputs to the insurer-insured interactions. The decision models are integrated with a regional catastrophe loss and retrofit estimation model. The framework has four key strengths: (1) the insurer and each homeowner is assumed to act in its own best interest and subject to its own constraints; (2) outcomes are considered from the perspective of each of the key stakeholders explicitly; (3) both insurance and retrofit are considered in the same analysis, allowing consideration of the interactions between them; and (4) detailed, disaggregated state-of-the-art loss estimates are integrated, and they properly represent spatial correlation and the possibility of multiple hurricanes in a short time frame. The models are all applied in a full-scale case study for hurricane risk to residential buildings in Eastern North Carolina, and results are examined under different policy configurations, i.e., introducing retrofit in addition to insurance, providing government subsidy to reduce retrofit cost, offering both government subsidy and insurance rebate to help pay for retrofit, and mandating insurance purchase with a cap on insurance premiums. The case study analyses suggest that it is possible to design different system configurations in which each stakeholder could be better off, that retrofit incentives for insured homeowners can be effective in linking and strengthening the benefits of retrofit and insurance, and that mandatory insurance can be effective together with capped profit loading factors and possibly retrofit rebates from the insurer to the homeowner. |
