Evaluating catastrophe reinsurance contracts: an option pricing approach with extreme risk

摘要

This study evaluates a government-sponsored Excess-Of-Loss (XOL) Catastrophe (CAT) reinsurance contract using the financial option approach with extreme risk. We show that the Generalized Pareto Distribution (GPD), a Peak-Over-Threshold (POT) model, can properly depict the extreme losses from natural disasters in Taiwan, and thus can produce the most moderate premium estimates compared to other tail distributions. We contend that the risk neutral pricing is applicable even if CAT is a systematic risk and the reinsurance market is incomplete. Lastly, the impact of choosing thresholds on premium estimates is also examined.