Assessing Catastrophe Risk Pricing in a Coastal Insurance Firm

Thesis Abstract

Coastal communities face escalating catastrophe risks that challenge the pricing accuracy and financial resilience of insurance providers operating in high-exposure markets. This study investigates how catastrophe risk is priced by a leading coastal insurance firm, examining the interplay between exposure, model assumptions, and market competitiveness to identify systematic mispricings and risk-transfer inefficiencies. The aim is to develop a robust pricing framework that enhances predictive accuracy, aligns premiums with true risk, and informs strategic underwriting decisions in a dynamic climate-risk environment. Specific objectives include (1) to quantify the contribution of physical and financial risk factors to premium setting; (2) to evaluate the adequacy of external catastrophe models and internal loss experience in informing rates; (3) to assess the role of reinsurance arrangements and capital-adequacy constraints on pricing choices; (4) to examine how policyholder characteristics and market bidding behavior influence observed pricing differentials; and (5) to propose an integrated pricing model that combines structural, stochastic, and behavioral components for improved accuracy and capture of tail risk. The research adopts a mixed-methods design integrating quantitative and qualitative components. A dynamic panel data approach analyzes 36 months of underwriting data from the firm, encompassing 2,500 coastal property and casualty policies across three regional hubs, supplemented by 5 years of external catastrophe model outputs (Hurricanes, floods, and storm surge scenarios). Regression techniques, including multilevel generalized linear models and quantile regression, assess the determinants of premiums beyond expected loss, while a Bayesian updating framework evaluates model risk and uncertainty propagation into pricing decisions. The qualitative strand employs semi-structured interviews with 12 senior underwriters and five risk actuaries to capture organizational heuristics, governance structures, and perceptions of model credibility. The study triangulates findings with internal loss-adjustment records and with market-level competitive pricing data from 18 comparable coastal insurers, obtained through industry benchmarks and regulatory disclosures. Instrumentation includes standardized underwriting pricing templates, model input checklists, and a thematic interview guide aligned with established catastrophe risk theories. Key expected findings include (a) identification of statistically significant gaps between external catastrophe model outputs and realized loss experience, resulting in systematic underpricing or overpricing in specific peril-band scenarios; (b) demonstration that price-to-expected-loss ratios are augmented by tail-conditional risk measures such as value-at-risk and expected shortfall, particularly for properties with high flood exposure and proximity to surge zones; (c) evidence that internal capital constraints and reinsurance layering influence premium levels through capital-at-risk considerations; (d) insight into behavioral pricing biases among underwriters, especially in competitive bidding contexts, that contribute to margin compression during peak hazard periods; and (e) development of a composite pricing model incorporating structural risk (exposure, vulnerability), stochastic catastrophe scenarios, and organizational governance factors, improving out-of-sample pricing accuracy by 8–12% in simulated deployments. The study contributes to knowledge by bridging actuarial theory with organizational pricing practice in a high-risk coastal setting. It extends the literature on catastrophe risk pricing by integrating agency theory with risk governance, the appraisal of catastrophe models under practice constraints, and the impact of reinsurance architecture on rate-setting strategies. The proposed pricing framework offers decision-useful insights for managers on optimal underwriting and capital allocation, while informing regulators about pricing soundness and resilience of coastal insurers. The main conclusion anticipates that a blended pricing approach—combining external model validation, loss-history calibration, and governance-aware adjustments—substantially enhances pricing fidelity and financial stability in the face of escalating climate-driven catastrophes. Recommendations include adopting a formal model-risk management process for catastrophe pricing, enhancing data-sharing with reinsurance counterparties to refine tail risk assessments, expanding scenario testing to incorporate compound hazard events, and institutionalizing periodic pricing audits to mitigate underpricing and adverse selection during high-risk episodes.