State Department of Insurance
Client: A Southeastern State Department of Insurance
Engagement: We reviewed a Homeowners filing for a major insurer that had a large market share in that state. The filing proposed significant rate increases on coastal areas as a result of recent heightened hurricane activity. We reviewed the filing and provided actuarial opinion on the proposed changes, giving special attention to the wind portion of the rate.
Merlinos’ Approach: Applying our in-house knowledge of hurricane models, we evaluated external adjustments to model output and built-in model options. Specifically, our analysis focused on the use of the near-term catalogue, adjustments for demand surge and storm surge, and on the trending of modeled hurricane losses to prospective levels. The near-term catalogue assumes that models can predict short-term variations in hurricane activity. Many modelers consider it less credible than the long-term catalogue; thus, it is controversial and warrants further scrutiny. In addition, the near-term catalogue has never been approved by the Florida Hurricane Commission. After reviewing the model output that we requested from the insurance company, we concluded that the adjustments for demand surge and storm surge were excessive. The trending of modeled hurricane losses to prospective levels using annual loss trend based on non-catastrophic losses is not an appropriate database for developing catastrophe loss trend as required by actuarial standards of practice.
We provided actuarial opinion on the net cost of catastrophe reinsurance loaded into the rates. We were concerned with the company’s approach of allocating the net cost of reinsurance to state and rating territories. Their approach resulted in reinsurance provisions which did not vary rationally or smoothly between adjacent rating territories. For some layers of loss, the insurer did not have an executed reinsurance contract. Rather, they loaded in a provision for reinsurance that was based on a hypothetical reinsurance premium which was quoted by the company’s reinsurance broker. For other layers of loss, the insurer considered reinsurance with an affiliated company. This affiliated reinsurance does not provide consumers the protection of transferring risk outside of the insurance group and is subject to a higher standard of scrutiny when included in the net cost of reinsurance. We were also concerned with the loading of primary insurer profit into the reinsurance provision. This is arguable, as reinsurance, if anything, reduces the primary insurer’s risk.
When we reviewed the hurricane deductible and ex-wind credits for consistency with the revised wind portion of the package rate, we recommended that the credits be increased to be consistent with the increase in the wind portion of the rate.
We considered the impact of other rating structures on the wind portion of the rate, such as rating structures related to consumer credit scores and those that assign credit/surcharges based on insured claims history. We objected to the use of the same set of consumer credit score factors for all rating territories. We had long suspected that, while consumer credit scores are a good predictor of non-catastrophic perils, such as fire, they are less predictive of hurricane losses. Thus, they should have less impact on rates in areas with high hurricane exposure than on the rates in inland areas. Because of our long-term relationship with this state’s Department of Insurance, we were able to monitor companies’ experience in this state and, ultimately, confirm that our suspicion was correct. We also objected to surcharges for catastrophic losses under the claims history credit/surcharge rating structure because these claim history surcharges and credits were developed using non-catastrophic data and did not consider catastrophic losses based on long-term averages or models, which is the industry standard.
We evaluated the non-hurricane portion of the proposed rates by rating territory. Because we discovered non-insurance historical data that clearly demonstrated greater non-hurricane wind peril inland compared with coastal areas, we objected to the use of a non-hurricane wind provision throughout the state.
Conclusion: We provided a summary of our analysis and conclusions to the Department of Insurance. This could be used, in addition to market and public policy considerations, to determine a course of action regarding the filing, to promote a healthy insurance market within the state.