FSU Catastrophe Center Discusses Mitigation Discounts
The Florida Catastrophic Storm Risk Management Center has released a report discussing implementation of Florida’s mitigation discount program for residential property insurance. This report is the third in a series of recent studies identifying significant concerns with the way the current system of mitigation discounts has affected the residential property insurance market. The Florida Commission on Hurricane Loss Projection Methodology published the results of its legislatively-mandated review on February 1, 2010. In March, Risk Management Solutions also released a report commissioned by the Florida Department of Financial Services.
The Risk Management Center’s new report agrees with many of the findings of the modeling commission’s review. In other areas, it expands upon or more clearly articulates concerns that have been plaguing the insurance industry over the last couple of years.
The Governor and Cabinet recently asked Commissioner Kevin McCarty to discuss underwriting results in the residential insurance market, including why so many insurers are losing money in the absence of recent hurricanes. The Risk Management Center’s study helps shed light on this issue, stating that problems in the implementation of the mitigation discounts are having a significant impact on the financial performance of Florida’s authorized insurers (and Citizens Property Insurance Corporation).
The report cites an initial decision to apply mitigation discounts to the “weakest” risk rather than an “average” risk as a major problem with the discounts. Insurers’ rates historically were based on their then-existing mixes of business. The mix of business would have included some houses with stong mitigation features, some with average mitigation features, and some with no mitigation features. When the discount tables were introduced, they in essence assumed that insurers’ policy portfolios were comprised of the weakest buildings, so homes with any mitigation features at all should receive discounts. This did not match reality, and therefore produced premiums that in the aggregate are too low considering that the underlying risk profile is essentially unchanged after intorduction of the discounts. This, in turn, has caused the industry’s loss ratios, expense ratios, and combined ratios to be high and its underwriting results to be poor.
The new study also points out a problem with using a single model to determine mitigation discounts. Although the modeling commission has determined that several models meet its standards, those models vary in their calculations and assumptions. This creates a potential disconnect between the model(s) an insurer uses to calculate its expected losses and the model upon which the mitigation discounts were derived.
The Risk Management Center’s review also addresses a significant problem that it terms “over-application” of the discounts. Mitigation features reduce the expected losses that homes would suffer from hurricanes. However, the premiums charged by an insurer encompass more than just expected losses. For example, the premiums include fixed expenses that are not reduced simply because certain homes have mitigation features. In addition, the executive summary of the initial ARA study clearly stated there had been no attempt to allocate expected loss reductions in condominiums or apartments among the buildings themselves and the contents separately insured by the various unit owners or renters. Applying the entire loss relativity differences, as converted into mitigation discounts, to HO-4 and HO-6 policies therefore does not appear to have support in the underlying study.
Another important point raised by the new report is the effect (or lack thereof) of the discounts on reinsurance costs. Applying the discounts to the entire wind premium implies that insurers will see reductions in their reinsurance costs commensurate with the discounts being given to consumers. However, this is not the case. The report points out that reinsurance premiums are designed to cover not only the reinsurers’ expected losses, but also a “risk load” they charge for taking on the catastrophe exposure. This risk load is not reduced by virtue of the mitigation discounts, meaning that the discounts actually are eroding premiums at levels exceeding the cost/expense savings to the insurers and leading to their poor financial performance. Likewise, insurers’ premiums include some recognition of the cost of capital, whether that capital is internally derived or obtained from capital markets. Insurers’ cost of capital is not reduced as a result of the mitigation discounts. In fact, although not explicitly addressed in the report, not only are the net cost of reinsurance and the cost of capital not reduced by the mitigation discounts, they actually can increase and eventually might become prohibitively expensive. When an insurer fails, its creditors, including its reinsurers, typically receive pennies on the dollar if anything at all. If reinsurers believe that Florida insurers pose significant credit risks in addition to the traditional catastrophe risks, the cost of reinsurance will be higher, at least for some insurers, than it would in a more stable environment. Likewise, when a large number of insurers show underwriting losses in non-catastrophe years, the Florida market looks relatively less attractive to investors than other opportunities, driving up the cost of capital.
The study mentions that some of the features for which discounts are granted are inherent to the homes, such as roof shape, whereas others are “actionable” by consumers, such as storm shutters. The study suggests that the state should stop referring to the inherent, non-actionable items as producing insurance “discounts.” Instead, those items should simply be rating variables. True discounts would be given only for those items where consumers take affirmative steps to improve the storm-worthiness of their homes when they have a choice whether to do so. This point relates to an excellent discussion in the RMS report pointing out that when insureds receive substantial discounts simply for having a form completed without taking any affirmative steps to mitigate their losses, the cost of taking affirmative steps to mitigate their homes is unattractive in light of the small incremental savings to be gained. Thus, while we might refer to Florida’s discount system as providing “mitigation” discounts, the system really only encourages consumers to have a form completed on their existing homes and provides little or no incentive for them to actually do anything to harden their homes.
The Risk Management Center’s report further suggests that the concept of affordability is a social issue than should not be effectuated through the insurance industry. The state should allow the insurance market to function according to economically sound principles, and should separately address affordability through need-based programs as considered necessary.
In sum, the Florida Commission on Hurricane Loss Projection Methodology, Risk Management Solutions (RMS), and the Florida Catastrophic Storm Risk Management Center have released reports in the last two months identifying signficant problems with the windstorm loss mitigaiton discount program in Florida. When trying to determine why insurers are losing money even in non-storm years, applying these discounts to portions of the premium that are not affected by the discounts, including the net cost of reinsurance, is a signfiicant contributing factor. In addition, the discount program is not achieving its major public policy purpose of incentiving consumers to strengthen their homes. Although there is uniform belief that improving the quality of Florida’s existing and future housing stock is a vital long-term goal, it is clear that the current system is not encouraging this result while unfortunately eroding the financial base that will be needed to rebuild after a storm. This disconnect produces signficant peril for the State of Florida, and the three recent reports should form the basis for revising the current system.