Challenges with drawing conclusions from the Clarity Data Act

The Alabama Legislature charged the Department of Insurance (ALDOI) with the responsibility of collecting data specified in the Property Insurance Clarity Act of 2012 and providing public access to that data. In November, 2013, the ALDOI published such data on its website, which may be found at the following link: (http://www.aldoi.gov/Consumers/ClarityActConsumers.aspx. These additional links are available on that website page:

• Industry Data by Zip Code

• Property Insurance Clarity Act

• Basic Flow Chart of the Homeowners Insurance Ratemaking Process

• Homeowners Ratemaking Process – Executive Summary

• Description of the Homeowners Ratemaking Process permitted by the Alabama Dept. of Insurance

• Blank Sample Exhibit of the Aggregated Industry Data

• Companies Reporting Clarity Act Data

• Companies Not Reporting Clarity Act Data (including exempt companies)

A number of interested parties have been studying the Clarity Act industry data by county and have reached several conclusions, such as:

• Coastal homeowners losses are less expensive than upstate homeowners losses.
• Mobile & Baldwin Counties homeowners have been over-charged for their homeowners insurance, and have been subsidizing the upstate homeowners insurance cost.

• The ALDOI has allowed draconian rate changes on the Coast despite not having historical loss and premium data by zip code or counties.

The ALDOI has been asked on numerous occasions for our opinion of these conclusions, and we have been asked how we intend to make use of the Clarity Act data to better evaluate homeowners insurance companies’ requests for rate increases.

Let’s look at the last bullet statement first. For many years before the Clarity Act was passed, the ALDOI has required insurance companies to include in their homeowners rate filings 5 years of premium and losses by territory and statewide to support their requested rate change. Each company defines their rating territories differently across the state; some populous counties are subdivided into multiple rating territories (though not necessarily down to each individual zip code), and some less populous counties are combined with other counties to form a single rating territory so as to give that territory statistical credibility. No insurance company is permitted to raise rates in any single territory unless that territory’s loss and premium data statistically justifies that increase. The ALDOI enforces this requirement without exception, to the point of having required companies to even reduce rates when a territory’s data indicated that the rates were too high. Since insurance companies are already providing their homeowners territorial data to the ALDOI in their rate filings, there is no added value to the ALDOI using the Clarity Act data to evaluate the rate filings.

In addition, the Clarity Act includes losses from hurricanes. In the past, ratemaking methodology required insurance companies to use 20-30 years of hurricane losses to project future losses, but that methodology contained a number of shortcomings. With the advent of hurricane computer models over the last 20 years, the models have proven to be a better predictor of future hurricane losses than the historical losses. Therefore, the current accepted standard ratemaking methodology no longer makes use of actual hurricane losses, but replaces actual losses with projected losses from those computer models. Insurance companies provide those model projections in every rate filing submitted, which the ALDOI has determined is more appropriate than the Clarity Act data.

With regard to the other bullet points cited above, the ALDOI has a number of concerns with the usefulness of the Clarity Act data for drawing such conclusions.

What is included (and not included) in the Clarity Act data? There are five key issues with using the Clarity Act data to draw conclusions about the profitability of individual counties, and the equity between premiums being charged by county.

(1)   Hurricane vs. Tornado losses: The Clarity Act data currently covers 2003-2012. This data for Mobile/Baldwin Counties includes hurricane losses from 2004 and 2005. These hurricanes were not atypical of hurricanes that we can expect to occur every 15-25 years on our Coast. The Upstate data includes numerous tornado losses, but especially includes the losses from the 2011 Tuscaloosa Outbreak. Tornado computer models developed by meteorologists and statisticians suggest that the Tuscaloosa Outbreak has a chance of reoccurring in Alabama once every 250 years. Comparing results on the Coast that include a few typical hurricanes to results Upstate that include a very large and rare tornado outbreak is not appropriate.

(2)   Surplus Lines insurer data not included. Surplus lines insurers are not regulated for the most part by the ALDOI; therefore they were not subject to the Clarity Act and did not report their data to the ALDOI. In Mobile/Baldwin Counties, the surplus lines insurers write approximately 20% of the homeowners policies (and their policies are most likely closer to the beach and thus more subject to wind losses), while Upstate surplus lines insurers write less than 1% of the market. With such a large percentage of the premiums and losses excluded from the Coastal data, one would be concerned that comparisons between the Coastal data and the Upstate data might not be valid.

(3)   Double-counting of Wind-Only Policies. There are many homeowners in Mobile/Baldwin Counties (approximately 15,000) who purchase two homeowners policies: one for wind-only (mostly from the Alabama wind pool), and one for all other perils from other insurers. Since both policies are included in the Clarity Act data, the Coastal data are over-stated by approximately 10%, thus distorting the policy counts and any comparisons between losses per policy on the Coast versus Upstate.

(4)   Missing data from homeowners no longer purchasing Wind coverage. Wind coverage for homes near the coast represents approximately 75% of the total homeowners premium. Many homeowners, facing tightened budgets, have selected to “go bare” and not purchase wind coverage any longer, but only purchase a policy for all other perils. Homeowners Upstate do not have this option. Therefore, the losses per policy on the Coast would be under-stated since they do not include any wind losses for these homeowners, while all wind losses for all homeowners Upstate are included. This distorts any comparison between the Coast and Upstate. The ALDOI has no means to estimate how many Coastal homeowners choose to “go bare” for Wind coverage.

 (5)   Clarity Act data only includes losses, no expenses. Compiling the Clarity Act data over the 10 years 2003-2012, one finds the loss ratios for Mobile/Baldwin Counties versus Remainder of State to be 51% versus 92%, respectively. Just looking at these loss ratios certainly could suggest that the Coast is less expensive for insurance companies, and that the Coast is subsidizing Upstate. However, there are numerous other expenses that insurance companies incur when writing a homeowners insurance policy, such as: claims adjusting expense, commissions to agents, internal overhead expense, premium tax, and reinsurance. In addition, insurance companies must be able to earn a profit over the long-run in order to stay in business. Most of these expense loads are identical between Coastal homeowners policies and Upstate homeowners policies, but that is not the case for reinsurance expense and for the necessary profit margin. Reinsurance (purchased mostly from Bermuda, London and Europe) is more expensive for Coastal policies than for Upstate policies. And since the Coastal wind losses are typically less predictable from one year to the next compared to Upstate wind losses, the profit margin (or cost of capital) must be greater on the Coast than Upstate in order to attract companies to invest their capital in writing insurance on the Coast. Overall, the cost of reinsurance and cost of capital on the Coast runs about 10% higher than it does Upstate. Combining all of the expenses and profit margins together with the loss ratios cited above, the total results for the Coast versus Upstate become approximately 104% versus 135%, respectively.

What do these total results suggest? This indicates that on the Coast over the last 10 years, for every $100 of premium collected from homeowners, insurance companies have paid out approximately $104 for losses, expenses, and a required profit margin/cost of capital. That means that they have not earned as much profit margin as is reasonably required on the Coast in order to sustain their company over the long run. Hence, many companies have reduced the amount of Wind coverage that they write on the Coast so as to reduce the amount of capital that they have to dedicate to the Coast. This clearly indicates that over the last 10 years, homeowners on the Coast have not been overcharged for their homeowners premium. Subsidizing generally implies that one person is overcharged in order for another person to be undercharged. If Coastal homeowners have not been overcharged, then it is not possible that they are subsidizing the premiums for homeowners Upstate.

What does this say about the premiums being charged to homeowners Upstate? Homeowners are being undercharged by insurance companies for their insurance. The ALDOI has acknowledged this fact for several years. Companies have chosen to be more competitive in their pricing of insurance Upstate because the losses Upstate have generally shown to be more stable and predictable (although the Tuscaloosa Outbreak was a very rare deviation from this).