Homeowners Insurance: Threats from Without, Weakness Within
December 1996
Executive Summary
Homeowners is the property/casualty insurance industry's third largest line of business by premium volume. But since 1980, homeowners has achieved poor financial results. Catastrophes certainly bear much of the blame. However, this study concludes that part of the problem is a pattern of enhancements insurers have made in their policy forms. The number of types of loss insured under a typical policy increased, contributing to an acceleration in losses that has outstripped premium growth.
Over the years, insurers repeatedly broadened the homeowners coverages to address growing customer needs, increase market share, and meet competition. But many coverage enhancements provided questionable advantages for insurers. The enhancements also made the programs more complex and added to their cost. The cumulative effect of the changes was more losses — and reduced profitability.
Homeowners costs reflect the evolution of coverage enhancements. Over time, through the use of insurers' loss and premium experience in ratemaking, homeowners rates recognize these costs to the extent permitted by marketplace forces. But it may take many years for the ratemaking data, and eventually rates, to reflect fully the costs of a single enhancement. A series of individual enhancements can have a continuing and cumulative effect, with rate levels constantly playing "catch-up" with the forces driving costs.
Industry Results
The catastrophes of the last seven years — particularly Hurricanes Hugo and Andrew — have led to unusually poor underwriting results. But even before 1989, the homeowners line of insurance exhibited high loss ratios, high expenses, and low investment return, making it a difficult line for insurers to write profitably.
From 1980 to 1984, homeowners operating ratios fluctuated little, averaging a marginally unprofitable 100.3%.[1]1. The operating ratio is the sum of the ratio of net incurred losses and loss adjustment expenses and policyholder dividends to net earned premium plus the ratio of net underwriting expenses to net written premiums minus the ratio of net investment gains to net earned premium. Between 1986 and 1988, homeowners was profitable, with an average operating ratio of 95.2%. Homeowners experience began to deteriorate in 1989, the year of Hurricane Hugo. From 1989 to 1995, the homeowners operating ratio averaged 115.6% — well above the average ratios of 97.3% for personal auto, 100.2% for commercial property, and 95.8% for other commercial lines. The homeowners operating ratio peaked at 151.9% in 1992, the year of Hurricane Andrew, and has not fallen below 107% since 1988.
Insurers generally pay homeowners claims very quickly. Without the long claim-settlement pattern characteristic of liability lines, homeowners cannot generate the same level of investment income per premium dollar. With low investment gains, insurers have to achieve higher levels of underwriting profit for homeowners than they do for longer-tailed liability lines to hit a target rate of return on surplus.
Homeowners expense ratios are high compared with personal auto and nonproperty commercial lines, partly because the average premium for homeowners policies is lower than for most other lines. Because many of the processing costs are the same regardless of the amount of premium, homeowners policies generate higher costs per dollar of premium.
Results by Market Segment
Expense ratios vary by company. The direct writers' average expense ratio increased 2 percentage points from 1987 to 1995, while the ratio for agency writers declined by almost 1 percentage point. Although the expense ratio for direct writers has increased, this group's consistently lower expense ratio — compared with the ratio for agency writers — and increasing market share have kept the industry ratio within a 1-percentage-point range (30.6% to 31.5%) from 1980 to 1995.
Results vary substantially by individual company and by type of company. Large insurers, generally direct writers, have experienced more favorable homeowners results than small and medium-sized insurers, primarily as a result of lower expenses. Yet higher catastrophe losses in recent years have hit large insurers harder than small and medium-sized insurers, because large insurers are more concentrated in catastrophe-prone states. This is especially true for the three largest writers — State Farm, Allstate, and Farmers. These three insurers accounted for more than 40% of homeowners written premium in 1995. All three are direct writers, and all have lower expense ratios than most other companies. But they also suffered extremely high losses during recent catastrophe years.
One reason for the three largest insurers' adverse results in 1992 was the group's high market share in catastrophe-prone states. The second reason was that these insurers tended to purchase little or no reinsurance, relying instead on their large, geographically diversified portfolios to manage catastrophe risk. In years with major catastrophes, such as 1992, reinsurance offers critical levels of protection to the smaller and medium-sized insurers. Sufficient reinsurance, particularly at existing rate levels, was not, and is still not, available for larger insurers.
Competition and Market Concentration
The countrywide property/casualty insurance market is competitive and not highly concentrated. The market includes a large number of companies, and new companies may enter the business relatively easily. Since 1991, the number of insurers writing homeowners (and other lines as well) has declined, because of consolidations and the potential for large catastrophe losses. This decline has increased market concentration. But the increased market concentration also reflects heavy competition for business, as insurers with competitive advantages grow in size. Competition has probably contributed to low premium growth and enhanced product offerings.
Market concentration for any line of business can vary considerably by state. This study looked closely at the five largest states. In the past 10 years, the number of insurers writing homeowners in three large, catastrophe-prone states dropped by 31% in Florida, by 29% in Texas, and by 23% in California. Catastrophes have not contributed to homeowners losses in Pennsylvania and New York as much as in California, Texas, and Florida. However, the number of homeowners insurers in Pennsylvania did decline 15% in the past 10 years. On the other hand, the number in New York increased 12%.
The drop in the number of insurers writing homeowners insurance in California, Florida, and Texas resulted in higher concentration in these three states — as well as an increased market share and disproportionately large share of countrywide catastrophe losses — for the three largest homeowners insurers.
Additional Analysis
From 1977 to 1995, premiums grew an average of 5.1% per year, while loss costs grew an average of 6.7% per year. Removing the experience of the worst catastrophe years — 1985, 1989, 1991, 1992, and 1994 — loss costs grew 5.7% per year — still half a percentage point higher than the average premium growth of 5.1%. The overall shortfall of 12.0% — the annual shortfall compounded from 1977 to 1995 — indicates that the larger catastrophes alone do not fully explain the poor homeowners underwriting results since the late 1970s.
Current Issues in Homeowners Insurance
In addition to inadequately priced increased coverage over the years and the recent spate of costly catastrophes, several important issues continue to challenge the future profitability of homeowners. The issues include the following:
- Availability and affordability problems in the Florida homeowners market — After Hurricane Andrew, the Florida Legislature created the Florida Residential Property & Casualty Joint Underwriting Association (RPCJUA, or the JUA) as an insurer of last resort. But the JUA has grown to become the state's second largest homeowners insurer, with more than 900,000 policies, representing roughly $100 billion in exposure. The JUA has operated at a loss since its inception. The growth of the JUA hinders efforts to attract private insurers back to the market, because the JUA bills insurers, based on their market share, for its deficits.
Four years after Hurricane Andrew, Florida homeowners insurers continue to struggle with overconcentration of exposures, and property owners still face insurance availability and affordability problems. The question of how to pay all claims from a major hurricane hitting a sizable metropolitan area remains largely unanswered in Florida, as well as in the rest of the country.
- Availability and affordability of earthquake insurance in California — Since 1985, California state law has required that insurers offer earthquake coverage with every homeowners policy. This requirement, in conjunction with the 1989 Loma Prieta earthquake and the 1994 Northridge earthquake, has led to widely reported availability problems in the California homeowners market, as insurers try to manage their catastrophe risk. The recently created California Earthquake Authority (CEA) is a state agency that provides residential insurance for earthquake damage. All companies must still offer earthquake insurance with every homeowners policy. But a participating company can (and must) shift all of its earthquake policies to the CEA. The CEA makes the California homeowners market more attractive to insurers that might otherwise have curtailed business or left the state. However, based on initial participation levels, the CEA covers only the first $7.5 billion in losses.
- Hawaii Hurricane Relief Fund — Hawaii homeowners insurers are vulnerable to significant losses from earthquakes, volcanic eruptions, and hurricanes. In inflation-adjusted dollars, Hurricane Iniki ranks as the eighth worst catastrophe for U.S. property insurers. After Hurricane Iniki in 1992, many insurers were reported to have temporarily stopped writing property coverage in Hawaii. In response, in 1993, the Hawaii insurance commissioner expanded the underwriting authority of the Hawaii Property Insurance Association, and the Hawaii Legislature established the Hawaii Hurricane Relief Fund (HHRF). The HHRF provides hurricane coverage. Private insurers can participate in the HHRF and continue providing basic property coverages for perils other than hurricane. Nonparticipating insurers must provide hurricane coverage directly to their policyholders. Although participation in the fund is not mandatory, most insurers have chosen to participate. Still, many insurers remain cautious about entering or expanding business in the Hawaii market, despite the relief the HHRF would provide.
- Catastrophe modeling — Because historical experience on catastrophes is sparse, and the population (and, consequently, exposure) has grown substantially in many catastrophe-prone areas, insurers have developed catastrophe models to overcome the deficiencies of relying on historical experience. A catastrophe model is a set of data bases and computer programs designed to simulate the frequency and severity of possible catastrophes using current exposure distributions. To date, state insurance regulators have not broadly embraced catastrophe models for ratemaking. As understanding of catastrophe models increases and such models improve in accuracy, ISO expects that models will become accepted tools for determining the portion of an insurer's rate needed to cover catastrophe losses.
- Tax-free catastrophe reserves — Some insurers and regulators advocate allowing the accumulation of tax-free catastrophe reserves to encourage the accumulation of funds to pay claims in the event of a catastrophe. The advocates of such an approach must overcome significant obstacles, particularly congressional reluctance to reduce tax revenues. A subgroup of the National Association of Insurance Commissioners is developing a proposal for allowing reserves for future catastrophes. The proposal will include consideration of how to obtain federal tax deferment.
- Capital market alternatives for managing catastrophe risk — Since 1989, insurers have faced increased catastrophe risk and limited means of transferring it. Insurers have been seeking alternatives to traditional means for managing this risk. These alternatives include catastrophe options, catastrophe bonds, contingent surplus notes, contingent equity, liability-backed securities, and catastrophe swaps. Insurers were optimistic that these ventures would attract capital markets. But to date, insurers have made few arrangements that engage the capital markets. Current stumbling blocks to capital market investment in catastrophe risk include:
- investors' lack of knowledge about catastrophe insurance
- the present lack of liquidity and low trading volume, which impede establishment of viable secondary markets for the investments
- the complex accounting, rating agency, regulatory, and tax issues involved
- the long time required to structure some deals
Rate provisions for catastrophe losses have two components — loss costs and risk load. The loss cost pays for expected average catastrophe losses. But large catastrophe losses are unusual and require the availability of additional capital or other funding, even if catastrophes don't occur, to support the obligation to pay claims. The risk load component recognizes the need for access to this additional funding by providing the returns to investors. Although past rates have con-tained contingency provisions, some regulators and insurers have not yet fully accepted the concept of risk load as a way to fund large catastrophe losses.
- Urban property issues — Since the 1960s, the availability and affordability of property insurance in the inner cities have been contentious issues. Some consumer groups contend that insurers practice "redlining" or other forms of discrimination against residents of minority and low-income areas. However, unfairly denying coverage to good risks would be counter to insurers' economic interests. Research to date has been inconclusive.
- Environmental issues — Most personal property owners have little or no financial incentive to remove or reduce hazards that may lead to liability claims for environmental impairment. And most homeowners liability coverages contain no restrictions on losses from the discharge of pollutants. But a growing number of insurers have implemented a pollution exclusion or a lead-poisoning exclusion or both for their personal lines policies. ISO has developed a new optional endorsement to address some of insurers' concerns in this area.
- Using credit information in underwriting — For many years, property/casualty underwriters have used credit reports as one of several tools in underwriting commercial lines policies. Recently, some insurers extended the use of credit reports as a competitive tool in underwriting homeowners and other personal risks. Some regulators and consumer advocates have challenged this use of credit reports, maintaining that the practice unfairly discriminates against homeowners in certain geographic areas or against members of protected groups.
Summary and Conclusion
Beginning with Hurricane Hugo in 1989, catastrophes have caused large underwriting losses for homeowners insurance. Much larger catastrophes and resulting insured losses are possible. Policymakers in California and Florida have tried to address the need for insurance to cover a moderately sized catastrophe, but a solution to the problem of a large catastrophe (or a combination of several moderately sized catastrophes) does not yet exist. Until society solves this problem and creates the necessary financial mechanisms, homeowners insurers will remain in a precarious situation, and insurance availability may remain a problem in catastrophe-prone areas.
But catastrophes are not the entire problem. Competition and customer demand for more coverage have generated an ongoing string of enhancements, resulting in prices constantly playing "catch-up" with the forces driving costs. Future profitability for homeowners insurance will hinge on insurers' ability to reflect the true cost of their evolving coverages in the prices of their products.
|
