Tackling the Real Wind Peril
By Stephen Clarke and Joseph Palmer
Historically, fire has been a leading cause of commercial property damage losses. Along with fire, hurricanes and tornadoes — specifically the wind peril — are a growing concern. The active 2005 hurricane season brought that sharply into focus, with 27 named storms and 14 hurricanes (4 of which reached Category 5 intensity, including Wilma, the strongest Atlantic hurricane on record).
Damage due to wind has been growing rapidly as a percent of overall commercial property losses. Tremendous increases in coastal populations have fueled much of that growth. In Harris County, Texas, the population today is approaching 4 million; while in 1930, it was less than 400,000. From North Carolina to Florida, the population has grown exponentially over the last 30 to 40 years. With that growth come more houses, more offices, more factories, and more stores. Today, roughly 38 percent of the total U.S. insured exposure value is in coastal counties. In New York, 62 percent of insured value is along the coast; in Florida, that number skyrockets to almost 80 percent.
Eighteen of the top 20 catastrophic events, as identified by ISO's Property Claim Services®, involve wind (15 hurricanes and 3 tornadoes). Only the Northridge earthquake and the terrorist attacks of 9/11 did not involve the wind peril. Of the 18 wind events, 13 occurred during the 2000s — when property values and population densities were peaking.
In 1926, a devastating hurricane struck just south of Miami. That Category 4 storm caused widespread death and destruction, with an estimated $100 million in property damage. With increased population and development in the region, that same storm today would result in insurable losses easily exceeding $100 billion, as estimated by AIR Worldwide.
The concentration of catastrophic wind events since 2000 is largely due to increased hurricane activity coupled with the growth in coastal properties. Long-term change in Atlantic sea-surface temperatures may be a reason for increased hurricane activity and intensity. The Pacific oscillation known as El Niño or global warming in general might be the cause. Whatever the reason, the "perfect storm" of population growth and building expansion in coastal counties combined with an increase in severe wind events has driven wind losses to historic proportions.
Hurricane Andrew, which struck south Florida in 1992 and caused almost $17 billion in insured losses, was a wake-up call for the industry. Homeowners and businesses filed more than 650,000 claims with their insurers. In the aftermath, 10 small insurers were left insolvent; 34 insurers announced plans to withdraw from the market; and 29 others restricted or reduced their writings. One major insurer announced that in 50 years of writing Florida policies, it had not collected enough premiums to pay for that one event.
In response, the industry looked for ways to foster the availability and affordability of coverage in light of the magnitude of the exposure. Coverage options were developed to spread the risk more evenly between insurer and policyholder. Financial markets were leveraged as an alternative source of capital through the development and sale of catastrophe bonds. The importance of effective (and enforced) building codes was recognized as we developed the Building Code Effectiveness Grading Schedule (BCEGS®). Notably, Hurricane Andrew spurred the development and use of catastrophe models, led by the efforts of AIR Worldwide, to evaluate the risk of catastrophic loss. But people continued to flock to the coasts, further expanding the growing exposure to loss.
Rating the wind exposure has historically been a fairly straightforward exercise. Because fire had been the predominant peril, the classification of buildings for underwriting purposes generally focused on building characteristics related to susceptibility to fire — primarily the building's construction. The wind peril was generally class-rated and used the building's fire classification, with only minor modifications to reflect a roof's uplift characteristics and height. When wind was a relatively small proportion of overall property losses, that approach made sense. But in certain areas of the country, wind has become an increasingly significant loss exposure. So, evaluating the resilience of a building with respect to the wind peril — separate from the fire peril — is becoming necessary.
Not all building characteristics that are beneficial from a fire perspective are also beneficial from a wind perspective. Light metal construction may be fire-resistive but can sustain significant structural damage during a windstorm. Conversely, a wood frame building with appropriate roof straps and tie-downs will perform well in a windstorm but not in a fire. The challenge is to formulate a property rating algorithm that can accommodate the differences between the fire and wind exposures.
We are embarking on a program to do just that — the Enhanced Wind Rating Program. The new rating approach will identify, evaluate, and measure the specific wind exposure of eligible commercial buildings in areas subject to significant wind losses. The key exposed areas are along the coast, where the hurricane exposure is significant, and the central part of the country, where tornadoes are the key threat. Using on-site inspections, the program will gather information relevant to the wind exposure for each eligible building, encompassing such areas as location and surrounding environment, roof and wall envelope, framework, building codes, and temporal variance.
We are applying an engineering-based, actuarially sound approach to evaluating individual building characteristics by leveraging our vast database of commercial property exposure information and the expertise inherent in models developed by AIR Worldwide covering hurricane and severe thunderstorm. The program will not only deliver a specific building loss cost particular to that building's wind-related characteristics, but it will also provide critical data for use during the underwriting and loss modeling process. The data collected can also be used with catastrophe models to manage books of business across geographic regions, evaluate aggregations, and even directly enhance model results.
We will begin collecting specific information on eligible buildings shortly and will start to make underwriting reports available later this year. Plans are to file the rating schedule and related manual rules for commercial property by year-end, with a 12-month lead time to implementation.
*ISO's Property Claim Services® (PCS®) unit provides property/casualty insurers and reinsurers with catastrophe loss information — estimates of anticipated industrywide insured losses arising from catastrophes.
Stephen C. Clarke, CPCU, is assistant vice president of commercial multi-lines at ISO. Joseph M. Palmer, FCAS, MAAA, CPCU, ARM, is assistant vice president and actuary of commercial multi-line actuarial at ISO.
Also contributing to this article: Kevin Kuntz, assistant vice president of risk decision services at ISO, and Peter Bingenheimer, assistant vice president at AIR Worldwide.