Insurance executives probably don’t relish discussing how they use catastrophe modeling in property insurance ratemaking.
Not only is the topic obscure and technical but, to the extent it is discussed publicly at all, the discussion usually proceeds on the premise that “cat modeling” benefits insurers at the expense of insureds.
Catastrophe modeling has been criticized by regulators, legislators, and consumer activists as an opaque process used by carriers to increase premiums on people due to conditions over which they have little or no control.
Among other things, critics of modeling question how processes supposedly grounded in science and mathematics could arrive at varying estimates of insured losses.
Well, there’s good reason for that, and there’s no reason for insurers to be defensive about catastrophe modeling, now that it has become integral to property insurance ratemaking.
Loss costs development at AAIS demonstrates how the incorporation of catastrophe modeling into ratemaking can benefit consumers, particularly those in distressed coastal markets.
(NOTE: While AAIS develops and files advisory rating information, companies are not required to adhere to it. Companies can and do deviate from AAIS rating information and manual rules, subject to regulatory approval, where applicable.)
AAIS’s most recent Homeowners manuals provide loss costs that incorporate modeled data derived from WORLDCATenterprise, software developed by EQECAT, Oakland, Calif.
EQECAT, a subsidiary of Houston-based ABS Group, is among the world’s first and leading providers of cat modeling applications for estimating the amount of damage likely to result if natural disasters strike certain areas.
Catastrophe models incorporate huge amounts of data related to natural hazards (fault lines, storm tracks, etc.), other natural
conditions (altitude, soil density, etc.), and building characteristics (masonry, frame, etc.) that affect the frequency and severity of disaster losses. The data is utilized by built-in mathematical algorithms to estimate potential losses for a certain risk or collection of risks.
Traditional loss data is still a critical factor in loss cost development, but historical data often does not reflect increased construction and other characteristics of an area that may have changed considerably since its last major disaster.
Modeling fills in the gaps left by traditional loss data and allows insurers to differentiate more precisely among risks with greater or lesser exposure to loss.
To date, AAIS has utilized WORLDCATenterprise components for three types of events: earthquakes, hurricanes, and wind and hail caused by tornadoes.
The most visible impact of the use of modeled data at AAIS has been a pronounced increase in the number and range of territorial loss costs, according to Greg Jaynes, AAIS chief actuary.
Across the country, he says, the number of AAIS Homeowners territorial “price points” has more than tripled, with more states still left to be developed and filed.
A “price point” refers to the loss cost for a particular policy (such as an HO 0003), for a particular amount of insurance (say, a $100,000 Cov. A limit), with a particular deductible (say $500) in a particular territory, with a particular type of construction (e.g., frame).
In the latest AAIS Homeowners manual, there are nearly 500 territorial loss costs for each such classification of risk, with more being developed.
In coastal areas, the modeled data also allows for more precise determination of separate deductibles for wind and hail loss, and of premium credits for wind loss mitigation efforts by policyholders.
“Perhaps most importantly for consumers in coastal areas,” says Jaynes, “is that the model helps us develop more deductible options our member companies can offer insureds.
Deductibles are the most important thing consumers have to manage their exposure to catastrophes.”
Territories
To see how this plays out, consider two newly created Homeowners rating territories in a coastal state. Territory 401 is along the Atlantic coast, where it is exposed to Atlantic hurricanes. Territory 409 is in the opposite corner of the state, away from the coast with relatively small windstorm exposure.
Using the EQECAT model, AAIS estimates that 35.2% of the losses to a property in the coastal Territory 401 will result from the perils of wind and hail. For inland Territory 409, AAIS estimates that only 8.3% of the losses suffered by an insured property will result from wind and hail.
With the ability to price wind and hail losses separately from losses resulting from other perils, and with the ability to assess separate wind and hail deductibles, a policyholder in coastal Territory 401 can get more benefit from his or her choice of a separate wind/hail deductible.
To illustrate, consider the options for hypothetical owners of two properties. One property is in coastal Territory 401 and the other is in inland Territory 409; each is insured for a $100,000 Cov. A limit under an AAIS
HO 0003 (open perils on structures; named perils on contents).
Suppose the owners take a very common approach and select a wind/hail deductible equal to 1% of the Cov. A limit, plus a $500 deductible for losses arising from all other perils.
Since base loss costs are calculated with the assumption of a $500 deductible, there is no change in the loss costs for the non-wind perils; the deductible relativity is 1.00.
The selection of a separate 1% wind/hail deductible has a considerable impact on the projected wind/hail loss costs, however. A wind/hail deductible at that level is projected to reduce the wind/hail loss cost by 13% in the inland territory, and by 18% in the coastal territory. Therefore, the wind/hail loss relativities are 0.87 and 0.82, respectively.
By selecting the $500 non-wind and the 1% wind/hail deductibles, the property owner in
inland Territory 409 realizes an overall deductible relativity of 0.99. That’s because he has chosen the base non-wind deductible, and the wind-hail deductible has little impact on his overall rate, because his wind exposure is relatively small.
The property owner in coastal Territory 401 sees a much bigger impact from the same selections, however. His overall deductible relativity is 0.94--which amounts to a 6% savings on premium directly tied to loss costs.
The impact grows even greater for insureds that select even higher deductibles.
Look at what happens if the owners of our two properties select deductibles of $3,000 for non-wind losses and 5% of the Cov. A limit.
For both policyholders, the $3,000 non-wind deductible is projected to reduce their non-wind loss costs by 27%, providing a 0.73 relativity to the base loss costs for that portion of premium.
In addition, the 5% wind/hail deductible projects a 46% reduction in wind-related loss costs for the inland property (which translates into a premium relativity of 0.54) and a huge 69% reduction in wind-related loss costs for the property in the coastal territory.
In all, the selection of higher deductibles, and the weighting of deductibles made possible with modeled data, produces savings on the loss cost portion of premium of 29% (a relativity of 0.71) for the inland property and 42% (a 0.58 relativity) for the property in the coastal territory.
The equations for these calculations are shown below.
Projecting this out for numerous territories and deductible options, one can seen how AAIS is expanding the number of price options available to insurers and insureds.
That would not be possible without catastrophe modeling.
.
Price Points 
