New Tool Helps P/C Insurers Manage Volatility of Catastrophe Models

A new risk management tool promises to help insurance companies understand wide swings in loss estimates from traditional catastrophe models and better plan for large loss events.

Karen Clark & Co. (KCC), founded by a woman who pioneered in catastrophe models and catastrophe risk management, has launched a new risk management tool that it calls Characteristic Events (CEs). CEs are defined-probability events created for specific peril regions that provide a wealth of information to insurance companies. The events are developed for the return periods of most interest to insurers, such as 1-in-100 and 1-in-250.

Clark said the information garnered through the CE compensates for some of the “inherent limitations” of traditional catastrophe models used by insurers.

“After working with dozens of companies over the past several years, we found the fundamental requirements for effective catastrophe risk management are stable sets of events to represent large loss potential and consistent yardsticks for measuring and monitoring risk,” said Karen Clark, president and CEO, of Boston-based KCC.

“This is what companies are looking for because, while catastrophe models perform many functions and provide a lot of numbers, the models can be problematic for risk management decision-making when the loss estimates swing widely from model to model and update to update,” she said.

Volatility in cat model estimates has been on the minds of P/C executives since modeling from RMS last year released a new hurricane model, version 11, that dramatically changed estimates of potential hurricane losses around the country.

Clark has in the past been critical of what she sees as industry over-reliance upon traditional catastrophe models.

Other cat model executives, including Bill Keogh, president of EQECAT Inc, have also called for better understanding of what the models can and can’t do. AIR Worldwide CEO Ming Lee has said some carriers and rating agencies may be misapplying cat models.

Clark said that by remaining constant year to year, CEs provide “more stable risk metrics for day-to-day underwriting and pricing decisions and monitoring the effectiveness of risk management strategies over time.”

The CEs can be used on their own or as a complement to the models.

How the CE Approach Works

Using the same scientific data underlying the catastrophe models, KCC creates CE intensity “footprints” to represent the event characteristic of the selected return periods and for specific regions. For example, in Florida, the 100-year CE will be representative of a Category 5 hurricane while in the Northeast, the 100-year CE will be representative of a Category 3 storm. These footprints are superimposed on company exposures, and damage functions by occupancy and construction are applied to estimate the losses.

For each exposed region, KCC floats the CE footprints to make sure that all exposed properties are covered and the resulting patterns of risk are smooth and logical, particularly at high resolution. Because the catastrophe models generate random events, the underlying pattern of risk at high resolution can be erratic and anomalous, according to Clark, who says model-generated ZIP code and location level losses are “particularly volatile.”

Clark said much of the volatility in the model-generated loss estimates is due to noise from uncertainty and changing assumptions in the hazard component of the models. She said CEs eliminate this volatility by keeping the events constant. The CE footprints change “only if there is tangible scientific evidence that significantly alters the consensus view of risk in a particular peril region.”

Because CEs remain constant and are the same across companies, they provide a “common currency” for comparing and monitoring catastrophe risk, according to KCC, which also says because they are transparent, the CE footprints can be peer reviewed by internal and external experts.

How Companies are Using CEs

Insurers can use CEs to accomplish multiple objectives, according to KCC:

• CEs provide a transparent view of the risk so companies can better understand their exposures, and they are model-independent scientific benchmarks for testing the model loss estimates. By comparing the CE estimates to the model-generated estimates, insurance companies can determine if any models are outliers for their books of business and which models are more credible in various peril regions. In addition, the model estimates can be compared to those of the CEs by region to determine appropriate model-blending weights.
• Rating agencies and regulators have growing expectations with respect to how well companies understand and can explain their catastrophe loss potential. KCC said CEs provide additional information companies can use to support their risk management decisions to external stakeholders as well as for effective internal communication.
• Also, CEs can be used on their own or in conjunction with the models as yardsticks for monitoring risk and the effectiveness of risk mitigation strategies. CE losses can be monitored at the corporate level and drilled down to individual business units, lines of business, policies and locations.

Clark said the industry response to CEs has been “overwhelmingly positive because companies now have an additional scientific method for assessing and managing catastrophe risk which is healthy for a competitive market.”

“Catastrophe losses now dominate many of the property lines, and will continue to grow, so the more credible scientific information that can be brought to bear on risk management decisions, the better,” she said.

Currently available for U.S. hurricane, future versions of CEs will address earthquakes and other perils, according to KCC.

Topics Catastrophe Carriers Profit Loss Hurricane Property Casualty Risk Management