Catastrophe Models and Property Insurance

The insurance industry has long been concerned with the impact of catastrophes (CAT) on their financial stability.  The reinsurance market exists to smooth our results when a catastrophe strikes to maintain carrier solvency and market stability. Twenty-five years ago, the CAT concerns were hurricane and earthquakes. These exposures were geographically centered and were managed by monitoring the values accumulated in given areas. Managing and reporting your exposure accurately to the reinsurance market was the best practice.

CAT Models and Insurance

CAT management has changed with the emergence of new catastrophe exposures. Convective storms that develop high winds, tornados, ice storms and hail are common in the central United States. This was formerly a safe haven from coastal CAT exposures. Wildfire has emerged as a CAT hazard that is present in every region. Floods are increasing in frequency and severity with the changing weather patterns and continued development along waterways. These changes require more management that simply adding up values in a given region. 

The reinsurance markets pioneered the use of CAT models. These would locate exposures using mapping technology and using mathematical models based on past history, building construction and damage patterns, they could develop an estimate of the cost of a given event. Computers allowed Monte Carlo simulations to model thousands of event variables to arrive at a predicted outcome. The number produced was a damage estimate and based on the predicted frequency of events, a needed premium to support those losses. This tool changed how reinsurance was priced. It also placed a new burden on insurance companies to have their exposure data configured to support the modeling done by their reinsurance company.  For the first decade of use, the models were a tool for the reinsurance markets.

Once the primary insurance companies saw the potential benefit of these models, they moved from reinsurance to primary insurance markets. The CAT model developers added new models for wind, hail, wildfire, flood and convective storms to the hurricane and earthquake tool set. The age of digital exposure monitoring and modeling arrived. It is a fundamental pricing tool for your property insurance. If you do not embrace this technology, you are left out of the decision-making process.

Data Becomes King

This tool is based on data. With inadequate or missing data, the systems will fill in the blanks using averages. The averages are always slanted towards the most pessimistic outcome. If your CAT model results are high, the estimated loss will be within the normal Cat exposure loss dollars. This is a safety net for the insurance industry. The lower the quality of the data, the higher the margins for the insurer.

Our role in this process is data management and data quality. The first variable for the CAT models is placing your property on a map via geocoding. If your address does not accurately map, the model will place it in the middle of the next available data set. Often a zip code, county or city boundary.  This is a significant variable in exposure. A county level geocode that contains significant rural areas will have a higher wildfire loss estimate than an urban area. The quality of geocoding will impact this model result.  A location without a zip code can geocode to a similar address that is far away from the intended location. An example of this was an inaccurate flood hazed determination that affected the placement of property insurance.

The second major factor is the construction class, number of stories and square footage of the building. These factors are used when predicting the amount of damage that a structure suffers in a modeled event. This data is derived from historical claim information from past events. These factors are often incorrectly reported on insurance applications.  Third party data providers such as tax assessors may also have flawed data. Bad data is often detrimental to your pricing.

Another key element is the age of the building and updates to the building shell such as windows, roof and siding. The effective age of your building shell reflects the building codes that were in effect in your geographic area. As building codes evolve after CAT events to reduce the impact of the next event, newer construction is expected to perform better. Building age and updates are often not accurately recorded.  A post hurricane Andrew study indicated that proper building code enforcement would have reduced the losses by 50%. This reinforced the need for accurate effective code dates for properties. This is not a variable easily accessed by third party data providers. It is a product of careful questions and reviews.

The Driehaus Difference

We have experience in running CAT models within insurance companies. We have developed internal tools to ensure that our data will properly geocode, has all of the needed elements and our surveys and interviews focus on accuracy of this data. Our data can be easily imported into models and allows a faster turnaround of data for quotes. We can get quotes on large property schedules faster because our data is designed to work with a CAT model environment. Leverage our knowledge to your benefit. Call us ta 513-977-6860 or reach us at www.driehausins.com to get us involved in your insurance program. We want to be your insurance provider.