Insurers Begin Modeling Casualty Catastrophes

Scenario-based modeling has been used for property risk for some time, but its application in the casualty area is recent, according to Robin Wilkinson, vice president and managing director of Casualty Analytics at AIR Worldwide, a Verisk business. In an interview with Claims Journal, she explains how modeling systematic and sudden trigger casualty occurrences can benefit insurers and reinsurers in many ways.

Liability catastrophe modeling can be applied to a wide variety of business lines.

“Casualty risk modeling tends to be focused on casualty catastrophes and they have the potential to impact almost all lines of business,” said Wilkinson.

Some lines that could be impacted include general and product liability, professional and directors’ and officers’ liability, business income, workers’ compensation and employer practices liability.

The events themselves, she said, can be systemic events or sudden trigger events.

Systematic events occur when there are many claims against a number of policies in one industry, like in the case of asbestos or sports concussions.

Classic clash events center on a single company resulting from a sudden trigger event, like the Deepwater Horizon explosion that impacted a number of liability lines, she said.

“Such as pollution, general liability, directors’ and officers’, obviously a little bit of property. Mostly liability, and losses exceeding actually in the case of $100 billion,” said Wilkinson.

While there is no standard model used for modeling liability catastrophes, she explained the scenario-based model shows the most promise because it’s not completely dependent on historical data.

“Historical data on casualty events may be scarce or not wholly predictive of future events,” said Wilkinson.

Though scenario-based modeling has been around for 20 years, it’s mainly been used in modeling property exposures. Insurers are now showing interest in modeling liability scenarios in order to improve portfolio data.

A scenario-based model allows for expert input and is transparent, giving the “user the ability to see and also stress test the inputs and assumptions that lead to their results.”

She shared the basic steps to modeling a liability catastrophe:

1. Generate a footprint of the impacted industry and lines of business.
2. Determine the economic parameters associated with the event.
3. Overlay a portfolio to understand how much the event will likely be insured within the portfolio.
4. Simulate the event several times to get a reasonable range of potential outcomes for the portfolio.

Insurers can benefit from casualty cat modeling in a number of ways, Wilkinson said. For example, it providers an insurer a clear picture of its concentrations of business exposures and the types of events that might have an adverse impact on those businesses.

“It provides an insurer a way to stress test a particular portfolio and that, in turn, helps to assess capital adequacy,” said Wilkinson.

The model can also assist insurers in considering new business when adequately priced, based on liability catastrophe modeling results.