Acceptance of Capital Modeling Grows; Requires Understanding of Use, Limits

July 11, 2008

Economic capital modeling (ECM) is a critical tool in the Enterprise Risk Management (ERM) process, and one that is being increasingly accepted by insurance company senior management, according to industry experts. In addition to the support of senior management throughout the business, effective programs call for a common understanding of what models can and cannot achieve.

Panel moderator Stephen Mildenhall, executive vice president, Aon Re Services Inc. described economic capital modeling as the part of ERM that calls on an actuary’s technical skills in quantifying, aggregating, measuring and monitoring risks. “ECM is the actuary’s problem and a central problem in the overall ERM endeavor,” he said, at the recent 2008 Spring Meeting of the Casualty Actuarial Society in Quebec City, Canada.

Mildenhall observed that the many successes of economic capital modeling have often been unfairly overshadowed by news accounts of shortcomings, such as subprime and credit issues affecting investment portfolios and catastrophe models that did not totally account for a mega-catastrophe like Hurricane Katrina.

On the positive side, members of the panel identified significant success in the acceptance of economic capital modeling in the insurance industry. “We have gotten senior management’s attention and that has helped move the ball forward,” said David Cummings, director – enterprise risk management, State Farm Insurance Cos.

Cummings pointed out that “we have been able to move the phrase ERM from a buzzword” to the point that “people understand that there is some value behind what we are doing.” He noted that economic capital modeling has helped “translate gut feelings into something more concrete.”

John Beckman, senior vice president and chief risk officer, CNA Insurance Cos., said economic capital modeling is increasingly embedded in strategic and operational decision-making.

The models are generating discussion and “articulating the amount of risk you are willing to take,” he said. For example, Beckman observed that models have impacted decisions on reinsurance retentions. Moreover, managements are now using economic capital modeling to decide whether an investment in a line of business will generate an adequate return on capital.

Stephen P. Lowe, managing director, Towers Perrin, said that the models are providing strong analytical support for some tough management decisions, such as withdrawing from markets in catastrophe prone areas, where returns are not as adequate. Most significantly, he said, is the “increased use of economic capital modeling as a platform for measuring economic value creation.”

Cummings said his company is using the models to help integrate different parts of its businesses. To do so, however, required life and property/casualty actuaries to “get on the same page. Actuaries are not uniform in how they approach problems,” said Cummings. Such distinctions required additional work to “build integrated views across diverse lines of business,” he said.

Lowe agreed, stating “getting everyone on the same page is a huge task.” However, he pointed out that the reward is the ability to look at risk – with economic capital modeling as a measure of risk – in the same way. “This is not an easy task; but it is an important task.”

He further pointed out that “embedding economic capital models in the business is a very significant organizational change, introducing new concepts, new language and new discipline.”

Lowe cautioned that actuaries “need to avoid the Icarus complex, not falling into the trap that brilliance is a requirement for these models.” Counseling a practical approach, he stressed the importance of back testing to see how models have performed historically, suggesting that building the underwriting history for back-testing often requires “a bit of an archaeology project.”

Beckman added that some deterministic testing is a way to “validate models and ensure that you don’t get surprised.”

Actuaries need to explain where models are useful and where they might go wrong, according to Cummings. “We want people to use and trust models, but don’t rely on them too much,” he said. “There is room for us to add a little extra caution,” said Cummings. “We need to understand what is modeled well and what is not modeled well.”

Cummings said that economic capital models are a useful input to standard ratemaking methodologies. Capital models provide a great framework to assess different businesses with different risks.

Panelists agreed that there have been some advances in capturing unknown and emerging risks through economic capital models. “We don’t know what they’ll look like, but we know they will occur,” said Lowe.

Beckman said that discussions about the modeling of emerging risks are important. He further noted that it would be a mistake to consider that such a risk would involve a single event impacting one policy. Actuaries and their models must consider that “it’s one event that hits multiple policies and possibly multiple years.”

Source: The Casualty Actuarial Society

Was this article valuable?

Here are more articles you may enjoy.