The current financial crisis did not result from a failure of enterprise risk management (ERM), but a failure to implement ERM processes, according to actuaries.
According to Parr Schoolman, senior vice president for Aon Benfield, an over-reliance on optimistic models and the people who built them have been blamed for some of the financial meltdown. He said one lesson for modelers is that even if models accurately capture the severity of a risk, an incorrect estimate of the probability can still lead to the wrong conclusion.
Schoolman was among the panelists for “Prudent Enterprise Risk Management: What is Learned by the Crisis?” at the Casualty Actuarial Society’s 2009 Spring Meeting.
The Aon Benfield actuary recommended that users of models recognize that they are not the real world and results of complicated models can be driven by one or two key assumptions. “For subprime mortgages and mortgage backed securities, the complexity of the models hid how much the analyses were relying upon continued home price appreciation,” Schoolman said.
Stress testing key assumptions, especially the correlation and likelihood of tail events, is important, as is acknowledging the limit of actuaries’ ability to parameterize a model from a limited data sample and its corresponding implications for model precision, he said.
Neil Bodoff, senior vice president of Willis Re, Inc., said that before lessons can be learned from an event, “we must be acutely aware of potential pitfalls and biases in our analysis,” which he listed as the fallacies of hindsight and myopia.
“Hindsight is the classic ‘Monday Morning Quarterback’ problem,” he said. “After the fact, we’re all geniuses with perfect knowledge of all that unfolded, but at the time things happened, in the event or in the moment, we don’t know exactly how things will play out.” He said the solution is to focus on what can be done to “improve our chances, before the blowup.”
The myopia, or shortsightedness, is the lack of seeing the broader landscape and instead focusing on the idiosyncratic aspects of the particular crisis, Bodoff explained.
The solution is to focus on those aspects of the crisis that have broad application, according to Bodoff.
The Willis Re executive said it’s important to include the cost of downside risk when measuring profit for financial statements and when measuring profit for compensation purposes. By doing this, incentives are generated internally to consider profit on a risk adjusted basis and those incentives can shape behavior of risk takers before a blowup.
Applying these lessons to property/casualty (p/c) insurance, Bodoff said modelers can look at whether p/c accounting reflects the cost of risk when measuring profits and whether p/c insurance enterprises include the cost of downside risk when calculating profit for compensation purposes.
“I believe many of the aspects of the current financial crisis are relevant to p/c insurance, as well,” he said.
Michael Wacek, executive vice president and chief risk officer of OdysseyRe Holdings Corp., identified five lessons: control leverage; anticipate that markets are not always liquid; maintain adequate liquidity; don’t totally outsource credit analysis to rating agencies, and understand the complexity of the business.
Wacek pointed out that many banks had asset leverage of 30:1 or more, and that while insurers typically had less asset leverage, their liabilities are risky, too. He said the “AIG predicament illustrates the danger of uncontrolled derivative exposure, which is not counted in the asset leverage ratio,” an issue that also affects banks.
Wacek said that ratings had proved to be a poor indicator of credit quality and he recommended that companies do their own credit analysis or limit credit exposure, even to the highest rated counterparties, and that they diversify intelligently.
“Some financial institutions apparently became too complex for their risk management framework,” he noted. “Enterprise risk management must be embedded in the organization, but is especially critical in the executive suite.”
Source: The Casualty Actuarial Society