CHICAGO--Financial market turmoil of the past 18 months and its effect on companies of all types has left many wondering why risk models failed to predict the various events.

But, according to some enterprise risk management experts gathered recently at the 2009 Enterprise Risk Management Symposium in Chicago, the models did function accurately in many cases. The problem, they said, sometimes stemmed from a failure to accept the results the models produced.

A more complex model isn't necessarily a better model, said Anthony Dardis, head of U.S. insurance at risk model provider Barrie & Hibbert Inc. in New York.

Speaking as part of a Modeling Lessons Learned session, Mr. Dardis said, "Good models do exist already." There are models that did very well in gauging exposures during financial markets events during the past year, but the bigger issue in many cases was executives' unwillingness to accept the results of the models, he said.

One can't do modeling in isolation from consideration of human factors, Mr. Dardis said.

"Modeling human behavior is certainly one thing, but then there's what do humans do with the results of models once they've got them," he said. While good models are available, there probably have been cases in which executives were uncomfortable acting on the results showing possibilities they'd never seen before, he said.

Daniel Rodriguez, chief risk officer in the Global Proprietary Trading Group at Credit Suisse Group A.G. in New York, said one failing comes from an assumption by users of models.

"One of the big failings...is that a lot of people make the assumption that the model continues to be applicable after a big change in the marketplace," he said. "You have to be a modeler and a hunter, continuously scanning the environment for things that aren't incorporated in your model," Mr. Rodriguez said. "You have to continue to adapt to your environment."

"I think the key point is continuous monitoring and continuous reassessment of the assumptions that we make," added Leo Tilman, president of strategic advisory firm L.M. Tilman & Co. Inc. in New York.

Sameer Vittal, senior technologist, advanced technology operations at General Electric Co. unit GE Energy in Greenville, S.C., said his group focuses on modeling the causes of events.

"Correlation does not imply causality," Mr. Vittal said. "There's a huge difference." His unit often goes back after an event to try to determine why the system behaved as it did and often find "it was sending off signals well in advance," he said.

Mr. Rodriguez noted that in terms of "applying the brakes" to a business activity based on model results or actual events, the brakes shouldn't be saved only for the negative.

"You should also have an upside stop," he said. "The upside stop says, 'Your strategy is up three times more than you expected. Why are you being so successful?' Anything out of the ordinary that you can't reconcile your model with, upside or downside" should be examined carefully.

Mr. Vittal said it's necessary that the person with the information generated by the models has adequate "clout" to drive necessary changes in the organization.

Mr. Dardis said it's important that top management have an understanding of the models that a company uses.

"I think certainly from a senior management perspective there needs to be a very clear understanding of what are the questions that my models are trying to answer and what are the implications for my decisions," Mr. Dardis said.

"Senior management clearly has to have a clear understanding of the complexities and the interrelationships of assets and liabilities," he said. "I think there does need to be some pretty deep understanding at board level as well."

Asked where ERM should fit in an organization, Mr. Tilman suggested that ERM should be part of a company's strategic decision-making, although it often is not. "I think every one of us in the room can relate to the difficulty of connecting risk management and strategic decisions," he said.

For example, he said, if you ask the chief executive officer about a planned acquisition, he or she will talk about business synergies or expanding the company's footprint. But if you ask how much risk the transaction will add to the company's balance sheet, he or she typically will find it difficult to answer.

Thomas Hettinger, managing director of actuarial consultant EMB America L.L.C. in San Diego, moderated the panel.