A model defense.

Risk modeling is the key to successful risk management. Not quite. Not even close.

It has been headline news for most of this year, though.

"Risk models got it wrong," the headlines said. Billions were written off or lost. Chief Risk Officers were fired and some seape-goated. By quick, yet unfounded extension, enterprise risk management has failed, just like so many pundits predicted.

Wrong again. Like every other component of risk frameworks, no one part is key to the whole. In fact, each part is important to a successful approach to understanding and successfully managing risk.

But what about risk modeling? No one model earl be expected to give just the "right" answer.

More often, multiple and varied data points are more useful to an effective analysis. Think traditional actuarial work in casualty insurance. Most good actuaries use multiple methods or models to get to their range of predicted values.

So it is in modeling all kinds of risks. More than one model gets you to a better answer most of the time. But a better analysis doesn't stop there. Supplemental approaches can often add a lot. One such approach is the use of expert opinion.

This technique leverages the experience and knowledge of people involved directly in the area of risk or inquiry. This is being used in areas such as terrorism where there is little or no loss experience to draw on.

As the author of "Blink", Malcolm Gladwell argues, "a powerful process in all of us is working subconsciously to sort through huge amounts of information gathered over a lifetime, make associations between data and extract key indicators to arrive at rapid, highly accurate conclusions."

Another technique is a scenario analysis which examines a range of possibilities. These might take the form of an expected loss scenario, an extreme or tail loss scenario and one or more other vital scenarios. It is designed to allow improved decisionmaking by allowing more complete consideration of outcomes and their implications.

Yet another way to look at and analyze risk is using influence diagrams as an alternative to more traditional decision trees. This approach peels the onion back on the inter-relationships between risks and the causes and effects surrounding them. The more deeply you dive on these inter-connections, the more you can understand the likelihood and impacts so central to valuing risk and determining which risk matters more than the next.

Risk modeling has gotten a bad rap of late. It doesn't deserve it since it's critical to an effective risk analysis. However, it's most effectively used when it's used in combination with robust alternatives to traditional analytical approaches. But you can't stop there. The more frequent reason risk models have taken it on the chin is management decisionmaking that ignores or minimizes the results of these analyses and the recommendations of many solid risk managers.

All the analysis in the world is worthless if management mishandles it. In some cases, it's down to a moral hazard that puts self interest ahead of serving shareholders. So a risk culture that promotes adherence to a corporate value system of honesty and integrity is critical to effective risk management. Stakeholders expect this and risk managers can't be effective without it. Analyze all you want but don't discount the value and impact of a solid, well understood and communicated risk culture to drive analysis that can make a difference.

CHRIS MANDEL is the enterprise risk manager for a leading financial institution and a former president of the Risk and Insurance Management Society.