When measuring risk, the primary goal is to ensure that your risk estimates relate to cold hard cash. Money is the resource that matters most when considering risk measures. Fortunately, if you’ve been measuring risk using EL and UL (or VAR) or using some form of standard deviation-based method, you’re nearly home. The best approaches for converting risks into cash are based on these techniques.

Traditionally, expected loss is expressed in terms of financials (dollars). Thus, it is used directly, since it is calculated as a measure of dollars likely to be lost.

For unexpected loss, or standard deviation, the number is usually converted into a measure of capital. This measure of risk-related capital is called economic capital or capital at risk(CAR).

Economic capital can be calculated using the capital multiplier approach or by establishing the confidence interval.

With the capital multiplier approach, you multiply the standard deviation by a number (the capital multiplier). That figure relates to the amount of capital a company requires to support most of the risk events it could encounter. The goal of a capital multiplier is to account for the shape of the distribution of losses and the relative safety factor (risk appetite) that the company is trying to achieve.

The confidence interval method is a bit more technical. It tries to directly measure the amount of risk that a company should cover by establishing a confidence interval for the distribution curve. If you have considered 95 percent of your risks and their potential range of outcomes, then you have a 95 percent confidence interval. Risk managers often think of confidence intervals in terms of the number of standard deviations, so there is a close relationship to the capital multiplier method. They also can use more sophisticated methods to analyze the shape of the distribution and to determine how big of a number is associated with the amount of risk covered by the confidence interval.

The process for determining the confidence interval that is right for your company is similar to that of the capital multiplier. Begin by making an assumption about the type and shape of the distribution curve, or use fancier methods of modeling the exact curve. The ultimate confidence interval is usually related to your desired degree of safety. It can also relate to your target or implied debt rating, whether or not you hold an actual debt rating. Remember that higher confidence intervals relate to higher levels of capital, as well as higher and safer degrees of risk coverage. For most companies these should be set above 99.9 percent—above “junk bond” status.

The best reason for choosing the confidence interval method over the capital multiplier method is that confidence intervals require companies to model the distribution curve itself. This makes it possible to account more precisely for the shape of the curve, and how that can change as the risk profiles change, which, in turn, changes the capital multiplier and the money involved.

There are more methods for computing economic capital. These are generally used by companies as they become more sophisticated in their analysis capabilities. These approaches involve understanding the precise shape of the distribution of risks and using more specific means to relate the absolute amount of capital required to their confidence in withstanding a severe financial loss. However, these approaches require explanations of statistics that are far too advanced to cover in this book. Sources for these models are available in Appendix E.