## Relative Risk Measurement

Problems sometimes arise when standard deviation is used to measure risk. If an investment project is relatively expensive and has large expected cash flows, it will have a large standard deviation of returns without being truly riskier than a smaller project. Suppose a project has an expected return of \$1 million and a standard deviation of only \$1,000. Some might reasonably argue that it is less risky than an alternative investment project with expected returns of \$1,000 and a standard deviation of \$900. The absolute risk of the first project is greater; the risk of the second project is much larger relative to the expected payoff. Relative risk is the variation in possible returns compared with the expected payoff amount.

A popular method for determining relative risk is to calculate the coefficient of variation. Using probability concepts, the coefficient of variation is

Coefficient of Variation = v =

In general, when comparing decision alternatives with costs and benefits that are not of approximately equal size, the coefficient of variation measures relative risk better than does the standard deviation. 