The market risk premium is calculated by subtracting the expected long-term risk-free rate from the expected market return. These figures should model future market conditions closely. There are two approaches:
“Historical” or “ex-post” risk premia approach, which claims that past market returns are the best estimator of future market returns. See Historical (Ex-Post) Risk Premium.
“Forecast” or “ex-ante” risk premia approach, which claims that current market information can be used to improve the accuracy of historically based estimates. See Forecast (Ex-Ante) Risk Premium.
The historical approach relies on the assumption that the market risk premium is basically stable over time. It uses an arithmetic average of past risk premia to estimate future risk premia. Because it relies on actual historical information, this method can be considered an objective measure of the long-term expected market risk premium.
However, those who use this method must decide subjectively how many historical periods to use in the average. Some people believe that using the longest available data period is most objective. Since market statistics have been monitored since 1926, this period is from 1926 until today. Other people select milestones such as World War II, on the assumption that the risk premium is more stable since that time.
Other financial professionals believe that information besides historical data can be useful in predicting future market risk premia. They believe that there may have been structural changes in investment markets that affect the market risk premium and therefore historical estimates should be modified by or replaced altogether with, present expectations of future market conditions. This approach is called “forecast,”“ex-ante” or “future” risk premium determination.
To calculate a forecast risk premium, a forecasted risk-free rate is subtracted from a forecasted market return. The current yield curve is a valuable source of information about forecasted risk-free rates. It is composed of the current yields to maturity of risk-free bonds of various maturities. Because future rates can be “locked in” today and realized later, many people believe that these rates offer accurate estimates of future rates. Therefore, they use these rates as a proxy for future risk-free rates in calculating forecast risk premia.
There is much less agreement on how to forecast future market returns. In fact, the main problem with the forecast approach is that it requires a great deal of subjective judgment by the person doing the calculation. Which forecast estimates for the expected market return should be used? Should historical information be used at all? If so, what time period or periods is used and how should they be weighted with forecast estimates?
Methods of forecasting future market conditions are as varied as the assumptions on which they are based. A desirable forecast risk premium takes full advantage of the information currently available in the yield curve, includes structural changes in the risk premium, but involves a minimum amount of subjective judgment.