On the Relative Accuracy of Discounting Based on Risk-Free and Risky Portfolios

The degree of risk that should be incorporated into the net discount rate that is used to estimate the present value of future lost earnings has been the subject of controversy. While some forensic economists insist that a risk-free discount rate must be used, others have offered economic arguments that support use of a risk-adjusted rate. Historical simulation studies have found that, when the discount rate is based on risk-free or low-risk securities, the historical averages method of estimating present value is subject to large forecast errors due to significant changes in net discount rates over time. This study explores whether basing the discount rate on mixed portfolios of equities, intermediate-term government bonds, and Treasury bills might result in more accurate estimation. Using the historical averages method with data covering the period 1926–2008, results are generated for four mixed portfolios of varying degrees of risk, and these results are compared to the results obtained with Treasury bills, intermediate-term government bonds and long-term corporate bonds. The historical simulations do show that the mixed portfolios often provide more accurate estimates. These results should be of considerable interest to forensic economists who believe that some degree of risk should be incorporated into the discount rate.