In a classic paper, Dulaney (1987) proposes a historical simulation- based method for evaluating measures of the present value of future earnings. This method compares a given ex ante estimate of present value with an ex post simulated value, in each time period. A key issue is the interpretation of what it means to have a good fit, when matching historical simulated present values. With best fit defined in standard statistical terms, I find that the total offset approach - whereby projected growth in wages is assumed equal to the projected interest rate - works best in the examples considered here and in Dulaney (1987). This finding violates convention as most forensic economists implicitly allow a gap between projected wage growth and the interest rate, when estimating present value. It does, however, jibe with the absence of a statistically significant long-run gap between U.S. annual wage growth and the interest rate.