Least Squares Predictions and Mean-Variance Analysis

We compare the Sharpe rations of investment funds which combine one riskless and one risky asset following: i) timing strategies which forecast excess returns using simple regressions; ii) a strategy which uses multiple regression instead; and iii) a passive allocation which combines the funds in i) with constant weightings. We show that iii) dominates i) and ii), as it implicitly uses the linear forecasting rule that maximizes the Sharpe ratio of actively traded portfolios, but the relative ranking of i) and ii) is generally unclear. We also discuss under what circumstances the performance of ii) and iii) coincides.