Optimal Strategy Design for Portfolio Selection: an Inverse Risk Weighting Analysis

This article analyzes the behavior of the portfolio selection strategy that assigns to each asset a weight inversely proportional to individual risk (PIR) in comparison with the classical mean-variance (MV), minimum variance (MINVAR) and 1/N strategies. In doing so and applied to the Colombian stock market, this study performs out-of-sample estimates and provides conditions under which PIR weights lead to less riskier strategies than the 1/N strategy. In conclusion, the evidence suggests that the PIR strategy outperforms classical strategies in terms of profitability indicators, risk, Sharpe ratio, Turnover (cost) and Turnover (stability).