Robust reward–risk ratio portfolio optimization

Abstract In this paper, we propose robust portfolio optimization models for reward–risk ratios utilizing Omega, semi‐mean absolute deviation ratio, and weighted stable tail adjusted return ratio (STARR). We address the uncertainty in returns on assets by varying them in symmetric bounded intervals. The proposed robust reward–risk ratios preserve linearity in the resulting models, and hence are tractable. However, the robust models involve a sizably voluminous number of constraints, especially when the number of assets and scenarios is large. We employ the cutting plane algorithm to solve the proposed models in a much reduced time efficiently. We evaluate the performance of the robust reward–risk ratio models on the listed stocks of FTSE 100, Nikkei 225, S&P 500, and S&P BSE 500. The robust portfolio optimization models are found to outperform their conventional counterpart models in terms of statistics measured by the standard deviation, value at risk (VaR), conditional value at risk (CVaR), Sharpe ratio, and STARR ratio.