The Nonparametric DEA Method for Portfolio Constructions in a Non-Islamic Stock Market: The Case of the US

We propose a novel approach to construct integral metrics for selection of stocks in a portfolio based on the nonparametric Data Envelopment Analysis (DEA) method. For the period 2007–2020 and a non-Islamic stock market (the US), we evaluate and compare the return and the return-risk ratios for 110 different stock portfolios. We reveal that debt burden, the Price to Book multiplier, the ratio of the current price to the 52 week high price, and the stable growth of revenue and operating income significantly impact on stock returns of the US companies. The sectoral affiliation and the phase of business cycle are also significant. Our approach allows us to construct portfolios which significantly outperformed the S&P 500 index in 2007–2020 by the total return, Sharpe and Treynor ratios and showed a positive Jensen's alpha.