Portfolio management strategy using predefined stock filters

Abstract

Capital market efficiency suggests that security values at any given point of time always reflect all available market information about the security in question and thus serve as a true indicator of underlying value. While there have been several empirical studies done to study and prove that capital markets are efficient, several still question its validity. There is a category of investors who adopt a value investment strategy and invest in stocks that they believe do not completely capture their true worth and hence are undervalued. The primary means of picking undervalued stocks are on the basis of a low P/E or P/B ratio. There is a large body of literature [1, 2, 3, 4, 5] that provides empirical results to suggest that stocks with low P/E or P/B ratios generate superior returns in the long run. These strategies are typically contrarian in nature as they usually involve buying stocks that the rest of the market is selling, and vice versa. In this report, in addition to P/B and P/E ratios, we look at a few more alternative measures and characterize their performance in being able to consistently pick stocks that deliver consistent superior returns. We also take a portfolio based approach [6] as opposed to a stock based approach, and strive to build portfolios based on the underlying properties of the stocks in the consideration set. In section II below, we cover the data set, the underlying stock parameters considered and the methodology of the analysis. In section III, we present the empirical results obtained based on our analysis. In section IV, we showcase a hybrid portfolio built out a combination of filters, and demonstrate its ability to consistently generate above-average returns. We conclude with a summary and implications of our results in section V.