Value Factors Do Not Forecast Returns for S&P 500 Stocks

Abstract

This paper investigates how effectively value factors can forecast future returns for stocks in the S&P 500.

Ranked portfolios and linear models are constructed from a set of quarterly value factor from 1998 to 2013. Portfolios are drawn from the quarterly S&P 500 stock universe to avoid survivor bias.

Over this time period, with a set of over 400 or more stocks per quarter, the returns from ranked portfolios or forecast by linear models produce at best weak performance compared to the S&P 500 index returns.

These results suggest that for this fifteen year time period, for the large capitalization S&P 500 stocks, the value factors examined here are not useful for constructing portfolios.

Keywords: Portfolio construction, value investing, linear models

JEL Classification: C31, C32, G11