Trading Volume and Stock Investments

Previous studies suggest that trading-volume measures may proxy for a number of factors, including liquidity, momentum, and information. For relatively illiquid (typically smaller) stocks, investors may demand a liquidity premium, which can result in a negative relationship between trading volume (as a proxy for liquidity) and stock returns. For relatively liquid (typically larger) stocks—the focus of this article—momentum and information effects may dominate and result in a positive relationship between trading volume and stock returns. Portfolios of S&P 500 Index and large-capitalization stocks sorted on higher trading volume and turnover tend to have higher subsequent returns (holding periods of 1–12 months) than those with lower trading volume.Previous studies suggest that trading-volume measures may proxy for a number of factors, including liquidity, momentum, and information. For relatively illiquid (typically smaller) stocks, investors may demand a liquidity premium that results in a negative relationship between trading volume (as a proxy for liquidity) and stock returns. But for relatively liquid (typically larger) stocks—the focus of our study—momentum and information effects may dominate and result in a positive relationship between trading volume and stock returns.We back test simulations of historical data of the S&P 500 Index and the largest 1,000 U.S. stocks as measured by market capitalization (Largest 1,000). We derive measures of trading volume (i.e., average daily trading volume over the past three months) and turnover (i.e., annualized trading volume as a percentage of shares outstanding) from January 1991 to December 2007. We show that two trading-volume measures—trailing three-month trading volume (i.e., shares) and turnover—are monotonically related to price-to-book ratio (PB) and market capitalization (MKT). We also discover a U-shaped relationship with momentum strategies (MOM) (i.e., past six-month “winners” and “losers” both tend to experience high trading volume and turnover).We next focus on the potential profitability of long–short portfolios sorted on trading volume and turnover. We form portfolio deciles based on the trading-volume measures and compare returns for the subsequent 1-month, 3-month, 6-month, and 12-month periods. Contrary to much of the existing literature, for our sample of larger stocks, we find generally monotonic patterns, with the less (more) traded stocks (i.e., on the basis of trading volume and turnover) having lower (higher) returns. For the trading-volume measure, we find that when we regress excess (of T-bill) returns on market excess returns (i.e., the traditional capital asset pricing model), the alpha is significant for the most heavily traded portfolio. These results are even stronger when we use the three-factor Fama–French model (RmRf, SMB, and HML) and the four-factor Fama–French model (with a momentum factor, UMD, added). The alpha is also positive and significant for the highest-turnover portfolio. Results are sensitive, however, to the nature of the market (i.e., bull or bear).Finally, we create new measures (“trading-volume factors”) in the spirit of the Fama–French factors and investigate their properties. We find that their betas are generally significant when added to the Fama–French four-factor model and regressed against portfolio quintile returns based on PB, MKT, and MOM sorts.Our trading-volume factors may be related to some of the findings in the behavioral finance literature. Regardless of what a trading-volume measure might be a proxy for, it is an important consideration in any quantitatively based investment strategy. Thus, our results suggest that we may look at trading volume not only as a cost of trading (i.e., related to liquidity) but also as a source of information.