By studying the effect of short interest on aggregate stock returns, the authors show that short interest is one of the best predictors of future aggregate cash flows and related aggregate stock returns.
What’s Inside?
Short interest aggregated across stocks is one of the best predictors of equity risk premium. Short sellers—creators of short interest—are informed traders who, through their macroeconomic research, can anticipate future aggregate cash flows and associated returns.
How Is This Research Useful to Practitioners?
The equity risk premium is an important indicator in many areas of finance, including portfolio construction, asset allocation, and capital budgeting. The authors study the impact of short interest, aggregated across securities, on the equity risk premium.
They construct a long, monthly time series of aggregate short interest from 1973 to 2014. The series represents a measure of total short selling in the market. The authors combine the variation in short interest (because of changes in the beliefs of short sellers) with the secular changes in equity lending conditions and the amount of capital reserved for short arbitrage to create a short interest index. The short interest index shows a higher predictability than 14 other indicators in forecasting future excess market returns at monthly, quarterly, semiannual, and annual time periods. Higher values in the short interest index predict a lower equity risk premium and, in turn, lower future excess returns.
The authors show that short interest taken across securities has the highest equity risk premium predictability quotient among other popular indicators. They provide proof that the ability of short interest to predict equity risk premium and, in turn, future market returns comes largely from cash flow channels.
How Did the Authors Conduct This Research?
The authors use a monthly time series of aggregate short interest data from 1973 to 2014 from Compustat. They combine monthly short interest data with equity risk premium data and data on popular predictor variables from the existing literature. Short interest is normalized by dividing the number of shares shorted in a given firm by the total number of shares outstanding in that firm. Stocks with prices below $5 and small caps are excluded.
The authors remove the effect of the expanding equity lending market and the increase in the amount of capital devoted to short arbitrage from the overall short interest so that only economically relevant variation in short selling that reflects the changing beliefs of short sellers is taken into account. Thus, aggregate short interest created from the monthly sample data is called a “short interest index.”
The results show that the short interest index is a statistically and economically strong predictor of S&P 500 Index excess returns for monthly, quarterly, semiannual, and annual time periods for in-sample data. Furthermore, the excess return forecast based on the short-term index is superior to the predictions based on 14 popular predictor variables for out-of-sample data. The short interest index—a construct based on short sellers’ trading at the macroeconomic level—anticipates future aggregate cash flows.