How Wise are Crowds?
Insights from Retail Orders and Stock Returns
Individuals who make their own decisions about which stocks to buy and sell are known as retail investors. These investors account for over a trillion dollars of stock trading each year and potentially play an important role in observed stock prices. While this role has been considered by numerous researchers in the past, it is not particularly well-understood.
In this paper, the authors analyze the largest sample ever studied of U.S. stock orders submitted by retail investors. The sample, which was provided by two anonymous market makers, contains over 200 million orders and accounts for $2.6 trillion in trades during the years 2003-2007. With this data, they study the relationship between buying and selling by retail investors and future stock returns. Specifically, they ask if and why net retail buying (i.e., the difference between buy orders and sell orders) for a given stock on a given day can predict future returns for that stock. They have three main results.
First, net retail buying in a stock is positively related the same stock’s return over the subsequent month. The strongest predictability occurs within the first week after trading is observed, but it never becomes negative even over very long horizons. This is important because it implies net buying by retail investors predicts permanent changes in a firm’s stock price and suggests these collective trading decisions provide new information to the market. The result is quite broad in that it’s true for aggressive orders (those designed to executive immediately) and passive orders (those submitted by patient traders that will only execute if the market price improves).
Second, the authors programmatically analyze the text of all firm-specific stories contained in the Dow Jones Newswires and extract measures describing the overall tone of these stories. They find that net buying by retail investors also predicts the occurrence of positive future news. However, this is only true for aggressive orders. This finding suggests that these aggressive orders collectively contain fundamental information about the firm. This information, which may be undetected by professional traders, could plausibly be based individuals’ geographic proximity to firms, relationships with employees, or insights into customer tastes.
Third, they show that net buying from passive orders (but not aggressive orders) is also related to prior and contemporaneous negative stock returns. This finding suggests that retail investors submitting passive orders do so when demand from other more aggressive traders has pushed or is pushing prices away from fundamental values. Thus, these retail investors are providing liquidity to other parties who want to trade quickly. The predictable future stock returns are therefore related to corrections to the stock price.This study improves understanding of the role of retail investors in the stock market, and it provides new insights into the mechanisms at play. The main results suggest retail investors are important for stock prices, both in terms of the information they reveal and the liquidity they provide. Moreover, in the current environment in which professional traders attempt to predict stock returns, observing the buying and selling decisions of retail investors in real time is useful to whatever extent it is possible.
Eric K. Kelley is an assistant professor of finance at the Eller College of Management, University of Arizona. Paul C. Tetlock is Roger F. Murray Associate Professor of Finance at Columbia University.
Published in The Journal of Finance.