When considering whether to buy stock in a company, investors often look to the trading activity of the company's top executives. If the CEO or CFO has recently made large purchases of company stock, investors tend to assume the stock price is about to go up; if they are selling stock, on the other hand, the meaning is less clear.
A new paper by the University of Notre Dame's Peter Kelly takes a rigorous look at the predictive power of insider trading—not the illegal kind, which is based on access to information not in the public domain, but the normal trading activities of corporate executives.
Generally speaking, insider purchases have very strong return predictability, but insider sales do not. Kelly, an assistant professor of finance at Notre Dame's Mendoza College of Business, wondered whether those insider sales might contain more insight into future returns than they appear.
While most traders purchase stock because they have good reason to think the stock will go up in the near future, there are any number of explanations for selling stock, he explained.
"You might sell because you think that you need to diversify your portfolio, or because of the fact that you need to pay for your kids' education, or because you need to make a down payment on your house," Kelly says. "So there are a lot of reasons why people could sell that would be unrelated to the company's underlying financial information."
Using historical stock market data, he found that a subset of insider sales do predict future returns. Specifically, he determined that when insiders sell company stock for a loss, the stock's subsequent six-month return is 188 basis points lower than all other firm-months. On the other hand, when insiders sell their stock for a gain, there is essentially no return predictability. His study, "The Information Content of Realized Losses," was published in The Review of Financial Studies.
Why is the subsequent performance of stocks sold at a loss so much worse than stock sold for a profit? Kelly argues that because investors hate to lose money on a trade, they need a stronger negative information signal to sell at a loss than to sell at a gain. "Since selling a stock at a loss is painful, an investor who sells at a loss must have particularly negative information," he explains. "And what you see is when stocks are sold at a loss, it predicts negative returns."
Portfolio managers may be able to use Kelly's research to generate excess returns for their clients. By purchasing stocks that have been recently sold at a gain by company insiders and selling stocks that have recently been sold at a loss, Kelly finds that a manager would earn 67 basis points per month higher than the benchmark index.
Kelly, whose Ph.D. is in finance, credits his interest in psychology with leading him to this research topic. "A lot of people entrenched in finance academia think that markets are sort of perfectly rational and very efficient," he says. "I tend to believe that behavioral biases (such as investors' strong aversion to selling stocks at a loss) play a larger role in the pricing of assets and investor decision-making."
Explore further: Researcher solves the mystery of the stock market's "beta anomaly"
Peter Kelly. The Information Content of Realized Losses, The Review of Financial Studies (2018). DOI: 10.1093/rfs/hhy013