Wharton’s Jessica Wachter discusses her research about superstition-driven investor behavior on Wharton Business Daily on Sirius XM.

Investors of all types, including the savvier ones and even corporations, let superstition guide their decisions at times. They tend to put a pattern around one-off events because they want to find structure and predictability, even if the underlying data does not support that desire.

Such superstition-driven investment behavior is often responsible for the high volatility in stock prices, according to a study by Wharton professor of financial management and finance Jessica Wachter and Hongye Guo, a doctoral candidate in finance at Wharton. The study, titled “‘Superstitious’ Investors”, attempts to explain stock market volatility with the premise that it is “too large to arise from rational expectations of future dividends.”

While it is hard to say exactly how superstitious investors are, “there’s no question that investors are superstitious,” Wachter noted in a recent interview with the Wharton Business Daily show on Sirius XM. (Listen to the full podcast above.)

Existing research on such volatility assumes rational investors, but fails to convincingly establish a relationship between the returns for investors bearing risk and the risk premiums, the authors contend. They propose a model for stock return volatility “that does not assume rational investors with full information.” The researchers clarify that they haven’t assumed that investors are irrational; instead, such investors have biases before they process the relevant data. “The superstitious investor model captures the correct magnitude of return predictability, and the lack of dividend growth predictability,” the authors write in their paper.

Wachter and Guo trace their findings to landmark research in the late 1940s by American psychologist B. F. Skinner that showed how pigeons developed bizarre habits of behavior when presented food at regular intervals. Research since then has shown that the pigeons’ behavior “illustrates a tendency to create structure out of randomness,” Wachter and Guo write. Research has also shown that “the strong tendency to find structure where none exists characterizes human subjects as well, both in the laboratory and real-world situations … and it persists even when subjects are trained to know what is random and what is not,” they add.

The researchers contend that investors wrongly believe they can forecast dividend growth “using a persistent signal,” arguing that dividend payouts are, in fact, random events. “Prices embed incorrect beliefs about dividend growth, and thus are excessively volatile,” they write. Their study analyzes data covering the period from 1927 to 2017, sourcing it from the Center for Research in Security Practices at the University of Chicago, a provider of historical databases in stock market research.

“The ones who have seen a bad stock market event actually exhibit more risk aversion.”

“An asset price is today’s forecast of the future outcome of a random process, such as a company’s dividend, or a country’s exchange rate,” the authors note. “Any information investors think they have about this future outcome will be in today’s price. And yet if the process in question is not in fact forecastable, the price will adjust to meet reality, rather than reality adjusting to meet the price.”

The Superstition Effect

Black Monday — October 19th, 1987 — saw the largest stock market drop in history. The Dow dropped 22.6% that day, known as the Crash of 1987, and also as the “October Effect,” joining a list of other stock market superstitions. However, there was also a very large stock market drop on October 26, 1987, the one-week anniversary of Black Monday, Wachter notes. “How would you possibly explain that?” she asked during her interview with Wharton Business Daily. “It’s almost as if, well Monday came and people were worried that the same thing would happen again. It may be that people entered the next Monday just not even realizing how pessimistic they were. It’s operating outside of the bounds of what we think of as our conscious control.”

For example, researchers have focused on stock market returns on ‘October Mondays,’ which for a while tended to be “particularly bad,” said Wachter. Another phenomenon is the ‘Monday effect’ where stock market returns tend to be lower on Mondays, she added. “What’s interesting about that is you don’t see a Tuesday effect or a Wednesday effect. The very fact that there should be any day-of-the-week effect is a psychological phenomenon, because why should stock market returns be predictable by the day of the week?”

Putting Order on Disorder

Oftentimes, investors attempt to force-fit a theory to give their actions a cover of rational behavior. “There’s a lot of evidence and psychology on motivated beliefs, where you take an action for some reason, and then you go back and construct a reason for it,” said Wachter. “But the reason that you construct is not really why you took the action in the first place.”

Wachter pointed to the fear around Friday the 13th as one “that can’t be consistent with the laws of physics.” She sees that as an example of “how people like to put order on things that are essentially unpredictable.”

When it comes to risk in the stock markets, Wachter agrees that in many cases the risk is baked into the pricing to begin with. “The question is whether there is too much risk baked into the financial markets,” she noted. “Essentially, people are too scared of the stock market on average. But often we are cued by our environment in ways that might be optimal or might have been optimal through most of human history, but are perhaps suboptimal when it comes to investing.”

“The very fact that there should be any day-of-the-week effect is a psychological phenomenon, because why should stock market returns be predictable by the day of the week?”

Evidence shows that even professional investors are not immune to being superstitious, Wachter pointed out. “The ones who have seen a bad stock market event actually exhibit more risk aversion,” she added. “You would certainly expect to see this in the domain of decision making even with sophisticated investors, and even within companies.”

What Guides Investment Behavior?

For sure, investors may consider some events as solitary and not part of a norm. One example of that is when a company’s earnings report contains some surprises, said Wachter. “One question that you’re faced with is: Is this a permanent change, or is this a temporary change?” she added. “We know from studies that investors tend to under-react often to earnings reports that produce momentum — something known as earnings momentum — over, say, a three-month horizon.”

Bad news is stickier, it seems. “People do look at a positive earnings surprise and think of it as a one-off, even if the data says otherwise,” said Wachter. “Deciding whether something is one-off or a persistent process is really tricky just as a statistical matter. And there’s plenty of evidence that people get it wrong.”

Another line of research says that people’s personal experience with the stock market influences whether they invest or not, or whether or not they participate in the stock market, said Wachter. “Somebody whose personal experience with the stock market is that returns are mostly positive is more likely to stay in the stock market than somebody [who has not received positive returns], and therefore they associate the stock market with negative returns. Of course, strictly speaking this is also not rational, because we all have access to the stock market data. It’s the same stock market data for everybody.”

Expecting a market trend to continue even if that hope is not backed by data is another hazard. “If you’ve seen a stream of positive returns, you may be more likely to think that [more] positive returns are coming,” Wachter noted, adding that such extrapolation is unwise. “If all else is equal, it seems like periods of high valuations in the stock market are followed by periods of lower-than-average returns. But timing the stock market is pretty hard to do.”

More research into investor behavior is needed to sharpen insights, according to Wachter. “We’re trying to formulate a set of hypotheses that bring together economics and psychology and then just go out and test them in the data,” she said. “The hope, eventually, is to come up with a coherent theory that incorporates some of the really robust findings from psychology as well as the findings in economics.”