The stock market isn’t always rational; often it’s swayed by superstitions. A case in point is the bump the market gets when Punxsutawney Phil doesn’t see his shadow.
Savva Shanaev of Northumbria University and two colleagues mapped a century’s worth of U.S. stock-market returns against the annual predictions of Punxsutawney Phil, the star of the long-standing North American custom of Groundhog Day. According to the tradition, if the Pennsylvanian groundhog sees his shadow on February 2, winter will last for six more weeks. If he doesn’t, spring is around the corner—and, the researchers found, the market gets a boost. The conclusion: When the groundhog predicts an early spring, investors get optimistic.
Mr. Shanaev, defend your research.
Shanaev: Generally speaking, the market is much more rational than you might think. It’s incredibly efficient when it comes to accurately condensing information about value. But sometimes it can be glaringly irrational.
HBR: So Groundhog Day isn’t the only anomaly?
Not at all. Researchers have also documented the “sell in May and go away” effect, which reflects the fact that the market tends to be at its worst from May through October. Then there’s the January effect, so-called because stocks often rise at the start of the year, and the Monday effect, in which market returns are lower than average at the beginning of the week. Some studies have also found that stocks perform poorly around the full moon and when Mercury is in retrograde.
Many of these effects are fairly consistent around the world, but other calendar- or superstition-based market anomalies are more localized. In China, for example, returns increase during Chinese New Year, and stocks whose ticker symbols include the lucky number eight tend to perform better than average, while those whose symbols include the unlucky number four perform worse. Israeli market returns are higher on Rosh Hashanah, a cheerful Jewish holiday, and lower on Yom Kippur, a somber one. And many Islamic countries’ markets exhibit abnormally positive returns during the celebratory month of Ramadan.
How much do Phil’s predictions affect the market’s performance?
Overall, I found no statistically significant changes after a long-winter prediction, but the market appreciates by 2.78% after an early-spring prediction. This is most likely because early-spring predictions are much rarer, occurring only once every four years, on average, so investors react more strongly to them.
Why would they be influenced by information that has nothing to do with companies’ actual value?
Some investors are genuinely superstitious, if only on a subconscious level. Others may not be but are swayed by the shift in public sentiment that often accompanies cultural events like this one. Still others may be crafting investment strategies according to how they think superstitious investors will react. In fact, the groundhog’s impact on the stock market starts to become evident two weeks before his February 2 predictions—suggesting that the latter explanation is at least partly in play.
How does that work?
Phil generally doesn’t see his shadow if there’s heavy cloud cover in Punxsutawney on February 2, and reasonably reliable weather forecasts are available up to two weeks in advance. Now if the Groundhog Day stock-market effect were driven purely by superstition or sentiment, you would expect to see it only after the prediction had been made. Because we start to see abnormal market activity two weeks before then, I suspect that some investors are tracking Pennsylvania forecasts and adjusting their market positions accordingly.
Specifically, if they see a cloudy day in the forecast, they know the market is likely to go up on Groundhog Day, and so they buy before then, causing the market to rise slightly. They then sell shortly after February 2, causing the post–Groundhog Day boost to be somewhat lower than it would be if driven only by superstitious investors’ trades on or immediately after February 2. In other words, the small but statistically significant pre–Groundhog Day effect suggests that there are at least some nonsuperstitious investors who are aware of the anomaly and are actively investing against it.
If early-spring predictions correlate with cloudy skies, couldn’t the abnormal market returns simply be driven by the weather?
Researchers have certainly documented correlations between weather and investing behaviors. However, their findings actually support my conclusions. Bad weather generally corresponds to worse-than-usual market performance—and yet when there’s bad weather on February 2, the market outperforms, suggesting that something else is going on. And in any case, the weather in one small town in Pennsylvania is unlikely to have much impact on national trading activity.
Was the effect stronger in certain industries or markets?
Some of my colleagues have explored whether Punxsutawney Phil’s forecasts have a larger effect on the stock prices of companies with strong ties to Pennsylvania, such as those in the steel industry, but their findings were inconclusive. In my own analysis I broke down results by sector but saw no significant differences.
Unsurprisingly, given that Groundhog Day is a North American custom, the effect isn’t seen in other markets. I analyzed returns in the UK, Australia, Germany, France, and Japan and found no indication of an anomaly around February 2 in those countries.
Do you expect the Groundhog Day phenomenon to persist?
In my research on other stock-market anomalies, I’ve found that the effects tend to diminish as the investing community becomes aware of them. When the market recognizes an anomaly, it usually self-corrects. But that analysis focused on more widely studied and publicized anomalies, such as the Monday effect, which has become so well-known among investors that it has largely disappeared. The Groundhog Day effect is much less well-known, so while our data suggests that some investors are already counteracting it to a degree, the market is unlikely to fully correct for it.
I suppose this article could change that!
If enough people read it, perhaps it could! But remember, the Groundhog Day effect is relatively minor in scope. It offers potential abnormal returns of less than 3% to investors only once about every four years, so it probably won’t become anyone’s top priority. Studies have shown that you can make nearly double those returns just by selling in May and buying after Halloween—and that’s assuming you have the patience and resources to make such a low-reward investment strategy feasible.
So you won’t be retiring on your Groundhog Day earnings?