Fly Me to the Moon, and Let Me Profit on My Stocks

SOURCE: The New York Times – November 19, 2006

Fly Me to the Moon, and Let Me Profit on My Stocks

By MARK HULBERT

THE next time you’re about to buy or sell a stock, you may first want to look up at the moon.

You read that correctly. Believe it or not, the stock market tends to do better or worse depending on where we are in the lunar cycle.

That’s the conclusion of two studies that have circulated for a couple of years in academic circles but that, until an article in the November issue of the Harvard Business Review, received relatively little attention outside academia. The first study, “Lunar Cycle Effects in Stock Returns,” was by Ilia D. Dichev, an associate professor of accounting at the University of Michigan, and Troy D. Janes, an assistant accounting professor at the State University of New York at Buffalo. The second study, “Are Investors Moonstruck? Lunar Phases and Stock Returns”, was by Lu Zheng, an assistant finance professor at the University of California, Irvine; Kathy Yuan, an assistant finance professor at Michigan, and Qiaoqiao Zhu, a graduate student there.

Both studies agreed on a crucial point: during the 15 days of the lunar month closest to the new moon — starting seven days before it and ending seven days after — the stock market’s average returns are much higher than those of the other half of the month.

The differences are quite large — as much as 10 percent a year, on average, depending on the stock index and the number of decades studied — and they are pervasive. The researchers found a lunar effect in all major United States stock indexes over their history, including the Dow Jones industrial average back to its start in 1896. Furthermore, after examining several dozen foreign markets, the study by Professors Dichev and Janes concluded that “if anything, the results are more pronounced for foreign countries.”

One complete lunar cycle — from new moon to full moon and back again — lasts slightly less than 30 days. New moons and full moons do not occur on the same days of each month in the modern Western calendar. That is important, because it means that the stock market’s lunar cycle is not a disguised version of any other seasonal pattern based on that calendar. Other research has found, for example, that the market performs particularly well at the beginning of each calendar month, especially at the start of each year. What would cause the lunar cycle to affect the market? Neither group of researchers studied this question directly. But both cited extensive psychological and biological literature that has found that the lunar cycle can heavily influence our moods. A full moon, for example, increases our tendency to feel depressed and pessimistic, so investors may be more inclined to stay out of the stock market at or near that time.

The lunar effect appears strongest for the smallest-cap stocks — a finding that raises the researchers’ confidence in their psychological explanation. After all, according to Professor Zheng and her colleagues, relatively few institutional investors hold such stocks, and individuals are more likely to be affected by the lunar cycle — in part because institutions typically have cumbersome decision-making processes, often involving committees, that immunize them from employees’ moods.

What should investors make of this research? Perhaps the most important lesson is that our emotions too easily trump our objectivity. According to Professor Dichev, the lunar cycle in stocks “serves to remind us how much changes in our moods can affect our behavior in potentially irrational ways.” To prevent that effect, we may want to pursue largely mechanical strategies — like buying and holding an index fund — that specify in advance what actions to take in our portfolios and when.

COULD a trading strategy be devised to exploit the lunar-cycle trend? Possibly, but Professor Dichev offers two big caveats. The first is the need for investments that minimize transaction costs, like a no-load mutual fund with no redemption fee for short-term trades. (To make full use of the cycle, you’d make more than a dozen round-trip trades a year.) The second is that there is no guarantee of success in every cycle. Based on historical data, the strategy would succeed only about 60 percent of the time, Professor Dichev said. Though that’s enough to make it profitable in the long term, you’d have to tolerate losses 40 percent of the time. The strategy worked in the most recent cycle: the Standard & Poor’s 500 index returned just 0.3 percent in the 15 days surrounding the full moon on Nov. 5, versus 0.9 percent in the days surrounding the preceding new moon. But the strategy failed in the previous cycle: the index was up 2.2 percent in the full-moon period and 1.5 percent in the new-moon period.

Of course, there is an even more fundamental problem: You have to be willing to bet your hard-earned money on the phases of the moon.

Mark Hulbert is editor of The Hulbert Financial Digest, a service of MarketWatch. E-mail: strategy@nytimes.com.