UNH Media Relations
EDITORS AND REPORTERS: Stephen Ciccone can be reached at 603-862-3343 and firstname.lastname@example.org. Ahmad Etebari can be reached at 603-862-3359 and email@example.com. The complete report is available for download at http://www.unh.edu/news/docs/ciccone-etebari_stocks.pdf.
DURHAM, N.H. - Investors looking for the best return in the U.S. stock market would be wise to invest in January, which generates the highest returns of any month of the year, according to a historical analysis of stock market returns conducted by two University of New Hampshire researchers.
Researchers Stephen Ciccone, associate professor of finance, and Ahmad Etebari, professor of finance, of the UNH Whittemore School of Business and Economics, present their findings in "A Month-by-Month Examination of Long-Term Stock Returns." The research was presented recently at the Northeast Business & Economics Association Conference.
The researchers conclude that of all the months of the year, January nets the highest return on stock purchases. On the other hand, investors may want to avoid stock purchases in September, which the researchers determined results in the worst return.
"The results of this study demonstrate the existence of monthly seasonal patterns in the stock market. These seasonal patterns continue to present a challenge to market efficiency proponents. Of particular interest is the persistence of the January Effect even though it has been well known for about 30 years. Despite much research, seasonalities remain among the more puzzling aspects of the stock market," Ciccone and Etebari said.
The researchers looked at 81 years worth of U.S. stock market data from 1926 to 2006 to determine whether any months generated abnormal returns. They found two - January and September.
If investing each year in an equally-weighted stock index only during January, $1 in 1926 grows to $87.40 by 2006. If the same $1 was invested in the second best month, July, it would grow to only $3.11. According to the researchers, January has an "astonishing 81.48 percent of positive returns." Outstanding small-firm stock performance fuels the January Effect, which persists across the 81-year period.
"Despite superior performance, the risk in January is rather average. January's strength appears due to its low frequency of poor returns, not a high frequency of exceptional returns," Ciccone and Etebari said.
On the other hand, that same $1 invested in only September would drop to 49 cents by 2006. The September pattern also was found in four of the five international markets evaluated -- the CAC 40, the DAX 30, the FTSE 100 and the Nikkei 225. And October isn't much better - $1 invested in 1926 shrinks to 53 cents by 2006.
"Extraordinary returns may affect the results. For example, the poor performance in October is often attributed to two major events: the stock market crash of 1929, which marked the start of the Great Depression, and the stock market crash of 1987, the largest one-day drop in market history," Ciccone and Etebari said. However, the conclusions did not change after dropping the three best and three worst returns for each month.
The researchers also note two additional seasonal stock patterns - the "Summer Rally" and the "Halloween Effect," often referred to as "Sell in May." Investors have anecdotally thought that summer - June, July and August - is a good time to buy stocks, despite little evidence backing it up.
While at a glance the summer months appear to be a good time to invest, with a 1926 dollar growing to $15.46 in 2006, the November through April period - also known as the Halloween Effect - generates a much higher return. A dollar invested each year from November through April starting in 1926 would grow to $3,891.98 by 2006. If invested in May through October instead, the $1 grows to only $6.42.Ciccone and Etebari note that the Halloween Effect primarily is driven by strong January returns and weak September and October returns.
"Overall, seasonal stock market patterns such as the January and September Effects pose serious challenges to notions of market efficiency. This is especially true given the fact that most seasonal patterns have been known for quite some time, yet they continue to persist," the researchers said. They caution, however, about using their results in a trading strategy, noting that January 2008 was one of the worst Januarys on record.