Predicting earnings in the stock market
Management prof's research shows stock returns as predictable as calendar

By JOHN DELLA CONTRADA
Reporter Contributer

According to Joseph Ogden, 66-85 percent of the average annual return on NYSE stocks is earned in the fourth and first quarters of the calendar year (October through March), while the second and third quarters (April through September) provide only 15-34 percent of the annual returnÑdespite evidence that risk is identical in the two periods.

Moreover, Ogden's research shows that when the market is down in the second and third quarters, losses tend to be fully reversed in the following fourth and first quarters.

Ogden's findings are based on the analysis of stock returns for the 50-year period following World War II, 1945-1995, and are documented in his paper "The Calendar Structure of Risk and Expected Returns on Stocks." The paper will be presented at the Eighth Annual Conference on Financial Economics and Accounting to be held at UB on Nov. 7-8.

According to Ogden, an associate professor, the findings challenge conventional Wall Street wisdom and academic theory, which state that the market is efficient and stock returns are not predictable. Instead, both short- and long-term investors can benefit from the "calendar-predictability" of stock returns as documented by his research, Ogden says.

"Among the immediate implications is that short-term investors should generally avoid stock investments during the second and third quarters because average returns are low during this period," Ogden says. "And if they do not hold stocks through the following fourth and first quarters, they will not benefit from the subsequent reversal of the market."

"For long-term investors, however, holding stocks through the second through third period is not irrational, even though the expected return is low. Any losses in this period will be reversed if you hang in there," Ogden adds.

He explains the calendar-predictability of stocks by tying it to a corresponding seasonal pattern in the salaries of white-collar workers, who ultimately are the major investors in the market. Using U.S. Labor Department statistics, Ogden analyzed the monthly aggregate salaries of white-collar workers from 1966-93. He found a 15 percent average jump in salaries during the fourth and first quarters, October through March, punctuated by a tremendous jump in January.

"Salaries grow at a much faster rate in the fourth and first quarters of the year, probably because business executives, Wall Street traders and others tend to receive most of their profit distributions, bonuses and pay raises in these quarters," says Ogden. "As a result, the demand for stocks, and thus stock returns, have a strong seasonal pattern."

Ogden's research also casts new light on the "January Effect"Ñthe tendency of small firms to provide extremely high returns in January. Conventional wisdom states that this effect is due to reversals of December losses induced by investors dumping small-firm stocks in December to realize losses for tax purposes.

"My research shows that tax-loss selling has a substantial impact on January returns," he says. "However, the January Effect is also part of a larger cyclical reversal of previous second- and third-quarter losses."