September’s Stock-Market Woes May Bode Well for October

The notion that October is a bear killer traces, as far as I can tell, to the late

Yale Hirsch,

the founder and editor of the Stock Trader’s Almanac. In an email, Mr. Hirsch’s son (and the Almanac’s current editor),

Jeffrey Hirsch,

says that his father first used the term in the 1969 edition of the publication. According to the younger Mr. Hirsch, eight of the Dow Jones Industrial Average’s 23 bear markets since World War II have ended in October—significantly more than the two you would expect if you divided the 23 bear markets equally across all months.

Midterm effect

Perhaps the first clue that October’s reputation may be more than just a statistical fluke is that four of the eight October-ending bear markets to which Mr. Hirsch refers occurred in midterm election years. An academic study published a year ago finds strong evidence of abnormal stock-market weakness over the six months up to and including the October before midterm elections and above-average strength in the six months after.

The study’s co-authors—

Terry Marsh,

an emeritus finance professor at the University of California, Berkeley, and CEO of money-management firm Quantal International, and

Kam Fong Chan,

a professor of finance at the University of Western Australia—believe this midterm pattern is caused by the heightened political and economic uncertainty that exists before midterm elections and the resolution of that uncertainty afterward.

That doesn’t mean we should expect the market to hit its midterm low on the exact day of the election, Dr. Marsh says. The midterm seasonal pattern is based on general tendencies, so the idea that the midterm low would occur in October is entirely plausible given the level of research precision. Also, he says, it’s possible that as the market tries to anticipate the election outcome, it in October will have a firmer idea than in prior months.

October as ‘Bear Killer’?

Percentage of Dow Jones Industrial Average’s bear markets since 1946 ending in these months:

September selling

The midterm pattern at most only helps to explain why half of the post-World War II bear markets came to an end in October. Several researchers and market experts point to other ways in which the stock market behaves unusually in the fall that could increase the likelihood of a bottom occurring around then even in non-midterm-election years.

Specifically, they refer to the above-average levels of selling that occur in the typical September. Crucially, this selling largely has nothing to do with fundamental factors, so the market can begin to recover in October when that artificial selling pressure abates. The likely causes of September’s artificial selling pressure include:

• End-of-quarter window dressing and tax-loss selling, according to

Lawrence Tint,

the former U.S. chief executive of Barclays Global Investors, the organization that created iShares (now part of BlackRock). In an interview, Mr. Tint says both factors should be particularly strong in Septembers of bear-market years. The first occurs when mutual funds sell losing stocks to avoid the embarrassment of having to report owning them in end-of-quarter reports. The second occurs when stocks are sold to realize losses that can be used to offset capital gains on which tax would otherwise be due. All mutual funds are required to end their tax years in October, and some managers will harvest tax losses in September to avoid the need to sell at the very end of the tax year when others must do the same.

• Seasonal affective disorder, known as SAD, is a depressive mood disorder related to the change of seasons that severely afflicts a small percentage of the population. According to

Russ Wermers,

a finance professor at the University of Maryland who co-wrote a study correlating SAD with mutual-fund flows, it also is “associated with a greater level of risk-aversion by a much larger body of the population.” Though we tend to associate SAD with the winter months, Dr. Wermers says that what affects the stock market isn’t the absolute number of those suffering from SAD at any particular point, but changes in that number. And the biggest month-to-month change in those suffering from SAD—and the risk aversion with which it is associated—occurs in September, when daylight wanes faster than in any other month, he says. That, in turn, leads to a significant net outflow from mutual funds in the average September, the research found.

This SAD effect works to the market’s favor in October, according to

Mark Kamstra,

a finance professor at York University and one of Dr. Wermers’s co-authors. Dr. Kamstra explains that SAD-afflicted investors will have largely made their portfolio adjustments by the end of September, at which point the marginal investor will be those who instead are “eager for risk and return.”

It’s impossible to know the extent to which the stock market’s large loss last month was caused by window dressing, tax-loss selling or SAD, and therefore difficult to even speculate how much the market may experience a bounce in October. But the odds of such a bounce are higher than they otherwise would be since this is a midterm election year.

Even if this expected bounce is strong enough to hasten the death of this year’s bear market, the timing of such a bottom is anything but precise. The midterms don’t occur until early November. The wash-sale rule, which requires a 30-day waiting period until stocks sold for tax-loss purposes can be repurchased, may lead to increased buying pressure throughout October. But Dr. Wermers, for one, wasn’t surprised by the stock market’s strong performance over the first two trading days of October.

Mr. Hulbert is a columnist whose Hulbert Ratings tracks investment newsletters that pay a flat fee to be audited. He can be reached at reports@wsj.com.