The Hollowness of the Halloween Indicator

(Brandon Bell/Getty Images)

Key Points

• The Halloween Indicator suggests stronger stock market returns from Halloween to May Day, with the Dow averaging 5.3% annually in winter versus 1.9% in summer.
• The indicator’s historical advantage is largely due to strength in the six months after midterm elections, not the upcoming Year 1 winter period.
• An econometric model predicts a 4.7% gain for the Dow in the coming winter, only 1.1 percentage points above the historical average for all six-month periods.

Halloween is likely to bring more tricks than treats to stock market investors.

I’m referring to the famous Halloween Indicator, which is based on the stock market’s seasonal pattern of producing stronger returns between Halloween and the subsequent May Day (the so-called “winter” months) than during the other six months of the year (the “summer” months). Since the Dow Jones Industrial Average was created in the late 1890s, it has produced a 5.3% annualized return in the winter months and just 1.9% in the summer months.

Significant as this difference is, however, there are reasons to be cautious about betting on this seasonal pattern—over the next six months in particular.

The Halloween Indicator works only in some years. The first reason is that the bulk of winter’s historical advantage traces to the stock market’s unusual strength in the six months after the midterm elections. That’s the period beginning after next year’s Halloween, 12 months from now, of course. During the winters of the other three years of the Presidential cycle, in contrast, the stock market’s average performance is unexceptional.

This is illustrated in the chart below. Notice in particular the Year 2 winter period that we’re about to enter: The Dow’s average return in comparable periods since 1896 is actually below the average of all six-month periods since then.

What about winters following unusually strong summers? Some followers of the Halloween Indicator nevertheless insist that hope isn’t lost, since we’re coming off a summer period that’s been uncharacteristically strong. In contrast to the average summer gain since 1896 of 1.9%, the Dow since May Day of this year is up 16.1%.

Though there is some modest support for these followers’ optimism, it isn’t strong enough to justify a particularly bullish bet for the next six months. That’s according to a simple econometric model based on past correlations between the stock market’s winter performance, the year of the Presidential cycle, and the market’s strength in the prior summer.

That model predicts the Dow in the coming winter period will gain 4.7%, only 1.1 percentage points more than the average across all six-month periods since 1896. That’s not far enough above that average to be significant at the 95% confidence level that statisticians often use to assess if a pattern is genuine.

Economic policy uncertainty. One of the root causes of the Halloween Indicator’s six-months-on/six-months-off pattern, according to researchers, is fluctuations in economic policy uncertainty. Equities tend to lose ground as such uncertainty rises, as investors demand greater potential return to compensate for the additional risk. When that uncertainty retreats, stocks tend to gain ground. Followers of the Halloween Indicator need to pay attention to this tendency because there are reasons to believe that economic uncertainty in coming months won’t adhere to the normal pattern.

That normal pattern, according to Terry Marsh of the University of California, Berkeley, and Kam Fong Chan of the University of Queensland in Australia, is for the Economic Policy Uncertainty index (EPU) to rise before the midterm elections by more than during any other six-month period—and then fall after those elections’ results are known. According to a 2021 study they wrote in the Journal of Financial Economics, this seasonal EPU pattern accounts for the historical tendency for pre-midterm weakness and post-midterm strength.

The current Presidential cycle may very well diverge from this normal pattern, however. The EPU’s seven-day moving average hit its highest level in history in April of this year, as uncertainty skyrocketed in the wake of President Donald Trump’s Liberation Day tariffs. That helps to explain why the stock market performed so poorly in early April, with the average stock dropping 13.3% on an intraday basis in just three trading days’ time. The EPU’s seven-day moving average today is 45% lower than in early April, which helps to explain why equities have performed so well since then.

The implication: For the stock market to perform well over the next six months, the EPU would need to decline even further than it already has. But with the government currently shut down and no resolution in sight, the risk of higher inflation, and a renewal of a trade war (especially with China), it seems just as likely that economic policy uncertainty will rise in coming months as fall even further.

The bottom line? Don’t expect seasonal factors to provide a strong foundation for the stock market over the next few months. That doesn’t mean that the market will perform poorly. It just means that if it performs well, it will be for other factors besides the Halloween Indicator.

Mark Hulbert is a regular contributor to Barron’s. His Hulbert Ratings tracks investment newsletters that pay a flat fee to be audited. He can be reached at mark@hulbertratings.com.

Write to editors@barrons.com