Should you care if the Dow Jones Industrial Average on Tuesday, Jan. 8 closes above where it ended 2018 â at 23,327.46? If you are believer in the famous (or should I say âinfamousâ) âFirst Five Days of Januaryâ indicator, your answer is âyes.â
According to this indicator, the U.S. stock marketâs 2019 result will follow its direction over these first five trading sessions. With four of the five trading sessions already complete, the Indicator is leaning bullish: The Dow
Â after those four days is ahead 204 points, or 0.9%.
Yet if you are a student of history, your answer is a resounding âno.â This five-day indicator rests on a surprisingly flimsy statistical foundation.
For starters, consider 2018. The first five trading days of January 2018 were spectacular, with the Dow gaining 2.3%. Thereafter through the end of the year, however, the Dow lost 7.7%.
Thatâs just one data point, of course. But experiences like last yearâs happen just as often as not. Consider all years since the Dow was created in 1896 in which it rose over the first five trading days of January. For the remainders of those years it gained an additional 7.1%, on average. That compares to a gain of 6.6% in those years in which the Dow fell over the first five trading days of January.
According to my PCâs statistical package, this difference is not significant at the 95% confidence level that statisticians often use when assessing whether a pattern is more than just a fluke. (See chart, below.)
Particularly revealing is the track record since 1975 for the First Five Days of January indicator. Since then, as you can also see from the chart, the Dow has actually performed better in those years in which the market declined over the first five trading days of January.
Given this, you might ask why anyone ever thought this indicator was worthy of attention in the first place. Itâs a good question, and the answer illustrates the perils of sloppy statistical thinking.
For example, there were many years in the early part of the last century in which the indicator worked like a charm â giving it at least a veneer of strong statistical support. But even in that period it wasnât able to survive statistical scrutiny.
For example, in this pre-1975 period in which the indicator worked better, there were several months other than January in which there was a stronger correlation between the Dowâs direction over the first five trading days and the remainder of the subsequent year. (Those months are April, May and December.) That should have been an immediate warning flag that there is nothing special about January.
Thatâs because this result points to the absence of any theoretical rationale for the First Five Days of January indicator. The presence of such a plausible explanation is crucial in order to protect you from being seduced by patterns that might on the surface look to be worth following.
As far as I know, no such explanation has been offered for this indicator. Many have been proposed, and they all boil down to the notion that the first days of the New Year are special. Anyone who had bothered to look at the data would have known that they are not.
The general lesson to learn: Subject your hunches to statistical scrutiny, and donât risk your hard-earned assets according to any indicator until you have done so.
For more information, including descriptions of the Hulbert Sentiment Indices, go to The Hulbert Financial Digest or email firstname.lastname@example.org .
More: Thereâs no method to the stock marketâs madnessÂ
Also: S&P 500 will climb 15% in 2019 â hereâs what to buy now