Move to January. At least for the other two months, you’ll be more predictive of stock market returns than you are.
January has a reputation for predicting the direction of the US market over the next 11 months. This claimed ability is known as the “January Predictor” and the “January Barometer”.
With the S & P 500, January is officially recorded as a “down” month, so you’ll see plenty of references to this indicator in the coming days.
1.1% slip. I wrote earlier that the January predictor is based on an unstable statistical basis.. Nonetheless, financial headlines will defeat the negative impact of the decline in the remaining January of 2021.
So let me point out some other ways the predictors aren’t worth following.
What’s so special about January?
Recall that January 2020 was another month of decline (down 0.2%), but in the following 11 months it was profitable at 18.4%, well above average (assuming dividends were reinvested). ..
This is just one data point. Another clue that there is nothing special about January is that there is even greater forecasting “power” in the other months when forecasting the direction of the stock market for the next 11 months. .. In fact, since the founding of the S & P 500 in 1954, June has the strongest predictive power, followed by February. January is third.
So why not read about the June predictor and the February barometer? My premonition is that supporters are less motivated by statistical rigor than their attention-grabbing stories and stories. From a behavioral point of view, a calendar year is a more natural period of focus than the period of February-February or June-June. However, psychological significance is different from statistical significance.
Importance of real-time testing
The January indicator has another clear sign that it’s not just about cracking down. It does not pass the real-time test.
This means tests performed after the first “discovery”. If the January predictor passed these tests, it would be much more confident that it wasn’t just the result of data mining, but enough time for the pattern to appear, and the historical data wasn’t the result of torture. ..
But it couldn’t. As far as I know, real-time testing of the January predictor begins in 1973. This is the first mention on Wall Street. According to academic research on the subject.. Unfortunately, the records since then are not very impressive. In fact, since 1973, not only is the 95% confidence level, which statisticians often use to determine if a pattern is genuine, not significant, but the 85% level is also not significant.
Don’t be surprised. In fact, the January Predictor is a good companion.consider A study that appeared in a financial research review last May.. It examined 452 possible statistical patterns (or “abnormalities”) that were found to exist in previous academic studies. The authors of this recent study were unable to reproduce these results in 82% of cases. The remaining 18% turned out to be much weaker than originally reported.
There is no correlation between the rise in January and the magnitude of returns over the next 11 months
Yet another clue that January’s predictors are based on an unstable statistical basis is that there is no correlation between the strength of the January market and the rise in the market over the next 11 months. If there is such a correlation, it may be possible to create a plausible account of investor confidence at the beginning of the year and take over the rest of the year.
However, there is no such correlation. Due to its absence, to believe in the effectiveness of the January predictor, one must believe that a gain of only 0.01 on the S & P 500 provides as much predictive power as a gain of 13.2%. It undermines credibility.
By the way, since the S & P 500 was founded in the mid-1950s, it was the biggest profit in January, so I chose this 13.2% in the illustration. It happened in 1987. From January 31st of that year to the end of 1987, the S & P 500 lost 9.9%.
To benefit from a statistical pattern, you have to follow it for years
Finally, even if the January predictor is based on a solid statistical foundation, it will need to act on it for years in a row in order to reasonably work to profit from it. As a rule of thumb for statistics, you need at least 30 samples for a pattern to be meaningful. For the January predictor, this means you have to follow it for 30 years. In addition, for the last 30 years, there will be a shift to a 100% stock allocation every January 31 when the stock market rises in January, and a 0% allocation if the January market declines. We do not trade.
Without that patience and discipline, we’re doing little to improve the odds over a coin toss.
What is the conclusion? For all intents and purposes, we cannot conclude anything from the January fall in the stock market as to where to stand on December 31st.
Mark Hulbert is a regular MarketWatch contributor. His Hulbert Ratings tracks investment newsletters that pay a flat rate to be audited.He can reach at firstname.lastname@example.org
Opinion: What does the January market fall mean for 2021 equity returns?
http://www.marketwatch.com/news/story.asp?guid=%7B21005575-02D4-D4B5-4572-D247081AD9A7%7D&siteid=rss&rss=1 Opinion: What does the January market fall mean for 2021 equity returns?