Stats › Seasonality and calendar › January barometer
The January barometer
As January goes, so goes the year. That's the whole claim, coined by Yale Hirsch back in 1972, and it's still repeated every single February. If the S&P 500 finishes January higher, the year is supposed to finish higher too, and if January is down you're meant to worry.
Its advocates publish an accuracy figure of 84.5%, which sounds impressive until you ask two questions. First, how often is the market up anyway? And second, is January inside the year you're predicting? The answer to the second one is yes, obviously, and that turns out to matter more than anything else on this page.
So what you'll find here is the barometer's own published numbers, taken from the people who promote it rather than from a critic, sat next to the base rate they need to beat. There's the headline accuracy, the stricter version that counts near flat years as misses, and the conditional figures for up and down Januaries, which are the genuinely interesting part. The reason I wanted this page in the library is that the barometer is a really good example of a statistic that isn't fabricated and still doesn't mean what people think it means. Two sources, both linked below with their date ranges. The Almanac's own page is the place to start if you want to argue with any of it.
The Stock Trader's Almanac puts the January barometer's accuracy at 84.5% across 1938 to 2021, with thirteen major errors. Counting flat years, meaning full year moves smaller than 5% either way, as errors brings it down to 73.8%. After an up January the full year has been up 86.5% of the time with an average gain of 16.2%. The problem is structural: the predicted period contains the predictor, so part of the apparent accuracy is arithmetic rather than forecasting.
Members only
The rest of this page is for members
Below this point there are 7 sections, 1 chart and 2 named sources, roughly 1300 words of it. Every figure carries the source it came from and the date the data is from.
You can keep browsing every statistic in the library for free. The intro and the summary are always open.