Last Updated on June 19, 2022 by Quantified Trading
The January effect is a seasonal trading strategy that worked over many decades. However, the traditional January Effect seems to be arbed away. But perhaps there is another January effect that works? This is the alternative January effect:
The gain in January determines the average gain for the next 11 months: If the S&P 500 declines in January, the average percentage change for the next eleven months is 3.3%. If the S&P 500 rises during January, the average percentage change in the next 11 months is 10.3%. This is the new January effect.
The January effect that works
On pages 279-280 in The Education of a Speculator, Victor Niederhoffer explains his version of the January effect: The gain in January determines the average gain for the next 11 months. If the close of January is higher than the close of December, the average gain is substantially higher the next 11 months than if the January close was lower than the December close.
Victor Niederhoffer’s book is old – published in 1997. Does Niderhoffer’s hypothesis still hold? It does:
The equity curve below is 100 000 compounded when January showed a decline (entry is on the close of January, and the exit is at the close of December 11 months later):
The first real drawdown happened in 1974, and the strategy never really recovered after that. The decade-long bull market of 2010-2020 made the strategy recover somewhat.
But when January is positive, the equity curve improves substantially:
Quite a difference, to put it mildly. When January is up, the win ratio is 83%. When January is down, the win ratio drops to 62%.
The reason why this strategy seems to work is because of the occasionally large losers when January is down. For example, it happened in 1974 and 2008, two years which went on to fall substantially the rest of the year. When you avoid big drawdown you can start compounding from a higher level.
Conclusion: the January effect that works
The January effect presented in this article is not bulletproof (of course): It was out of the market in both 2020 and 2021 (in both years the S&P 500 went on to rise substantially). Nonetheless, it has had a fairly good prediction rate over the last 60 years.