I am writing this in the second week of January, so we already know that the first week, the first five days of the year, saw stocks decline. Some pundits have paid a lot of attention to that by saying they believe the first five days of the year have predictive value for the rest of the year. I have not studied that, and I would not put a much stock (pun intended) in the power of any five days’ market movement to predict a period as long as a year. Another “barometer” of stock movements that has been written about[1] extensively is the claim that the direction of stocks in January is a good forecast for the entire year. Using the first month to predict the entire year is like having a head start in a race. A better idea is to see how well January predicts the rest of the year, the February through December period. That is what I will do in this perspective.
The measure of the stock market I will use is
the S&P 500 index, and I have monthly data for it starting in 1940. That
gives us 65 years, enough to get a good idea of whether January is a good forecaster
for the rest of the year. Last year, the S&P gained 1.7% in January and was
up 7.1% for the rest of the year, so the barometer was “right” in 2004. For the
period 1940-2004, it was correct 69% of the time, which meant the market moved
in the same direction in the last eleven months of the year as it did in the
first month. At first glance, that is impressive, far better than flipping a
coin to make the projection. However, on second glance, the market gained in
72% of those Feb.-Dec. periods, so the “stopped clock” approach of just saying
the market will gain has a better chance of being right than our hoped for
barometer.
Not surprisingly, we need to go beyond a couple of glances to make a meaningful evaluation. We start by recalling the old saw about investing that “being right and making money are not the same thing.” In this case, the accuracy measure does not tell us anything about how much the market moves in the last eleven months of the year. After all, we can buy the stock index and profit from its percent gains, but unless we can find a casino or bookie that is willing to take action on the direction of the market over a period, we can’t make our gains and losses based on whether we can predict the direction correctly.
To dig further into the barometer’s predictive ability, let’s compare the years when January saw rising stock prices with those when it was a down for the month. In 37 of the 65 years, the S&P rose in January, and in 86% of those 37, the market rose during the rest of the year. More importantly, the average gain after an up January was 12.4% compared to the 65-year average of 6.9%, which is quite a difference. In contrast, when the index was down in the first month, only 54% of the following eleven months saw stocks gain, not much better than tossing a coin, and the average change for those 28 Feb.-Dec. periods was a small loss of 0.4%.
That tells us that the movement of the S&P 500 in January is a useful tool for evaluating how the rest of the year is likely to go. When the index gains, it is in effect a green light for the rest of the year, but when the first month sees falling stock prices, it is essentially a guess about how the next eleven months will fare. I did a correlation analysis between January stock movements and those the rest of the year, and I will provide details upon request. In summary, the correlation is significant from a statistical viewpoint, but does not have great predictive power in that a particularly strong or weak January does not make it more likely that the rest of the year will also show a larger than typical percent change in stock prices in the same direction.
At least two comparisons with the other months are worth making. The first is to compare the percentage of times the month has rising stock prices. Perhaps a surprise, January ranks ninth, and only September, which is known to have been the worst month for stocks historically, has a much lower frequency of gains. (December is the best with 47 up months in the 65 years, followed by October at 43, and March at 42. February and June are slightly behind January.) We can speculate that since an up January is less frequent than other months, it is more meaningful as an indicator of higher stock prices to come.
This leads to the second
comparison. Namely, how does January rank as a predictor of the following
eleven months? Without going into details, January is by far the best in that
regard. April and June also seem to have some predictive power, but the
contrast between up and down years is not nearly as great as it is for January.
My talk at the
In summary, January is an effective forecaster for the rest of the year, particularly when stocks rise during the month. How could one use it? If January is down, it indicates that it is wiser not to buy and hold for the rest of the year. Instead, other trading tactics that would move one in and out of stocks in general or a broad market index would be more appropriate. If January is an up month, then one could consider buying and holding an index like the S&P. Similar effects to those shown above hold for the following twelve months. Accordingly, one could buy the index after an up January with the idea of holding until the end of the following January. If that month is also up, then hold for another year. Another way to take advantage of the expected good stock market over the rest of the year would be to adopt more aggressive trading techniques that might be too risky for use in weak markets.