Popular media and investors commonly state that summer is the least favorable time to invest in the stock market. Arguments for why this is the case are fairly speculative, but include family vacations (school is out) paid for with cash taken out of portfolios, and also the seasonality of the fiscal cycle with full year results being reported in the summer. A summary of these ideas can be found here. If we look at historical performances of the main stock indices, do we see this trend?
The first plot shows the average performance in the S&P500 (red) and NASDAQ (blue) for each month since 1972. So, for example, the January bars shows the average gain (in %) by the S&P500 during all the Januaries since 1972. Note that this is the real gain; I have adjusted for inflation using the full CPI. The dashed lines show that the average real monthly % gain by the Nasdaq is 0.31%; 0.18% for the S&P500. Thus, from the big picture, even though the Nasdaq is considered more volatile, it has rewarded investors who took on that extra risk with a 70% larger return than the S&P500. The main caveat here is that this is for 40 years of data. If we choose any 10 year periods, this may not hold (for example, the late 90's dot-com crash). Thus, one would truly have to be a long-term investor to see this gain.
click to enlarge images
As suggested by the opening paragraph, the plot above clearly shows that the July-September months have consistently seen much worse returns, actually losing real value, compared to the rest of the year. Here we also see that the Nasdaq basically shadows S&P500 gains except during December and January, when it outperforms by a factor of two or three.
Although this data looks convincing, it is clear that in any given year, this trend is not directly obeyed. Because these values represent averages, it is also very useful to inspect how the averages are obtained. In the next plot, I have shown this by again separating all months and then sorting each month's historical gains from lowest to highest. I have chosen the S&P500; the results are fairly similar for NASDAQ. The black plus shows that median value (the median is the middle value after sorting the data- it is NOT the average). The red circles show the 67th, 75th and 90th percentiles, while the blue circles show the 33rd, 25th and 10th percentiles. Comparing the median with the average from the figure above shows fairly good agreement during most months, but not so good agreement during the summer. For example, during July and August, the plot above shows average real losses, while the median value (below) shows either no gain or a slight gain. This suggests that a few extreme months where stocks lost a lot of value are strongly affecting the average, making it less representative of what has occurred.
Also from the chart below, it is interesting to not that the most robust growth in the S&P500 has occurred during November and December. In particular, the S&P500 has appreciated in real value in 9 out of 10 Decembers.
As my curiousity was piqued from the first graph showing the lack of gains during summer, I decided to run a quick thought experiment. I don't think this is novel (given that people already know of the summer minimum), but it was interesting to me nonetheless. I wanted to know what would happened to the real value of $100 if invested on January 1st, 1972 using two methods:
- Keeping this money invested the whole time (in the S&P and/or Nasdaq).
- Investing in these indices only during November through May, followed by liquidating for the summer months.
The plots below show this where the magenta line is method (1) and the green line is method (2). The results, at least before the caveats are included, are quite shocking. A simple cashing out during late May followed by reinvestment during November would have roughly doubled your gain in the S&P and nearly tripled your gain in Nasdaq.
Now, the caveats. I have not considered dividend reinvestment plans (DRIPS) and perhaps more importantly, I have not considered that by holding for less than 1 year, an investor would have been taxed at his/her normal income tax rate versus the more favorable capital gains rate. The latter point would definitely degrade that gains significantly, however, I think the results clearly suggest it pays off to keep in mind the time of year when investing. I am also repeating these quick analyses for certain individual stocks and/or sectors. If anyone is interested in seeing the results for a particular stock, please contact me.