When it comes to stock market seasonality there are many rules of thumb or adages that have developed into myths and have a very strong following. Everyone knows that they should "sell in May and go away" or that the fall period from September to November has seen some of the worst market crashes in history and should be treated with caution.
Most people realize two things; that the effect, if any, is psychological, and that this fact does not mean it's not real. The market after all is just composed of individual decision makers who are influenced by their own individual hopes, fears and feelings. If these market participants all hear the same stories and they act on them in the same way, then the stories create a self fulfilling reality. If everyone sells in May because they hear "sell in May and go away" then it will push the market down which will confirm the old adage. A shrewd investor cannot ignore this even if it is irrational.
You might think that these are old superstitions and today's market is far too sophisticated for them to work. You may also think that with the tight international relationships in today's markets these effects might have become muted. I used the S&P500 (.INX) as a market proxy and decided to check how well these sayings hold in recent history. I looked at the S&P500 level at the end of each month since 1989 and calculated the return for the month by dividing this level by the end of the previous month level. This simple calculation gave me the monthly return for each month since January 1989. It was a simple task from there to calculate the average monthly return for each calendar month. The following are the results:
What is evident from the data is that "sell in May and go away" has amazing staying power. I am a little baffled as to why this might be the case given that today's S&P500 returns are affected by people from all over the world who have probably never even heard the "sell in May" saying. However, if you were to sell in May and buy back in September every year since 1989 you would, on average, save yourself 135 basis points per year (ignoring trading costs).
However, not all adages appear to be confirmed. It seems that the period from September to November is actually one of the best times to invest. This period has seen some of the worst market crashes but it has also seen some of the biggest rallies. This appears more consistent with the idea that September to November has increased volatility rather than it being a bad time to be in stocks. After all market theory tells us that wherever there is increased risk/volatility there is also potential for increased return. If you were to wait until December to buy into stocks you would, on average, cost yourself 269 basis points per year. This is far more than getting out of stocks in May.
I went a bit further and broke the year down into quarters based on the returns in the above table. Instead of using calendar quarters I used the monthly returns as a basis for breaking the year up into quarters.
The above is the average return for each quarter. In theory it should be possible to beat the market by staying invested only during Mar-May and Sep-Nov while substituting some other type of investment for Dec-Feb that will yield over 57 basis point during that period.
I agree this is all a bit wonky and seasonality is a concept with little theoretical underpinning. However, the numbers do not lie, and if the next year is anything like the previous 23 years then this will turn out to be a great time to purchase broad market US index ETFs such as SPY, DIA or QQQ.