Sell In May: Fact Or Fiction?

Jun. 4.13 | About: SPDR S&P (SPY)

There are many different calendar anomalies that are supposedly true, one such anomaly is referred to in the saying "sell in May and go away." This refers to the idea that the stock market generally underperforms from May to October. However, according to the efficient market hypothesis market anomalies are not supposed to exist at all. If markets efficiently price everything all the time, why should markets price stocks relatively lower in summer months? It seems to make no sense. This should lead investors to ask many questions. Is this calendar effect based on fact? Is it likely to continue to hold in the future? In this article I will look at historical data and compare it with recent data in attempt to determine whether the saying "sell in May" is justified. I will also try to come up with reasons as to why it may be true.

Historical Data (source: Bloomberg)

In the years from 1928 to 2012 the S&P 500 returned an average of 6.86% per year before dividends. Therefore, the average monthly return of the S&P 500 was about 0.57% not including dividends. However, when including just the summer months of every year (which for all my analysis will include May, June, July, August, and September) the average monthly return of a summer month was only 0.32%. Even more shockingly when looking at the most recent data, only since 2000, the summer month's average return was -0.52% when compared with an average recent monthly return of 0.1%. The chart clearly shows that the summer anomaly seems alive and well.

Click to enlarge

Possible Explanation

Another calendar anomaly is about presidential cycles, one hypothesis is that the summer's average performance is pulled down because going up to elections uncertainty drags the market down. To test this I split up the summers based on the year in the presidential cycle. For example the first year is the summer after a president was elected (we are currently in one of these years). The fourth year is the president's last year before election. According to this theory during fourth year summer the S&P should perform terribly. After looking at the chart this is clearly not true, in fact historically the fourth-year summer has been the best of all four. Even after separating more recent summers the fourth summer is still not the worst. The current summer we are in is also neither the worst nor the best. Therefore, the presidential cycle cannot explain the summer anomaly.

Click to enlarge

Click to enlarge


Although the summer anomaly seems to be alive and well until a concrete reason for it is shown I believe it is just a coincidence. Just look at the presidential cycle graphs. A pattern seems to occur when looking at all the years from 1927 to 2012. However, when looking at the last 10 years almost the exact opposite pattern occurs. Unfortunately, it is hard to argue with 84 years' worth of data. As for now people who simply invest in broad ETFs like SPY might want to take this age-old saying's advice. The summer anomaly lives on and the saying "sell in May and go away" seems justified.

Disclosure: I am long VB. I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.