Another reason ‘sell in May and go away’ is wrong «
The S&P 500 is at an all-time high. The month of April has ended. You can now expect the repetitive droning on about the old Wall Street adage, “Sell in May and go away.” Unfortunately, that saying is likely for the quantitatively challenged, because as the statistics will show, that the saying is false.
To begin, we must ask, what does “sell in May and go away” mean? The reasoning is that once the markets turn from April to May, one should abandon the stock market. It is not clear by that adage when to come back into the equity markets.
On the one hand, since we are talking adage and October is the month of “crashes,” perhaps one should reenter the markets at the end of October, no?
However, when looking at monthly data (my data for the S&P 500 goes back to 1950), on average, the S&P 500 has gained 0.74% in October — so much for crashes. Hence, one can assume that the intention is to return to the markets after September. So let’s give the adage followers the benefit of the doubt that the plan is to sell at the end of April and reenter the market at the end of September.
The simple average annual index price return for the S&P 500 from 1950 through 2012 is 8.71%. On an annual compounded basis, the S&P 500 gained 7.32% per year. For the “go away” months, the S&P 500 performs as follows on a simple average annual basis:
Month Percentage gain
Thus, if you follow the adage, you would forgo an actual 0.55% of gains on average annually by selling in May and going away. Here is what it would cost. If you invested $100 for 25 years compounded at 7.32%, you would accumulate $584.81 at the end of those 25 years. However if you sold in May and came back in October, the lost opportunity of 0.55% annually would result in that same $100 growing to only $514.32 at the end of 25 years. In other words, you would have generated 12.05% less in appreciation if you followed the adage.
Perhaps the adage should be to sell in August and go away until October. Then you will avoid the normal average declines in the months of August and September.
So, now that the “sell in May and go away” adage has proven to be a profit loser, can we save it? Intuitively, the “sell in May and go away” axiom would imply that gains in the year are front-loaded into the first four months of the year. Let’s say that the S&P 500 from January through April outperforms its annual simple return. Should we then sell and go away?
In order to test this hypothesis, I looked at returns for the S&P 500 for the first four months of every year in my data base going back to 1950. Then I culled out years in which the index gained less than 8.71% in that four-month period, leaving me with periods when the S&P 500 gained more than 8.71% for the first four months of a year. Next, I calculated the returns from May through December for those years. Did that prove “sell in May and go away”? Quite the opposite, as can be seen from the data below:
Year Percentage Gain
The data is overwhelming. Not only do we not want to “sell in May and go away,” but we want to stay long. Out of 16 occurrences, the S&P 500 increased on 14 occasions from May through December. The average gain was 6.93%.
So as we see, the “sell in May and go away” adage is fallacious. Consider this old saw as just another deceptive homily of conventional wisdom. Another such bit of conventional wisdom that I recently disproved was the Pavlovian behavior which causes traders to sell after one-day discrete events , such as the recent Boston Bombings. You want to buy at the end of one-day discrete events and not sell.
So, next time someone tries to convince you to abide by some old Wall Street wives’ tale, ask them to show you the data to support their position before you take any action.