It’s that time of year folks. When the “Sell in May” statisticians provide the historically compelling numbers to make the case for being out of the stock market for the next 6 months. If you are not aware, the odds are heavily stacked against you if you do. Or are they? Much of the evidence is due to the fact almost all of the nasty bear market declines (’29, ’87, ’00 and ’08) began during this period. But the facts are the facts. On que, Dana Lyons posted an obligatory but non-confirming article that looked at the data in a different light and dug a little deeper than the raw numbers. Because of what he found and the fact it is contrarian, I thought it worthy of passing along…
----------------------------------------------------------
There is some evidence to suggest that stock investors would be well served by NOT selling in May and going away this year.
One of the most famous of all of Wall Street’s trading bromides is “Sell In May & Go Away”. Of course, the saying refers to the tendency of stocks to perform worse during the 6 months from May through October than they do from November through April. Perhaps the reason why it is still so popular is that, unlike some of Wall Street’s other sayings, there is actually solid historical evidence to back it up, including recent history.
Specifically, here are the average returns in the Dow Jones Industrial Average during the 2 periods since 1900:
November-April: +5.45%
May-October: +1.61%
Not only does the average return for the May-October period lag badly, but the consistency of positive returns has been much less reliable:
November-April: 69% Positive Returns
May-October: 61% Positive Returns
Plus, as mentioned, unlike many seasonal tendencies that lose their effectiveness over time as the edge gets arbitraged away, the Sell In May pattern has held true even as of late. For example, 6 of the last 7 years, 10 of the last 12 years and 18 of the last 23 years saw the Dow stronger from November-April than from May-October.
So is there anything that bulls can hang their hat on during the forthcoming 6 months? Well, first of all, despite the fact that May-October has generally lagged its 6-month counterpart, the historical average return for the period is still positive. So it’s not like the whole period has been a disaster, though there have certainly been some of those.
In parsing the data, however, we have found one historical trend that may actually encourage traders to stay invested during this year’s May-October period, rather than the mere expectation that their investments will just be biding their time. It is based on the performance of the just completed November-April period – specifically, the Dow’s strong (but not too strong) +15.4% performance.
Here’s what we mean. We again went back to 1900 and measured the return in the Dow during every November-April period. We then broke the returns down by those less than -15%, those greater than 20% and all 5% intervals in between. We then looked at the subsequent May-October periods to see if there was any correlation.
As it turns out, negative November-April returns had a strong tendency to lead to negative May-October returns (except for the very worst years which typically saw mean reversion). Conversely, positive November-April returns had a strong tendency to lead to positive May-October returns. And the historical sweet spot for May-October periods came following November-April returns ranging between 15%-20%, e.g., this year.