The Trouble With ‘Sell in May and Go Away’

As if investors need any more reasons to be gloomy these days, another is likely to start hitting headlines in the coming days: The calendar’s turn from April to May. “Sell in May and go away” is an old investing adage that encourages people to sell their shares in May and return in late autumn. It started out as “sell in May and go away, and come back at St. Leger Day,” citing British stockbrokers’ tendency to holiday from May until the St. Leger Day horserace in September. Conventional wisdom said market liquidity was lower then, potentially bringing weaker returns, hence investors were supposedly best off staying out of the market. In the following decades, it morphed, with many market historians pointing to equity markets’ history of lower average returns in late spring and summer than during autumn and winter, which we will show below. This year, we suspect the bear market (fundamentally driven broad equity market decline worse than -20% from a relative high) will heighten the chatter, with “Sell in May” touted as a way to protect yourself against the a possible second leg down. In our view, it is impossible to know what markets will do this summer—short-term volatility is unpredictable—and whether 23 March is the bear market’s low is unknowable today. However, we think seasonality is a poor reason to cut equity exposure, especially now.

For one, we think summer months’ weakness has always been overstated. The most common version of “Sell in May” that we encounter in financial commentary focuses on returns from 30 April through Halloween. Those are indeed weaker, on average, than returns in the other six months. Yet at 2.8% since reliable data for the FTSE All Share Index begin in 1933, they aren’t negative.^[i] They just aren’t as positive as the 9.8% from Halloween through 30 April.^[ii] Yes, returns in individual years vary greatly, and the dataset spans bear markets as well as bull markets. Accordingly, don’t be shocked if pundits cite the dismal 30 April – 31 October returns from 2008 as evidence you should sell now.^[iii] Perhaps they will bolster this with 2001 and 2002’s back-to-back sizable drops during the subpar 30 April – 31 October stretch in the bear market that accompanied the Tech Bubble’s implosion.^[iv] In our view, the trouble with that logic is that those bear markets, like the one that began in February, had fundamental causes unrelated to the calendar. Our analysis shows their extended length and multiple downdrafts had nothing to do with the calendar, either. Coincidence isn’t causality.

At the simplest level, we think selling when May arrives and sitting out the summer likely results in one of two scenarios: You miss the bear market’s second leg down, or you miss its V-shaped recovery. (Yes, there are other potential scenarios in between, such as a flattish spell or a second downdraft and quick recovery, but we are simplifying it for illustrative purposes.) If this bear market has material, longer-lasting downside ahead, it will likely have a fundamental cause, as our research shows all other prior bear markets’ second and third waves have. Correctly identifying that before the general investing public does—and, hence, before markets have an opportunity to factor it into pricing—is crucial in any decision to reduce equity exposure, in our opinion. Absent that, our analysis of past market returns suggests selling just because summer months in bear markets have been awful—or because summer returns are lower on average—has the potential to be a costly error.

To see why, consider the other what-if: that this summer features a strong equity market recovery. We aren’t making a short-term forecast or declaring the bear market over, but if you owned shares through an awful March, we think the payoff for that decision is the strong returns that have historically arrived early in a new bull market.^[v] Those gains compound throughout the rest of the ensuing bull market. Miss them, and your overall long-term return weakens, which could set you back from your goals if they necessitate achieving market-like returns over your investment time horizon (the length of time your assets must be invested to meet your needs).

Selling in May might have proven beneficial during the past two bear markets, but there are other related times when historical returns suggest it would have been a grave error. In 2009, May arrived less than two months after the bear market’s low, amid widespread belief that the nascent recovery was built on sand—a lot like the commentary we see today. But a new bull market was indeed underway, and the FTSE All Share Index gained 21.2% from 30 April – 31 October.^[vi] The 2000 – 2003 bear market ended in March, creating a similar scenario as the FTSE All Share returned 14.2% between 30 April and Halloween in 2003.^[vii] In other words, Sell in May hasn’t worked on a number of occasions when bear markets had been occurring—and weren’t known to be over.

So we encourage readers to tune out the calendar and chatter about seasonality. Our research shows equity markets care about a lot of things, but the month isn’t one of them. Whatever shares do over the next six months and beyond will likely rest on investor sentiment and whether that is too dreary or hot relative to what is likely to happen over the foreseeable future. Focus your analysis there, look forward, and avoid the temptation to act on widely known information.