Sell In May and Go Away? History Says No

Hickey and Walters (Bespoke) submit: If you have watched, read, or listened to any market commentary over the last twenty four hours, you undoubtedly have encountered a story discussing the Wall Street adage, Sell in May and Go Away. After the January Effect and Santa Claus Rally, this is probably one of the most widely discussed market axioms. As the chart below shows, the market, on average, is certainly weaker beginning in May and through the summer.

But just because it is weaker does not necessarily mean it tends to go down. The table below the chart shows the percentage of time the market rises from May 1 through September 1 over various time frames. Over each time frame covered, the market has a positive return at least 60% of the time.

As far as the likelihood of a significant gain is concerned, history tells us that the Dow is twice more likely to have a 5% gain over the next five months than it is to have a 5% decline. Since 1929, there have been 30 years where the Dow went up more than 5% between May 1st and September 1st, while there have only been 14 years where the index declined by more than 5%. There have been 14 years where the index went up more than 10% versus only 8 occurrences of double digit declines.

So while the Dow typically does not knock the cover off the ball over the next five months, it is not necessarily the time to go away.

click to enlarge

Comments

[S A:]

It is quite amazing - the charts definitely show two different markets:

- The one between 4/30 to 10/30 which is clearly flattish/down

- The other between 10/30 to 4/30 - in a clear and definite upward trend.

So the data clearly DOES support the "Sell in May and go away" adage.

Why the misleading headline - "History says No"? The data says "YES".

SA.

2007 May 02 01:42 AM

[SA Editor J...:]

SA,

Not sure what's misleading about the title? If you read Hickey and Walters' analysis, it clearly explain why historically investors have been better off holding than selling in May, with positive gains occurring more than 60% of the time and the likelihood of significant gains more than double that of significant losses.

JL

2007 May 02 03:31 AM

[Chris G.:]

The article is misleading because you don't just "sellin May" and "go away forever"...the rest of the thought process generally follows a return to the market after "traditional september/october lows."

So yes, the idea that you sell in May, then return in October might actually have some data backing it.

2007 May 02 08:53 AM

[ishvara:]

Very confusing! The article speaks of a five month period, but states the five month period to be May 1 to September 1, which is four months. Since September is the worst month of the year, it is not an idle confusion. Do the statistics include September of not? Ishvara

2007 May 02 03:35 PM