Sell in May - it works

Yesterday, I did something I usually do only twice a year. No, not talk to a girl. I took investment action - namely, switching out of shares and into cash.
This is because the old advice to "sell in May" really does work.  I've estimated (maybe subs reqd) that a seasonal investor - one who was in shares from October 31 to April 30 and in cash the rest of the time - would have grown £100 into almost £78,000 between 1965 and now. £100 left in the All-share would have grown to just over £24,000. Here are average total monthly returns on the All-share since 1965, with standard deviations in brackets.
January = 3.7 (9.3)
February = 1.7 (5.7)
March = 1.1 (5.4)
April = 3.0 (4.9)
May = 0.4 (4.4)
June  -0.1 (4.5)
July = 0.5 (4.6)
August = 1.2 (5.4)
September = -0.7 (6.1)
October = 0.5 (6.0)
November = 1.4 (5.5)
December = 2.6 (4.2)
This raises the question. If seasonal investing is so profitable, why doesn't everyone exploit it and so remove the anomaly? I reckon there are four reasons:
1. Dealing costs. Selling in May and buying back in November is expensive, and exposes you to a capital gains tax liability. This doesn't bother me, as I use my personal pension to make the switch for free.
2. People wrongly trust their judgment. They can always see reasons to hold shares in May, or to sell in November. They commit the availability error, of focussing on available information and forgeting that the background probabilities are against them.
3. Risk aversion. As the nights draw in in the autumn, people get depressed and anxious and so reluctant to take on equity risk. Equally, the better weather in the spring cheers people up and encourages them to hold onto risky assets. Why do you think so many big horse races occur in the spring? Coz that's the time people are most willing to gamble.
4. Absolute risk. The winter is a risky time for western economies (pdf). The UK has gone into recession in every one of the last 30 first quarters of the year; last year, GDP fell by an annualized 15.6%. You don't read about this as the official published figures are seasonally adjusted - but the raw figures are here (the code is BKVT). People who hold shares over the winter are therefore taking on economic risk. Some will be reluctant to do this.
So, I reckon "sell in May" is, for me, an exploitable phenomenon. Points 1 and 2 suggest it is consistent with the efficient market theory (which says expected returns can vary if risk varies). Points 3 and 4 suggest it's a behavioural thing. I'm not fussed.

Comments

[dave heasman]

Didn't work last year, did it? Would you wait until November to buy again? Ledger day is mid-September isn't it?

This looks to be an uncharacteristically good Tuesday, too.

[chris]

Sure, it doesn't work every year. But equally, there's little sign of the anomaly disappearing - 8 of the last 10 winters (Nov-April) have seen better returns than the preceding summers.
And the cliche to "buy on St Leger day" does seem wrong - I think because St leger day falls earlier these days than it used to.

[dearieme]

"Winter" - does it work in Oz & NZ too?

[chris]

It does, dearieme - the link in point 3 shows that the Oz market does relatively well in their winter/our summer.

[Charlie Whitaker]

Chris, would it be impertinent to ask how your pension is set up to allow you to do this?

[chris]

Not impertinent at all. A few years ago, a colleague advised me to take out a pension with the Equitable Life. I misunderstood, and got a unit-linked one rather than the with-profits one that went belly-up.
This unit-linked one allows for many cost-free transfers between various assets - I only use cash and All-share tracker. I'm told that most defined contribution pension schemes do this.

[dave heasman]

Bugger - this year it looks as if you're right, Chris.
(After implementing an Investment Trust valuation system 20 years ago I decided to put my Equitable Life pension into their Investment Trust units. Bacon saved. However, if the dollar keeps plummeting it will reach my stop-loss point quite soon..then back into cash.)

[dave heasman]

So far, Chris, I'm in your (moral) debt to the tune of about £2000. Without your prompting I wouldn't have got out until tomorrow. So thanks.