Sell in May and go Away but buy back on St. Leger Day

May 15, 2008 8:22 AM

Academic discussions of stock markets frequently reference The Efficient Markets Hypothesis; an idea that share prices are fairly valued, their prices reflecting all available information. However folklore such as "Sell in May and go away", which proved prudent in 2007, clashes with this theory.

The history of the equity markets is long, and like any discipline with a rich history there is a deep trove of lore and near superstition adhered to by many market participants. But sometimes such folk tales are true.

For example, The Weekend Effect, documents negative a correlation between Friday and Monday returns. Some believe the Weekend Effect is driven by short sellers purchasing shares Friday afternoon to close open positions. But others speculate this is caused by the corporate practice of "burying" bad news by releasing it on Friday afternoons after the markets are closed, thus leading to a Monday morning sell off. Regardless of the cause, The Weekend Effect has been observed in most G20 stock markets, and in some cases for almost a century. Schwert (2002) looked at The Weekend Effect in detail [.pdf].

While there are other market phenomenon linked to days of the week, we also see large number of what are called "calendar effect" anomalies. These tend to be observed at or during specific times of the year, and once again cut across national and cultural boundaries.

Consider The January Effect, which refers to the tendency of the equity markets annual returns to follow the results of the first five trading days in January. In other words, if the market finishes the first week of the year on a positive note, so will the entire year and vice versa. Haug & Hirschey (2006) conducted a very detailed analysis [.pdf] of the January Effect.

Lest one think stock traders are cold hearted capitalists, The Santa Claus Effect gives all good (actually only the longs) equity market participants a present in December in the form of an end of year rally and higher share prices. Some, however, believe this effect reflects nothing more than end of year buying as tax exempt vehicles (e.g., an IRA) are capitalised, or bonuses are received and invested.

Halloween brings seasonal market advise - "Sell in May and go Away but buy back on St. Leger Day". A phenomenon that not only has been observed globally but also documented in England since 1694. Adherents believe in liquidating share holdings each May, not re-entering the market until "St. Leger Day", a date in late September which refers to the running of a horse race at Doncaster in England. Jacobsen & Bouman (2001) studied this effect globally [.pdf].

So how about 2008? Well many will easily recall last summers carnage, with The Dow, the S&P500 and the NASDAQ all losing hundred of millions of dollars of shareholder wealth under much more benign economic conditions.

With the S&P 500 at a relatively low price to earnings ratio, The Fed pushing rates down to 2%, commodity prices trending higher, the housing market sharply down, the American economy slowing and the US Dollar trading at record lows - the summer of 2008 as well as the autumn will no doubt prove very interesting.

-------------
Complete citations to papers referenced
Haug, M., Hirschey, M., 2006, 'The January Effect', Financial Analysts Journal, Vol 62, No 5
Jacobsen, B. , Bouman, S., 2001, 'The Halloween Indicator, Sell in May and Go Away: Another Puzzle', Massey University Working Paper
Schwert, G., W., 2002. Anomalies and Market Efficiency, Handbook of the Economics of Finance, pages 937-972

posted by Mutant (11 comments total)

This FPP presents a brief overview of the totally fascinating field of Behavioural Finance. While much attention and many resources have historically been focused on quantitative modeling, finance hasn't much addressed the human element, with the inevitable consequences.

We are interested in these phenomenon as they are almost always global, and not easily explained away in terms of local market regulatory or cultural considerations. Truth of the matter is, we simply don't know why these phenomenon exist. But they do.

A word of warning: don't invest your money trying to trade these phenomenon, for several reasons.

First of all, the problem with trying to time the market is dual; picking both a time to sell and a time to buy back in. Historically, the best equity market returns are captured in a handful of trading days, and if you're out of the market you've missed the opportunity.

Second, any trading strategy driven by time / date considerations will have transaction cost and tax implications that buy and hold investors don't encounter. If the gains on every trade aren't sizable, over a sufficiently long period of time and large number of transactions these costs will rapidly erode profits, and in fact my eat into one's capital.

Finally, while these phenomenon are statistically significant, most of the documented effects are relatively weak. Large sums of money would have to be devoted to such strategies and, truth be told, market participants such as banks or hedge funds, capable of raising such capital, tend to pursue more arcane, more profitable and much, much less risky trading strategies.

Bottom line: you aren't going to make any money on any idea presented in this post; in fact you are very likely to lose money. Leave stock trading strategies to the professionals, purchase a no load, low cost index tracker, invest for the long term, then sit back and watch all the excitement.

You've got better things to do with your time.
posted by Mutant at 8:24 AM on May 15, 2008

You've shorted a bunch of stocks, huh?
"Hmm.....How can I get a lot of people to sell?...."
posted by salvia at 8:32 AM on May 15, 2008

Dude, how did you miss links to BehavioralFinance.net?

- Intro to Behavioral Finance (PDF)
- History of Behavioral Finance

Since best place for behavioral finance on the internet. Well sorted and with citations!

That being said, my pet theory is that quantitative finance and behavioral finance are not exactly different things. At least behavioral finance doesn't give the illusion that it is engineering, and can at least better express the unseen data. Interesting field but still completel in its infancy. For every intelligent paper I read in it, I find comes really close talking about time cubes. Perhaps that's a bit far, but it sort of reminds of the scene from Annie Hall where Woody Allen is obsessing over photography as an art form without an established set of aesthetic crtieria, which is where we sort of are at with behavioral finance.
posted by geoff. at 9:04 AM on May 15, 2008

reflecting all available information

Under the Weak- and Semi-strong form of EMH prices reflect all available public information.

Also, if your investment horizon is great enough, you shouldn't necessarily pay attention to short-term trends/anomalies. Interesting, nonetheless.
posted by preparat at 9:11 AM on May 15, 2008

The funny part to me is that it rhymes. Names like "The Efficient Market Hypothesis" are more what I expect. But "sell in May and go away?" What about the slightly less well known maxim "Buy In July And See Profits Fly?"
posted by salvia at 9:38 AM on May 15, 2008

"Put When Plants Root, But Call In The Fall"
posted by salvia at 9:38 AM on May 15, 2008

I'm well versed in the "sell at the bottom, broke by autumn" school of investing.

Excellent post, btw. Thanks.
posted by googly at 10:01 AM on May 15, 2008

I redirected my pension from stocks to treasury bonds back when the Dow Jones hit 10,000 (April 1999). I imagined some sort of "big number" effect would occur. People would push up until they got over the big number and then the number would define rallies and sell-offs. So far I haven't been disappointed. The Dow Jones has grown 3% per year since. The NASDAQ was at 2600 in April 1999, so it has actually fallen.
posted by dances_with_sneetches at 11:02 AM on May 15, 2008

Excellent post Mute.
posted by zemblamatic at 11:04 AM on May 15, 2008

Let us not forget the wisdom of the private financeer S. Jackson, who wisely let us know, "Buy commodities futures in June, corn prices be high soon."
posted by adipocere at 1:40 PM on May 15, 2008 [1 favorite]

"Don't buy the stock if you can't take a shock."
posted by salvia at 4:27 PM on May 15, 2008