PsyFi Search

Wednesday 30 December 2009

Rock On, January Effect

Born Lucky

Calendar related effects are amongst the most obvious manifestations of the way that collective human mental misbehaviour can impact stockmarket returns. Anomalies like the much observed January Effect, the outperformance of small cap stocks in the US markets during January, point to some kind of systematic bias in the structure of markets.

In fact there’s little doubt that the calendar does impact upon human behaviour but what’s in question is how it does so. Consider, for instance, the odd fact that people born between May and July consistently rate themselves as luckier than those who enter life’s great adventure between October and December. Nothing strange in that, you might think, until you discover that summer babies really are luckier than their winter siblings.

Quirky, Lucky and Wealthy

The psychologist Richard Wiseman specialises in the quirky side of human nature, pulling out peculiar behaviours in an attempt to illuminate the decidedly strange workings of our minds. In various studies he’s shown that people who think that they’re born lucky generally seem to experience more good fortune than others. Disentangling this, as he does in his book Quirkology, elicits some ideas that are both amusing and, then, bemusing.

For instance, a Wiseman experiment got people to rate how lucky they were and then had them scan a newspaper to count photographs. However, half-way through was printed a large announcement offering a financial reward merely for noticing it. Guess what? Yep, the people who noticed this were overwhelmingly from the lucky group.

Hemispherically Challenged

There are probably a couple of things going on here. When the born lucky experiment was re-run in the Southern Hemisphere the correlation between calendar months and luck was maintained, but the order was reversed. So a theory is that this is temperature related – babies born in the winter tend to be swaddled more and have less chance of exploring and grow up to be less social. As child psychologists will tell you – at length – the first few months of life can have a huge impact on the rest of it.

Anyway, people who are less social tend to get fewer opportunities to experience luck, while the experience of the lucky of just being being lucky tends to make them more relaxed and more likely to notice the unusual – hence the results of the newspaper experiment. Lucky people have more open body language and so on and so forth. Basically the lucky people are lucky because they behave lucky and that’s probably because they’re more open to new experiences and have wider social networks – the sort of, so-called, "small world" effect of the Kevin Bacon game described in Sexual Trading. Yet when it comes to winning the lottery they’re no more likely to do so that an unlucky person – real luck is something you make yourself, not an artefact of skewed probability.

Weather, Mood and Returns

It’s possible, therefore, that there are real human biases underlying stockmarket calendar effects. One possible, related, source of bias that’s high on the list of known mood triggers, right next to large quantities of drugs and one’s partner spending the month’s food budget on a nice new outfit with matching polka-dot accessories or a lawnmower that does nought to sixty in six seconds, is the weather. Good weather tends to make us happier, bad weather gloomier. So, the theory goes, good weather should spur investors into splashing out on a lovely new stock or two while storm clouds should leave them glowering and cowering in their non-trading bunkers.

Now obviously that’s far too simplistic an analysis because it’s unthinkable that something as basic as the weather could really make any difference to trading patterns in an industry dominated by large institutions and heavily researched investment models. Obviously.

Well guess what? Hirshleifer and Shumway’s results covered twenty six cities and stockmarkets and showed that stockmarkets do go up more – significantly more (in a statistical sense) – on sunny days than overcast ones. The theory here is that sunlight triggers feelings of optimism and optimistic investors feel, well, more optimistic and go out and buy stocks. Presumably without spending an awful lot of time analysing their decisions.

SAD Investors Are Risk Adverse

Another study investigated whether the seasonal changes that trigger Seasonal Affective Disorder (SAD) may also have an effect on short term stockmarket performance. SAD is caused by decreasing hours of sunlight as summer fades to winter and leads to a range of depression related illnesses. So, the hypothesis was that reducing sunlight might increase investors’ levels of risk aversion and hence decrease appetite for buying stocks. Changes in the seasons – such as lengthening periods of sunlight after the winter solstice at the end of December in the Northern Hemisphere (aka January) – might then trigger changes in investor buying and selling behaviour.

Once again logic tells us that the scientific and analytical process of stockmarket investing can’t possibly be prey to something as obvious and unrelated as the change of the seasons. And once again the evidence suggests logic is somewhere else on a long vacation and has switched off the Blackberry: Kamstra, Kramer and Levi have indeed found a correlation between the waxing and waning of sunlight in different countries and stockmarket returns.

But We Don’t Know Why

However, Jacobson and Marquering while agreeing that there is a strong seasonal effect don’t find any conclusive evidence that this is down to weather or SAD. Indeed, they suggest that applying a simple rule – "Sell in May and Go Away", or the Halloween Effect – start buying at the end of November, works just as well. Roughly, they conclude, there is some case to answer but we don’t really know how to do it.

Looking at the January Effect, Haug and Hirschey come to virtually the same conclusion. The effect is real but the main explanation offered – that this is something to do with investors selling in December and then buying in January due to the end of the tax year doesn’t hold up – not least because the January Effect is observable outside of the USA, in countries with different tax years.

Get Lucky, Get A Calendar

None of this really helps us as investors. Simply knowing that some effect occurs without knowing what causes it isn’t helpful because we have no idea when it will suddenly stop working. If these effects are caused by physiological factors, such as sunlight, then they ought to be robust but if they're simply caused by behaviourally induced triggers then we can be fairly sure that as soon as people start to try and exploit them they’ll start to fail.

Indeed, one of the problems of all such anomalies is that once they become well understood and get built in to automated algorithmic processors they’ll disappear. So perhaps the best thing we can do is to continue to regard them as odd side-effects of human superstition and, while claiming to ignore them, exploit them to their maximum extent while we can.

The possibility that there’s a link between the calendar and stockmarket performance, mediated by some combination of behavioural bias and weather related conditions can’t be ruled out. Richard Wiseman’s research shows that no matter how unlikely we may think it the combination of these factors in odd and unusual ways can change people’s lives. What he’s also shown, however, is that recognition of these issues can turn unlucky people into lucky ones: beginning by taking advantage of bizarre stockmarket effects might be as good a start as any. So rock on, January Effect.

Related Articles: Stocks, Science and Superstition, Don't Lose Money in the Stupid Corner, Your Financial Horoscope


  1. This is the best written money blog out there. You work it hard, Tim.

    Good luck to you in the New Year!


  2. Very interesting - Thanks

    And Good Luck for 2010

  3. I've always thought the January effect is a bunch on baloney.

    However, when tracking my portfolio's returns (I'm usually about 30-40% in small-caps), I've noticed that almost all of my outperformance comes in December and January.

    Based on only 4 years of data, my results strongly confirm the January effect.

    But I still think its a bunch of baloney...