PsyFi Search

Wednesday 24 October 2012

Seasonally Affective Investing Disorder

Not Not Die Out

Behavioral bias occurs in all sorts of odd ways, but basically can be traced back through evolutionary history to when humanity’s survival depended on deep co-operation and the invention of sewing.  Our ability to cope with the huge changes in weather conditions experienced over our short history defines our species: we didn’t go extinct.

It shouldn’t therefore be particularly surprising that we’re unconsciously affected by environmental conditions: there was a time, not so long ago, when nothing else mattered.  Now, cosseted by central heating and air conditioning, we’re remote from our material world but, it turns out, for investors our material world is not remote from us.

Seasonally Affective Disorder

Norman Rosenthal and colleagues originally described SAD back in 1984, in Seasonal Affective Disorder, A Description of the Syndrome andPreliminary Findings with Light Therapy:
“We define seasonal affective disorder (SAD), which is a condition characterized by recurrent depressive episodes that occur annually … We describe 29 patients who suffered depressions in fall and winter; these depressions remitted by the following spring or summer”.
The research goes on to point out the similarity between the circannual rhythms of animal life and the seasonality of SAD, and speculates that SAD may be an evolutionary hangover from a time when the changes in the seasons, as marked by the varying amount of daylight, provided markers for hibernation and reproduction.  As the paper points out SAD sufferers are twice as likely to have children born in late spring or summer.  Interesting, although hardly conclusive.

Sadflows

Given all of the other types of bias we’ve seen and the issues posed by SAD it  perhaps isn’t so surprising that it turns out to be implicated in some rather odd investor behavior.  However, the strength of the results testifies to a underlying seasonal rhythm which impacts all investors, not just those for whom sunlight deprivation is a health hazard.

Lisa Kramer, Associate Professor of Finance at Toronto, has a track record of fascinating research into the effect of the environment on investors.  So, for example, Kramer, Mark Kamstra and Maurice Levi have shown that US Treasuries have an astonishing seasonal variation in monthly returns.  In Seasonal Variation in Treasury Returns they find an 80 basis point change (that’s a 0.8% change to you and me) between April and October which can’t be explained by (deep breath):
“Macroeconomic seasonalities, seasonal variation in risk, cross-hedging between equity and Treasury markets, conventional measures of investor sentiment, seasonalities in the Treasury market auction schedule, seasonalities in the Treasury debt supply, seasonalities in the FOMC cycle, or peculiarities of the sample period considered.”
In fact, they find that the only convincing explanation is a correlation with the incidence of Seasonal Affective Disorder.  Underlying this is the idea that depressed people, as we’ve seen many times before, tend to be more realistic and are more risk averse: so if we tend to be made miserable by the winter months we’re less likely to optimistically invest (see: Depressed Investors Don't Need Feedback. Everyone Else Does). As the researchers comment:
“Specifically, Treasury returns are significantly positively correlated with the clinical incidence of seasonal depression in North America.”
Daylight Saving  Investing

Rummaging back through the record the paper also alights on a structural break back in the early 1970’s when seasonal effects first came to the fore.  Despite rosy childhood memories it turns out that this wasn’t because the sun used to shine more often and more frequently but seems to be correlated with a change in interest rate setting policy by the Fed.

Prior to this rates were set by the Fed itself irrespective of the market, but afterwards they were increasingly based on market driven auctions: by opening up rate setting to the free market it also opened up the intervention of behavioral bias. So, unsurprisingly the result seems to have been that a more seasonally adjustable rate based less on fundamental economic considerations and more on the hours of available daylight.

How rational are we?

Winter Blues, Winter Reds

In This is Your Portfolio on Winter Lisa Kramer and Mark Weber show that SAD sufferers do display financial risk aversion: they’re significantly (in the statistical sense) more likely to opt for “safe” choices during the winter months than people who don’t have this problem, yet in the summer months there’s no difference in risk aversion between the two groups.  Of course correlation is not causality, but sometime a cigar is just a cigar: further research finds a stronger effect in more northern climates (OK, Canada) where daylight hours are more restricted and a reverse seasonal effect in Australia, where the Northern winter is the Southern summer (although according to my Aussie friends the sun’s always shining).

Of course the idea that a subset of humanity is affected by seasonal effects is an interesting idea, but what’s even more interesting is the suggestion that the whole market is impacted by these issues.  Of course, if SAD only impacts a subset of investors – roughly 6% of the US population, with another 14% suffering from a milder condition – then this raises some interesting questions because the gross effects being demonstrated seem out of proportion to the overall market. Of course, life and investing is not so simple that we should expect a linear relationship; this is behavioral finance.

Non-random flows

So, in general we might regard markets as being fairly random; we all have different views and these views will cancel out, subject to the minor issue that occasionally there’ll be some external factor that will influence investors as a group.  So, en-mass, we’re pretty random.  However, the idea here is that there’s a significant subset of investors who aren’t behaving randomly but are being impacted by SAD-type effects – and they’re all basically operating in the same way.

In a generally random market if a major subset of people suddenly start demonstrating the same non-random behaviour we should expect this to have a disproportionate impact on market movements (see: The Proper Etiquette for Market Panics).  It’s as though 20% of traders all suddenly decided to sell stocks, or buy them, for no obvious externally discernable reason.  And, once you have a significant portion of the market aligning itself and trading in a similar fashion, you suddenly have a non-random influence on market behavior.

Evolutionary Misapplications

Even better, of course, we’ve seen that herding and group effects can cause otherwise fairly sensible investors to abandon their stock screeners and copies of the Intelligent Investor in order to chase the next latest trend: so you might expect to see some kind of seasonal impact on investing fund flows (see: Herd of Investors and Risky Shifts).  And, indeed, in yet further research Kamstra, Kramer, Levi and Russ Wermers have found evidence that US mutual fund investors shift money from equities to government bonds in the fall and send them back the other way in the spring. This doesn’t yet translate into evidence that this genuinely impacts market movements, but it wouldn’t be a great surprise if it did.

The idea that our evolutionary legacy translates into misapplied risk aversion and risk seeking isn't a new one, but the fact that something as simple as sunlight seems to have a major impact on the way investors behave really does emphasise the issues associated with intelligent investing.  To the reams of self-help books and internet screening tools add a UV light.  And a dash of realism when everyone else is getting high on sunlight.

Related articles:

1 comment: