PsyFi Search

Wednesday 7 April 2010

Do Behavioral Funds Work?

No. Maybe.

For all the impressive research on behavioural finance the acid test is ultimately whether it can yield better returns for investors. A few intrepid souls have set out to discover whether there’s any evidence that investing along behavioural lines can produce such returns and have re-emerged from the mutual fund jungle, somewhat battered and worse for wear, with the broad answer of, err, “No, it doesn’t”.

Of course, there’s a bit of a puzzle behind the research because the concept “behavioural investing” is somewhat amorphous. It would be unsurprising, given the parasitic mutational qualities of the financial sector, if the rise of behavioural finance didn’t attract managers who see it as the next destination of hot money. It’s enough to make your head spin: behavioural finance itself could be the next behavioural finance anomaly.

Behavioral Funds

In Behavioural Finance: Are the disciples profiting from the doctrine? some researchers from Florida State have taken a hard look at the rise of the behavioral mutual fund to try and figure out if these vehicles are providing investors with abnormal returns: that is, returns above and beyond what you might expect to get from simply investing in the market. The idea that the subject is a religion, its promoters disciples and the approach doctrinaire seems to beg the question a bit, but as an inveterate question beggarer myself I reckon we can reasonably skip that and look, instead, at the data.

Now, the idea of a behavioral fund is a bit dubious in itself. Behavioural finance research shows that markets exhibit anomalous behaviour unpredicted by standard economic models while also showing that investors are predictably irrational in a variety of peculiar ways. The research has also been able to show correlations between these findings and thus explain various market anomalies. So far, so good.

From Market Anomalies to Investor Irrationality

However, we’ve also seen that merely pointing out these anomalies is often enough to cause human behaviour to adjust to make them disappear. The idea that markets are adaptive is a strong theme in finance at the moment, and the evidence is compelling if not yet overwhelming: one of the problems of adaptive models is that they can pretty much explain anything which, if not controlled for carefully, means they explain nothing. Still, it very much looks like changes in human behaviour can explain extreme changes in markets.

This, however, is a long way from being able to use behavioral finance to directly make money out of it. Individuals, recognising their own psychological weaknesses, can undoubtedly improve their investment performance over extended periods: even if this merely means moving from a scattergun activist approach to a more passive buy and hold one. There’s also the intimation of an idea that rational investors may be able to detect the various extremes of irrationality: unfortunately it seems it may actually be quite difficult to make money by betting against the markets at peaks because the market will have been captured by zealots. It’s probably a bit easier in the troughs.

However, peaks and troughs are extreme events, by their very nature. Despite the nasty reality that extreme events occur far more often than classical models predict they’re still relatively rare. Mostly markets wiggle about between extremes and alpha seeking fund managers – that happy breed who seek to show that they can produce risk-adjusted returns in excess of the market while drawing their fees regardless of performance – must live with this environment while trying to make a turn. So a behavioral fund will somehow have to find a way of making money from human misbehaviour during normal times.

More Investment, Less Returns?

When the researchers from Florida examined a list of sixteen funds they identify as following behavioral principles they found that these funds were successful at attracting excess investment dollars, suggesting that the buzz around behavioral finance is penetrating some part of the mass investor psyche. It’s interesting to note that the researchers generally base the fund starting date from the arrival of a manager known to be associated with behavioural finance: nothing wrong with that in research terms but highly interesting in respect of the relationship between investment funds and star managers.

Moving along, however, we discover that the researchers find no evidence of abnormal returns in these funds although they do find that they’ve outperformed the S&P500. What’s really interesting, however, is that the reason for this outperformance appears to be a loading on the French and Fama HML factor. HML – High Minus Low – is a value measure, essentially indicating the premium to be obtained by investing in stocks with a high book to market value over their supposedly more growth oriented cousins. In straightforward terms the behavioural funds appear, over the period measured, to be emulating value funds.

A Behavioral Fund Bubble

Now, of course, this isn’t conclusive. Over the period in question value funds did outperform, so you’d kind of hope that behavioral funds would track this outperformance. The acid test is that by the time value starts to underperform then the behavioral funds should have switched tack and be following whatever the alternative theme is that is attracting the crowd. Quite how behavioural finance helps you develop foresight of this kind isn’t clear, but perhaps if the rise of behavioural finance continues then a good bet would be to start tracking behavioural funds. A behaviourally induced bubble in behavioural funds would, at least, be a mildly amusing artefact of the whole, strange situation.

It’s genuinely impossible to know whether funds labelled as being “behavioral” are somehow able to implement a strategy that will outperform other approaches over significant periods of time. In fact we don’t even know if these behavioral funds are actually operating on the principles of behavioural finance (whatever they are). As the researchers put it:
“The mere statement that a fund uses behavioral finance in its investment decisions offers no guarantee that it really does. Also, it is possible that the only funds admitting the use of behavioral finance are those that are prospering from it. There could be a larger group of funds using behavioral finance but not talking about it because of poor performance. However, we believe this potential selection bias strengthens the main conclusion of our paper.”
Alice in Behavioural-land

Well, maybe. Or maybe those funds that use behavioral finance but don’t admit it have figured out that if they talk about it that it’ll stop working? On the other hand, if behavioural finance is the next hot sector it’s kind of hard to believe the mutual fund industry would be shy about publicising this. Also underperformance might be because there are genuine behavioral strategies which work, but which can’t be implemented due to costs – nice clean academic theories are often sunk by nasty real-world facts.

Still, what we do know is that hot fund sectors which attract excess amounts of investor money tend to lead to investor underperformance: if only because investors who are attracted to hot money tend to be serial flippers of funds. It’s all horribly Alice in Wonderland-ish but the acid test will, as usual, be time and investors' capital.


Related Articles: Save Our Short-Sellers, Finance: Where the Law of One Price Doesn't Apply, When a Dollar's Not Just a Dollar

6 comments:

  1. Hmm...

    It it one thing to posit that investors are irrational, and that we can beat the market by avoiding (some of) their irrationality.

    It is quite a different thing to say that we can predict the timing of their irrationality, and profit from it directly.

    ReplyDelete
  2. Robert Wasilewski7 April 2010 at 12:04

    I wonder where you get "...unfortunately it seems it may actually be quite difficult to make money by betting against the markets at peaks because the market will have been captured by zealots. It’s probably a bit easier in the troughs."
    It seems to me to be the opposite - easier to take money off the table in roaring bull markets than to step in when the market is falling off a cliff.
    I know many now look back comfortably to 3/9/2009 and say what a wonderful opportunity it was but at the time I recall people running for the exit. Advisors were on the phone all day long begging clients to hang in a bit longer.
    The market very easily could have dropped another 20% at which time I believe there would have been a much more massive capitulation than we actually saw.

    ReplyDelete
  3. My view is that people over-analyze this question and thereby miss the obvious (yet rarely noted!) practical point of Behavioral Finance. The purpose of Behavioral Finance is to learn about investor irrationality and to take it into consideration when making investing decisions. Investor irrationality always evidences itself either in overvaluation or undervaluation. For investors to value stocks improperly is irrational. By definition. So to invest rationally one must respond in some way to the improper valuations.

    The obvious way to profit is to lower your allocation when prices are insanely high and to increase your allocation when prices are insanely low, thereby keeping your own risk profile roughly stable (this is "Staying the Course" in a meaningful way). This strategy has shockingly enough generated dramatically higher returns at dramatically less risk for as far back as we have records of U.S. stock performance (1870).

    Ssssh! Don't tell anyone!

    Rob

    ReplyDelete
  4. valueinvestor20107 April 2010 at 17:45

    I was just curious, can automated systems be used to test the behavioral theory?
    For example let us say we give $1 million to a particularly emotive fund manager and another to an automated investment system?
    The fund manager is told purchase High ROE,Low Debt, Low EV and Good Growth stocks with solid cash flow(typical value criteria) in an implicit manner. For the automated fund this is coded in an explicit manner.
    After time period x we come back and check the performances of both. The human is the experiment group and a system being emotionless is the control group? All the difference is likely to be mostly attributable to the behavior pattern of the human.
    Do you think such a process would help in testing and identifying the impact of behavior?? Is there any research stuff on SSRN etc on this??

    ReplyDelete
  5. Yeah, um, that's the thing: it's not an easier investing in an inefficient market. And the cognitive biases underlying much of behavioral finance are stronger than most realize. I mean, we're talking human nature here, so don't kid yourself: it's not nearly enough merely to be aware of them. I'm afraid it's gotten worse actually. Everyone still refuses to acknowledge uncertainty, ambiguity, etc. Even worse, the average Joe these days has deluded himself into thinking he is financially savvy and that education is overrated cause he now understands him some derivatives CDS contracts (they're just like insurance!).

    If you think markets are inefficient and question assumptions in your sleep, then take a breather. If you think that humans are generally irrational when making investment decisions, then fine. Go out and do some research and handpick you some winners (and screw the whole diversification bullcrap!). Unfortunately, my guess is you're destined to discover that it's no easier investing in an inefficient market. Also, be careful, and remember, no one predicted the last financial crisis, including yourself. Therefore, you won't see the next one coming either.

    Reminds me of a quote from Newton, "I can calculate the motion of heavenly bodies, but not the madness of people."

    ReplyDelete
  6. Yep, it isn't necessarily easier investing in an inefficient market and, so far, behavioural finance doesn't make it easier. It helps explain why it all goes wrong, but it doesn't help predict what's going to happen next. Indeed it may be impossible to do so because some uncertainty is never reductible.

    The idea about peaks and troughs and zealots and so forth is more or less the one I set out in the article on shorting: when markets go up it can be very difficult to reverse them because everyone who holds stock believes it will go up and there are limits on how much risk short-side investors can bear. This asymmetric skewing of risk also applies on the downside, so it's (theoretically anyway) easier to reverse a falling market than a rising one.

    ReplyDelete