PsyFi Search

Thursday, 23 July 2009

O Investor, Why Art Thou Rational?

Rationality Good, Irrationality Bad

One of the revealing things about a lot of modern financial theory and economics is the underlying assumption that rationality is obviously good. We’ve moved on a little bit from the belief that we’re all omnisciently perfect investment analysts but the ideal of the perfection of rationality haunts us still, like a ghoul in a very wonky machine. Imagine Béla Lugosi in a Ford Edsel.

However, we’ve somehow built a global society without being particularly rational most of the time. A significant proportion of the world’s population manages to simultaneously combine a belief in a benevolent yet invisible deity it's worth killing for with the imminent prospect of abduction by sex-obsessed aliens. Yet we’ve managed to probe the fabric of the universe, develop great works of art and keep most people fed, most of the time, through the magical wealth creating properties of the global economy. So what need has humanity for rationality when it’s managed to get this far without it?

Sloppy Neurality

Our sloppy thinking is due to design – the fabulous parallel processing capability of the brain comes at the price of reusing lots of bits of neural pathways for different purposes. This means that although we’re pretty good at processing lots of information we tend to do so inaccurately. Which doesn’t matter a lot of the time – we can perform amazing acts of computation practically instantaneously in order to do important things
in spite of our neural sloppiness. For instance, we are the only animal that can watch football while walking upright and carrying a tray of beers and a mound of nachos simultaneously. Of course, we're the only animal that would want to do this, but that's a whole 'nother ballgame.

This reuse means that we take processing shortcuts which lead to mistakes in areas where rational thought would often come in handy. Such as deciding whether to cut someone open and jiggle their innards around with a semi-circular saw. Or whether to send them to war against battle tanks armed with hand-held tin openers. These processing shortcuts lead us to assume such things as that attractive people are always intelligent (they’re not), that any short-term trend will continue indefinitely (it won’t) and that we can foresee what’s yet to happen (we can’t). In all of these cases proper rational analysis of the evidence might lead us to make different decisions.

Rationality in Decision Making

There are cases where it’s obvious that rationality is needed in decision making. Surgeons who keep on performing pointless operations, generals who carry on sending troops into hopeless battles and politicians who continue pursuing reckless economic policies are only a few of the situations where a bit more rationality would lead to a lot less unhappiness.

Despite this human irrationality quite clearly hasn’t stopped us being a decently successful species - only 100 million years or so to go now, before we match the dinosaurs. Our question, though, is whether a dose of rationality is beneficial for investors. There’s plenty of behavioural finance theory that shows how our sloppy processing leads to irrational financial decisions – giving different answers to the same investing question when it’s phrased in different ways, refusing to sell hopeless shares because they’re trading below our anchoring purchase price, buying stocks because they’re the popular flavour of the moment rather than because of their investment fundamentals and so on. The list of these mistakes is almost infinite.

Irrationality Can Be Rational

Yet investing isn’t purely a personal affair, it’s a social one. Making an irrational investing decision doesn’t happen in a vacuum, because lots of other people are making similar irrational decisions. So a rational investor can end up betting against the crowd. Is it really rational for a social creature to step outside of the group and expose themselves to what lies beyond?

Although we can surely agree that the behaviour of financial institutions in lending money they couldn’t afford to pay back over the past few years was wantonly irrational it’s not entirely clear that the same is true of the officers and employees of those institutions. If you were heavily rewarded for flogging dodgy loans with no possible comeback what would you do, rationally?

It’s fairly easy to show that what may be irrational for one group of people is perfectly rational for others and this is a real danger for investors. Consider, if fortune tellers and psychics were any good they wouldn’t be writing dodgy horoscopes, they’d be staggering out of Vegas, weighed down by their foresighted winnings. Similarly, if investment fund managers were genuinely able to produce excessive long-term returns they wouldn’t be running investment funds, but would be making fortunes on their own behalf.

Logically and rationally this suggests one of two things. Either fund managers don’t think they can outperform the markets or they think the risk of trying to doing so isn’t worth forgoing the rewards they can obtain by staying in place. Rationally, therefore, private investors shouldn’t invest in their funds. The fact that hundreds of millions of private investor money does go in this direction is evidence enough that investor irrationality does matter.

Professional Exploitation of Irrationality

There’s worse to come since there are people out there deliberately trying to find ways of profiting from the irrationality of investors. This has been done before in an off-hand kind of way. Betting against so-called odd-lot investors – small investors buying odd-sized packets of shares – has long been considered a way of capitalising on amateurs’ involvement in markets.

Now, however, the professional investment industry is putting lots of time, effort and money into developing automated models to take advantage of the findings of behavioural finance. So called Quantitative Behavioural Finance aims to develop models that take advantage of the inherent biases and irrationality of markets. Alan Kirman came up with a model back in the 1990’s that was able to generate suspiciously realistic market behaviour simply by including two sorts of investors – fundamentalists, who invested based on the basis of valuation – and chartists, who invested on the basis of market direction.

The terms ‘fundamentalist’ and ‘chartist’ are overloaded with various meanings but roughly we can equate the former with rational investors trying to take advantage of market mispricing and the latter with irrational investors trying to take advantage of market psychology. Interestingly Kirman finds that generating realistic looking markets depends on the possibility of chartists converting to fundamentalists and vice versa, dependent on market conditions. The ability of people to be rational at some times and irrational at others is well attested to: virtually all financial bubbles are evidence enough of this. Indeed it may be perfectly rational to be a fundamentalist at times and a chartist at other times – the problem is figuring out when to make the switch.

Boring Rationality

Clearly investment houses would love to find models that effectively trade on these rationality deficiencies and one can only speculate on what impact this might have on markets. That some existing models increase market fluctuations by exacerbating the upswings and downturns now seems to a common theory. Whether quantitative behavioural models would damp down these swings or simply cause different kinds of trouble is unknown, but the evidence of history isn’t encouraging.

Keynes is supposed to have said that markets can remain irrational longer than investors can remain solvent, and there’s more than a grain of truth in this. However, unless investors can manage to stay fairly rational it’s hard to see how they can ever make any money on the markets. The alternative, as ever, is to remove oneself from the equation and simply invest in an index tracker. Rationality is not always interesting, but it’s often better than the alternative.

Related Articles: Perverse Incentives Are Daylight Robbery, Unemotional Investing Is Best, Don't Lose Money In The Stupid Corner


  1. C.K. Chesteron wrote about this. He said that the entirely rational man is the man who eventually goes entire insane. We're on the way there with the investing advice that we have gotten behind over the past 30 years.

    We are not 100 percent rational creatures. Life isn't 100 percent rational! It is the greatest failing of most economic theories that they try to squeeze human behavior into rational boxes in which it can never fit.

    It is irrational to pretend that humans (or the markets they build) could ever be 100 percent rational!


  2. The problem with this post is that the rationality posited by, for example Harry Markowitz is really quite limited and much less than the broader concept of rational. The only requirement in Markowitz rationality and much of economic utility theory is that transitivity not be violated with respect to preference functions. That is if I prefer A to B and B to C then I also prefer A to C. Now collectively one can construct voting schemes where transitivity is violated (see Ken Arrow and his Nobel Prize) and that is clearly part of the problem with democratically allocating resources. But for most individual actors the condition of preserving transitivity is not that much of a stretch with respect to determining consumption preferences and risk preferences. Clearly humans are not rational in the broader sense but in the narrow sense posited in utility and expected utility theory it is not a ridiculous assumption.

  3. On Markowitz, to make portfolio theory work correctly you've got to estimate expected returns, variances and covariances correctly. Get any of those wrong and your results are off with the birds. That's a much wider definition of "rationality", of course, but it's the only one of interest to investors as opposed to economists.

    Economists have been well aware that humans aren't rational in the sense of utility maximisation for over forty years. The problem is doing something about it: behavioural finance is like an annoying younger brother who can tell you you've done your homework wrong but doesn't have a clue how to fix it. So mostly economists have tried to introduce human irrationality into existing models that require rationality. It's an uneasy fit, but in the absence of anything better they're likely to persist.