Frogs and Philosophers
In Games People Play we alighted upon the germ of an idea that bears further investigation – that although individual investors may themselves be as mad as a box of frogs on mescaline their combined behavior can somehow result in relatively rational markets. The idea is that, somehow, rational markets arise out of irrational traders doing stupid things.
Odd though this may seem as an idea the possibility of new properties emerging from systems made up of complex component parts isn’t new, although it does serve to make a lot of scientists and philosophers extremely annoyed. Which is a good reason for pursuing the idea, is it not?
Most science has been built on the idea of reductionism, breaking down complex ideas or systems into their component parts in order to understand the underlying structure of things. Reductionism is one of the most powerful mental tools humanity has yet created and like so many great ideas was first postulated by the Ancient Greeks and then turned into a useful technique by one of the great scientist-philosophers of the Enlightenment; in this case René Decartes.
Decartes is a towering figure in the history of science, and the fact that he was wrong about almost everything he set about understanding doesn’t change that. It’s a salient lesson in an age where being right is usually regarded as the only touchstone of success, that being wrong in the right way can often lead to more interesting, and far reaching results. So when he wasn’t thinking himself out of existence or arguing that animals were mere machines or that the universe was purely mechanical, Descartes was busy developing a philosophy of thought that still guides much of scientific endeavour today.
Yet despite the power of reductionism there’s a strong argument that it is ultimately a limited technique because there are properties of complex systems that can’t be deduced by repeatedly decomposing them into smaller and smaller units. So somehow properties that are greater than the sum of the whole emerge from these systems: hence they’re known as emergent properties.
Summarising the taxonomy of emergent properties would take a lifetime, but one example may suffice to express the idea: consciousness. After all, why tackle small problems when you can tackle big ones? So there is an argument that consciousness can’t be detected in the neural pathways and synaptic junctions of the brain, and can’t be found by further reducing these to minerals and chemicals: it’s not hidden in the atomic composition of our nervous systems, but is a spontaneously emergent property of a brain that got complex for other reasons, most of which probably involved competition for mates. Sex is the usual explanation for most things we can’t understand.
This isn’t the only explanation, mind you. There are as many explanations for consciousness as there are behavioral biases, although my particular favorite is that it’s a learned behavior: it’s a cultural adaptation to modern life. And if you don’t believe a modern human can function without being conscious then you’ve obviously not had much exposure to teenage boys.
However there’s a strong school of thought that the concept of emergent properties is simply an output of the meandering minds of weak-thinkers, who simply haven’t grasped the idea that reductionism has conquered every other area of science, and will one day reveal the answer to the problems we’ve not yet solved. Of course, this doesn’t quite explain how the idea of emergent properties has emerged, but we’re beginning to stray from the point; which is can a system comprised of palpably irrational investors exhibit behaviour that looks rational?
Well, unless you’re completely new here you’ll already have spotted the heavy handed signalling which suggests that the answer is possibly yes. This comes out of an absolutely fascinating piece of research conducted by John List and Daniel Millimet entitled The Market: Catalyst for Rationality and Filter of Irrationality.
In the Field
We’ve met List before, and he consistently produces research which challenges the boundaries of what we think we know about investor rationality and the reliability of the findings of behavioral economics. His speciality is field experiments – finding naturally occurring environments in which to conduct research as a counter to the unnatural laboratory conditions in which most psychology experiments take place (see: The Death of Homo economicus). As we’ve also previously seen it’s amazingly difficult to get people to behave naturally in labs, because they tend to try and figure out what the experimenters want in order to please them – which means that the experimenters have to go to inordinate lengths to disguise what they’re after (see: Be a Sceptical Economist).
List’s work, on the other hand, is based outside of the lab, which makes it difficult to obtain the repeatable conditions normally demanded of a scientific experiment, but which also makes it easier to get people to behave like real people; largely because they don’t feel that they’re being treated like a rat in a maze. The List-Millimet research uses this type of approach to investigate whether or not participants exhibit rational behaviour and whether or not rational behaviour is influenced by experience in the marketplace.
Their test of rationality is preference consistency – the idea that people have stable preferences over time is one of the main tenants of rational choice theory, but the possibility that people may modify their preferences by learning from the market is a challenge to it (see: Economics and Psychology:The Divorce). The researchers found:
“Empirical results generated from observing more than 800 experimental subjects indicate that (i) only about 31% of agents exhibit behavior consistent with rational choice theory, and (ii) market experience facilitates the development of such behavior.”
Which is interesting, although not that surprising based on other research we’ve seen; after all if investors were rational I’d be wasting my time here. However, a critical point that List and Millimet make is that this implies that economic rationality isn’t based on the individual’s self-interest but is a social construct, arising out of interactions with other investors through the medium of the market. It’s a key observation, and one that points a dagger at the heart of standard economic theories (see: Arbitraging Embeddness and Facebook Friends Can Make You Poor and Happy People Make Terrible Traders).
This is interesting enough but, the researchers then go on to ask an even more interesting question: does this learning process facilitate the creation of rational market outcomes out of irrational investor behaviour? Well, the outcome is off-the-scale interesting:
“Our findings suggest that even in markets populated entirely by irrational actors, several fundamental features of markets, such as price and quantity realizations, meet neoclassical predictions after a few rounds of market experience.”
How bizarre is this: irrational actors, operating on the basis of market experience end up creating what looks like a rational market. Market efficiency, in this view, is an emergent property of social interaction, not the outcome of multiple rational actors independently behaving sensibly.
This, of course, would explain a lot of behavioral biases: if social interaction is the medium of the exchange of rationality then if it goes wrong, as it often does in real life when messages get mixed, or there’s an influx of irrational actors or information gets garbled in translation what you will get is a decidedly inefficient market. People aren’t behaving any more or less irrationally than they normally do; it’s just that their environment has gone awry.
It’s a fascinating idea, and a clever piece of research. Even better, it was carried out on children in shopping malls. Who says economics is always divorced from reality?