PsyFi Search

Thursday 13 August 2009

Bacteria, Boids and Market Instability

The Gaps Between People

Old-time economics saw investors as rational individuals, all behaving autonomously in a logical fashion, rather like Mr. Spock umbilically attached to Deep Thought. Today not even economists really believe that this is how people actually operate, but figuring out something better is a not insignificant task. Psychologists, however, have long known that what happens in the gaps between people is as important as what happens in the gaps between their ears – so is there something going on in the interactions between investors, which causes market instability?

One possible answer comes from the study of bacteria. Just as we might have suspected all along, stockmarket investor behaviour can be modelled by examining the way a bunch of brainless, single celled and barely animate creatures interested only in food and reproduction disport themselves on a Petri plate. Sometimes analogies are just too sweet.

Investing Earthquakes

The critical thing about any economic model is that it arrives at results that look like what we actually see in markets. Mostly the jargon fixated commentators who dominate the media are happy to talk in terms of business cycles when, in reality, the only cycles seen in most investing circles are the ones used by the boys and girls delivering lunchboxes. What we actually get, if we look at stockmarkets and stock prices, is something that looks like the readout we see from a seismograph when an earthquake occurs.

If we start by making a few assumptions about what investors actually do in real life – like, for instance, that they don’t behave rationally and that they tend to copy successful behaviour from people they’re closely connected to – we can rapidly create a model that produces outputs that look very different from those generated by models of people who behave independently and rationally. In fact the output of these models looks a lot like the readout we see from a seismograph when an earthquake occurs.

So it seems that the interactions between investors and how these interactions affect their willingness or otherwise to invest is the critical thing in these models. Fundamental value is, in fact, not the most important issue most of the time. Indeed, what seems to be really important is the internal behaviour of the participants, not the external behaviour of companies out in the real-world. Which simply confirms the suspicions of harassed leaders of industrial companies – the markets don’t care about real business issues.

Of Birds and Bacteria

Oddly enough the synchronised behaviour necessary to generate the type of herding needed to make this model work has been seen in various naturally occurring systems, including birds and self-propelled flagellant bacteria. The vast flocks of birds that sometimes arise, wheeling in perfect synchronicity, are utter marvels of the animal world. How can one bird, at the edge of the flock, move in tandem with another on the other side? Some deep thinkers have suggested that this can only be due to animal telepathy, but these people are, frankly, idiots.

The explanation is much more prosaic. It turns out that we can replicate this behaviour simply by using a few rules about how any given bird interprets the behaviour of the birds around it. A change at one edge of the flock can ripple through the whole in a matter of seconds purely by each individual bird reacting to the ones around it. There's been lots of computer modelling done to show this working with so-called boids. Certain types of self-propelling bacteria demonstrate the same type of behaviour. However, if you introduce noise into the system, so that the creatures can’t communicate very effectively, then the synchronisation breaks down.

If the standard unsynchronised bacteria are our investors in their normal modes and the synchronised ones are what happens at market extremes when noise levels drop it’s fairly easy to see how the connectivity of the various agents is critical. In a world in which mass, instantaneous communication is prevalent it’s perhaps no surprise that we’re seeing more instability in markets and economies. The transmission of information, fear and greed is simply faster and easier than ever before – and we should therefore expect far greater and faster displays of irrational exuberance.

Phase Transitions in Markets

These types of models where we see sudden switches between different states are characteristic of something known as a phase transition. Phase transitions occur naturally and are unquestionably hard-wired into the structure of reality. Only we don’t quite know how or why.

For example, we see a phase transitions when water boils. This transition happens across the whole substance instantaneously, as if all of the molecules simultaneously decide to stop doing one thing and start doing another. This looks an awful lot like what we witness in our bird flocks as each molecule reacts to the ones around it and the change flows through the material.

A number of researchers have suggested that a type of phase transition (technically known as log periodic) marks the point at which markets crash: when suddenly all the parts of the system instantaneously align their behaviour and everything changes on the lazy flap of a single butterfly’s wings. Indeed they go further and argue that because such events have a typical signature in terms of oscillatory behaviour it’s possible to detect them ahead of time.

Marxist Instability Rules

Lots of work has gone on to prove this by looking at historical records of crashes. This research by Sornette and Johansen gives an overview but is a few years old now and work has continued apace in the meantime. The fits that the researchers have been able to obtain have been rather impressive. Unfortunately hindsight doesn’t make anyone rich and attempts to predict future market crashes have so far been, and let’s be kind here, completely, utterly and abysmally wrong. Worse still, if the predictions did come true and people started to believe their models then the predictions would start to fail because behaviour would change to incorporate the new reality. It’s really, really tough being an economics researcher.

Karl Marx theorised that the capitalist model demanded that there was inherent instability – that the periodic booms and busts we see are an inevitable property of the market economy. He went on to draw some rather larger conclusions about the result of these fluctuations which ultimately led to the deaths of tens of millions of people and some really bad architecture so we shouldn’t be blasé about the power of economics to affect our world. Still, his underlying premise seems to be as sound as ever: markets are fundamentally inclined to instability. There is no end to boom and bust and there never can be.

And if you don’t believe me, ask a bacterium.

Related Posts: Newton's Financial Crisis, The Tragedy of the Financial Commons, Akerlof's Lemons


  1. Today not even economists really believe that this is how people actually operate, but figuring out something better is a not insignificant task

    In the investing area, we have the perfect tool to tell us how much irrationality there is in the market at any specified time -- the P/E10 value.

    Imagine that humans were 100 percent rational. What would the valuation level be? Fair value! The logical thing to do is to respond to overvaluation by selling stocks since the long-term value proposition has worsened (bringing prices back down to fair value) and to respond to undervaluation by buying stocks since the long-term value proposition has improved (bringing prices back up to fair value).

    Markets are self-regulating so long as investors are encouraged to take valuations into consideration when setting their stock allocations. All overvaluation and undervaluation is caused by promotion of the Passive Investing concept, the idea that it is not necessary to change one's stock allocation in response to price changes.

    The P/E10 level (the price of an index over the average of the last 10 years of earnings) tells you how emotional investors are at any given time. In January 2000, the P/E10 was 44, three times the fair value P/E10 of 14. That told us that that was the most emotional market we have ever seen in U.S. history and the most dangerous market for those looking for a place to invest their retirement money.


  2. I am mostly agree with everything but this part: "Fundamental value is, in fact, not the most important issue most of the time." What about the high correlation in the long run between positive cash flows and stock prices.

  3. Another fantastic post Timarr.

    Economists who thought agents in the market were perfectly rational and independent entities had obviously never bought a share...

    No excuse today, since you can get a laptop and wi-fi into any Ivory Tower...