The evolution of markets and strategies

Is it possible to make good risk-adjusted returns on equities? My answer to this has changed. Back in 2006-07, I'd have said yes - by holding defensive stocks. These had for years done better than they should. Today, though, I'd give a different answer.

Recently, lower risk stocks have done no better or worse than they should. Instead, it is momentum stocks that have done well, whereas before about 2007 they did no better than their riskiness would warrant.

This shows that the success of particular strategies changes over time. Why? The answer lies in an idea proposed by economists such as Brian Arthur at the Santa Fe Institute in New Mexico, Andrew Lo at the Massachusetts Institute of Technology and Eric Beinhocker at McKinsey Global Institute. Financial markets, they say, are evolutionary systems*.

The basic premise here, says Professor Lo, is that investors have bounded rationality. They can't know everything, so they adopt, after trial and error, some simple heuristics or investment strategies - such as momentum, trading, value investing or buying tail risk. Just as species compete for food, so strategies compete to make profits. Successful strategies, like successful species, get lots of food and so reproduce. And unsuccessful ones die.

The "buy defensives" strategy fits this pattern. Before 2006-07, it ate lots of food (made good profits) and so reproduced - investors adopted it. But in multiplying, this strategy/species depleted its food source, and so recently has merely ticked over. It's not done so badly as to go extinct, or done so well as to multiply further.

At this stage, advocates of efficient market theory say that this sort of thing should happen all the time. "Fit" strategies, which eat lots of food, should multiply at the expense of unfit ones; stupid investors should run out of money and leave the market. As this happens, the market should converge towards efficiency.

This certainly happens in many cases. It's usually very hard to make super-normal profits by trading foreign exchange (unless you have near-inside knowledge) or government bonds. But it doesn't happen always, quickly and everywhere. Just as biological evolution has not (yet?!) produced super-fit master-species, so market evolution has not driven us towards full efficiency.

There are good reasons for this.

One is that sometimes a food source is so abundant that even weak species can thrive. If central banks run easy money policies, lots of strategies might make money that would fail if monetary conditions are tight - hence the adage, "don't fight the Fed."

A good example of such abundance came during the era of low volatility created by the "Great Moderation". This encouraged the multiplication of strategies that involved taking tail risk - for example, selling credit insurance or holding collateralized debt obligations. Such strategies thrived when their food source - low volatility - was abundant. But when this food disappeared in 2008, they died.

A further reason why natural selection doesn't bite hard enough to kill off the unfit is that healthy species don't always eat all the available food, thus leaving some for maladapted species. One factor here is the short sales constraint. Many institutional investors are prevented from shorting stocks. And even those who are free to do so are often reluctant to go short. It is risky to fight the madness of crowds; markets can stay irrational for longer than you can stay solvent. As Bernard Dumas at Insead has shown, there is nothing much that rational investors can do to exploit excess volatility. This means that bandwagon-hopping strategies ("have some of these - they're going up") can thrive for a while at least.

Natural selection, then, doesn't necessarily kill off market inefficiencies quickly.

What's more, new inefficiencies can sometimes emerge. Researchers at the University of Nantes have found that although the FX market is usually efficient, there have been occasions, such as in the mid-late 80s or 2007 - when simple momentum trading made money.

Their research gives us a clue as to one way in which such opportunities arise - through policy intervention. In 1985 governments agreed that the US dollar was too high and wanted it to fall. Betting against the dollar therefore became an easy way to make money - until the food (policy-makers' desire for a weaker dollar) disappeared. A similar thing might be happening now with quantitative easing. This is providing good profits for those strategies which bet on assets that benefit from printing money (such as gold). But some time, the food will run out.

Another source of occasional inefficiencies lies in the distribution of investors' beliefs. Professor Arthur has shown that quite small changes in these can give rise to apparently inefficient behaviour such as momentum or over-reaction.

Brock Mendel and Andrei Shleifer, two economists at Harvard University, describe how this happens. We can, they say, think of stock markets as comprising three types of investor: well-informed insiders, irrational noise traders, and those who are rational but lack inside information. Imagine, then, that a share's price rises and you are the rational but uninformed investor. What do you do? It depends upon your belief about the balance between the stupid money and the smart money. If you think noise traders are dominant, you'll sell the share. But if you think the smart money is buying, you'll want to buy.

And here's the rub. Whatever you do, you might be mistaken. Sometimes, you'll buy as the noise traders buy and so amplify irrational price moves. In this way, market inefficiencies can arise (and disappear) either from changes in the balance of investors' beliefs - the balance between smart and stupid money - or from changes in beliefs about those beliefs: do others believe that it's the clever or stupid money that's driving prices?

There are two simple messages in all this. First, we shouldn't think of markets as being efficient or inefficient. They are instead, says Professor Lo, adaptive. The profitability of strategies waxes and wanes as natural selection operates with varying degrees of force and as the environment alters to provide food for some strategies and to remove it for others. The challenge for investors is to ask whether selection and environmental forces favour a particular strategy or not.

Secondly, investment success requires flexibility and a range of strategies, simply because single ones can and do go extinct. Professor Lo says: "survival is ultimately the only objective that matters."

* In saying this, economists are not so much borrowing an idea from biology so much as claiming it back; Darwin and Wallace took the idea of natural selection from the economist Thomas Malthus.

visible-status-Standard story-url-evolution_markets_140111.xml