Some of this progress consists in understanding why it is that the economy is unpredictable. Economists used to believe that the economy was a simple system in which shocks to a few so-called representative agents (such as “the consumer” or “the corporate sector”) had linear effects. Recently, though, they have progressed beyond this.
1. Everything looks normal – until it suddenly doesn’t One example is a paper from January called A Creepy World by Didier Sornette and Peter Cauwels at the Swiss Federal Institute of Technology Zurich.
They show that the idea of creep in materials science should be imported into economics. This says that if you place some metals or concrete under some constant level of stress the material might for a long time appear stable and robust – until it suddenly breaks.
Financial markets behave similarly. For example, in the 00s, banks’ high gearing and reliance upon wholesale funding didn’t much worry investors – until it suddenly did. High government debt in southern Europe seemed stable and sustainable for years – until things suddenly changed in 2011. And the oil price seemed for a long time resilient to the Chinese slowdown and rising supply of shale gas – until it collapsed this year.
2. Why was banking crisis so catastrophic? Networks Another feature of the new economics of complexity is the recognition that the precise structure of networks between people and firms matters enormously. Two surveys in the latest Journal of Economic Perspectives by Matthew Jackson and Vasco Carvalho show how.
If a key hub in a network collapses, it can drag down the whole network, in a way in which the collapse of a peripheral unit doesn’t.
This is why the banking crisis was so catastrophic: banks were hubs and their failure hit both other banks and firms which were dependent upon credit. In this sense, network structures can cause individual company failures to have macroeconomic effects.
All this implies that experts cannot foresee the future because they don’t have – and, for now at least, can’t have – detailed knowledge about the structure of networks or of when non-linear effects will happen. Research by Andriy Bodnaruk of the University of Notre Dame and Andrei Simonov at Michigan State University confirms this. They show that the personal investments of unit trust managers do no better than those of non-specialist investors.
“Wealthy investors appear to be as good individual investors as professional asset managers” they conclude.
3. Friends can influence investment decisions – and the weather might too However, this doesn’t mean that non-specialist investors are especially good. Other research this year has reminded us of their shortcomings. Annie Zhang at the University Auckland shows that our asset allocation decisions are unduly influenced by our peers: people’s investment choices tend to resemble those of the people around them. And researchers at York University have shown that the weather affects our investment choices and share prices.
4. “Past performance not indicative of future results” – but many think it is A further failing has been demonstrated by researchers at Georgia Tech. They got people to trade an artificial asset under laboratory conditions where they knew there was a probability that trading would cease at any time, leaving them with a lowly-priced asset.
The researchers found that even when subjects were told the fundamental value of the asset, they pushed prices above it.
This, they say, is because traders based their expected prices not just upon the fundamental value, but upon past prices; if they saw prices rise, they extrapolated further increases.
This is not merely a quirk of laboratory experiments. Mannheim University’s Klaus Adam and colleagues show that the same is true for real investors’ expectations of US share prices.
5. A rational market can arise even if many traders are stupid
You might think that, if investors are so daft, it should be easy to beat the market. You’d be wrong. Work by Yale University’s Shyam Sunder and colleagues shows why.
They created an artificial market on a computer in which half of traders were well-informed whilst half were ignorant and updated their beliefs about future prices using a simple rule of thumb.
They discovered that in this market in which many traders, by design, had zero intelligence prices quickly converged to what they would have been if all traders were intelligent.
This shows that a rational market can arise even if many traders are stupid (the converse can also be true – but that’s another story). Whether a market is efficient or not depends not upon the rationality or not of its individual traders, but upon how they interact. This is an example of complexity: the properties of a system (such as a market) are not merely those of its individual units writ large.
6. Even when you don’t expect it, overconfidence can strike again The above finding implies that market efficiency is a fragile thing. A paper by Jiali Fang of Massey University shows some of the conditions in which it exists.
She studied how technical analysis strategies – based on past price moves – worked in 50 national stock market indices over the last 20 years.
If markets are efficient, such strategies shouldn’t work because any information contained in past price moves should be discounted immediately by markets.
However, she shows that technical analysis does work sometimes – where markets are under-developed and when investors are uncertain about fundamental information. These conditions, however, are rarely met in more developed economies – which is why Arvid Hoffman and Hersh Shefrin have found that technical analysis often loses money in these.
It’s in this context that we should read a paper by Meir Statman of Santa Clara University. He points out that the same sort of cognitive errors that sometimes cause markets to be inefficient also lead us to over-estimate our ability to beat the market; not least of these errors is simple overconfidence.
Perhaps there’s a common theme here. It’s that the economy and markets are complex and unpredictable things, and that humans lack the cognitive resources to make big systematic profits in them.