Why are economists not better able to predict financial crises, and possibly even prevent them?

Doyne FarmerJ. Doyne Farmer is Director of the Complexity Economics program at the Oxford Martin School, and Professor at the Mathematical Institute

By Olivia Gordon

This is the problem mathematician J. Doyne Farmer is trying to solve, with a blossoming branch of maths known as complexity economics. The Oxford University Mathematical Institute professor (and external professor at the Santa Fe Institute) is Director of the Complexity Economics Programme at the Oxford Martin School’s Institute for New Economic Thinking, and described his work this month to an Oxford lecture theatre packed with mathematicians and economists.

Complex systems is a mathematical approach rooted in insights from the physical and natural sciences. Most equations are non-linear; they’re complex. Computerisation increasingly allows mathematicians to simulate models of such complex dynamics. By applying complex systems to economics, Farmer thinks we might find intriguing new ways to predict and mitigate what he calls ‘our collective effect on ourselves’, for example, the complicated situations that create financial downturns.  Doyne FarmerWhile a graduate student in the 1970s he build the first wearable digital computer, which was successfully used to predict the game of roulette

Farmer has long been renowned as an innovator who puts daring theories into practice. As a graduate student in the 1970s, he famously built the first wearable digital device, a three-kilobyte computer worn in a shoe which successfully predicted where a ball would land in the game of roulette. He went on to contribute across the fields of chaos theory, artificial intelligence, time series analysis and theoretical biology. In the 1980s he founded the Complex Systems Group at Los Alamos National Laboratory in the US, and these days his focus is on financial systems, trying to model business cycles, specifically the agents of financial instability that make GDP and unemployment plunge periodically. As he puts it, ‘I’m trying to do something useful with my life.’

He was surprised to discover how little investment goes into researching financial crises, despite the fact the last one cost the world tens of trillions of pounds. As he explained in his lecture, we’ve got pretty good at predicting everything from the weather to climate change to asteroids to traffic, but ‘the really hard challenge is to understand our collective effect on ourselves as our world becomes more and more artificial.’ Financial panics are endemic and yet ‘we still don’t understand they’re coming even as they come upon us’. 

Fundamental to Farmer’s ideas is a rebellion against the notion of mathematical ‘equilibrium’ which has traditionally underpinned economic theories. As a boy of nine, playing noughts and crosses, Farmer grasped that it’s easy to work out how to play so that you always win or draw. Equilibrium describes the fixed point reached in such a game. But his recent research has shown that there can be a ‘chaotic zone’ in games with even two players, which means there’s sometimes equilibrium but also sometimes not. The chaos increases with the number of players. The leverage cycles of stock markets are in effect a multi-dimensional game. As Farmer noted, equilibrium is such a fundamental assumption in almost all economic models that the idea that it might be fundamentally wrong in many circumstances ‘makes economists really nervous’.  Doyne FarmerHis current research is in economics, including agent-based modeling, financial instability and technological progress

Simulation plays an ever-increasing role in mathematics, and experimental mathematics like Farmer’s could hopefully make a real change to global economics. ‘My vision of the future would be simulation of the major economies of the world,’ Farmer said. Models could be used to formulate policies. He added: ‘Economics is a hell of a lot harder than physics – but that doesn’t mean you have to throw up your hands and give up.’

Images © Olivia Gordon

Comments

By Peter Fernie
on

I would not be inclined to agree with his statement "we’ve got pretty good at predicting everything from the weather to climate change ". Climate change models have been demonstrably inaccurate.

By RH Findlay
on

Once mathematicians can predict the behaviour of a shoal of frightened and hungry fish, they could be about halfway to predicting the Stock Market and the behaviour of millions of greedy and frightened investors. And even then there will be unpredictable hidden stupidity built into the system, such as that which triggered the Great Financial Scam (Crash) of 2007-2008.

I wish the mathematicians luck in their pursuit. If they can crack that problem they will become much sought-after and well-paid consultants.

By Anthony Michael...
on

There is a distinction to be made between analysis aimed at prescribing a policy (which might start from the idea of providing a prediction) and analysis with the more modest aim of estimating risks. Analysis does not provide a prediction that you will fall to your death if you walk too close to the edge of a cliff but it does show that there is a RISK that this will happen. A number (a minority of course) of commentators pointed out the risks being run in the US economy before the crash. An economy based on personal consumption cannot sustain growth when the overwhelming majority has static incomes. The short term solution - credit growth based on funny money - was required to run for too long - the money became increasingly funny peculiar as a result. The crash was due; timing was hard to predict precisely.

Add new comment