AdaptUnpredict – Study of game theory models paves the way for better financial market management
Game theory sounds like something more relevant to chess than economics, but it could hold the key to understanding the market failures that trigger financial crises.
© Fotolia, 2012
Implementing new models, like those developed during Professor Peter Hammond's time as Marie Curie Excellence Chair at the University Warwick, UK, could improve the way the financial sector handles such issues in the future.
Game theory models consider the interactions between several 'players', whose decisions will be influenced by their expectations of the others' strategies. This means they are inherently unpredictable.
Since giving up playing chess seriously over 40 years ago, Professor Hammond had been interested in the shortcomings of game theory models. In 2007, the once-in-a-lifetime opportunity afforded him with the offer of the Marie Curie Excellence Chair at the University Warwick, UK, mobilised him. He took early retirement from Stanford University to return to Europe.
"Sometimes you want to do something that seems really adventurous," he explains. The Marie Curie Action (MCA) grant for the AdaptUnpredict project gave this talented researcher the chance to "do something I'd never really quite had the courage to tackle head on".
Financial markets are often regarded as multi-players games, with economists striving to understand and improve the 'rules' governing how governments, producers, and consumers interact.
But unpredictable events can render these models "inadequate in various ways," explains Professor Hammond. "Their validity is only ever temporary."
Using the Eurozone crisis as an example, Professor Hammond explains that rigid rules can break down catastrophically if faced with unforeseen circumstances. For this reason, economists need to think about how to create more robust institutions, with more relaxed game theory rules. The idea is that they should be able to change if the unpredictable consequences of previous decisions require it.
AdaptUnpredict began a pioneering investigation of so called 'enlivened models'. These models are constructed to ignore any built-in flaws that would often cause others to fail. They do this by adapting to previous decisions.
Professor Hammond hopes that they will be incorporated into the financial marketplace, but acknowledges that this will not happen overnight, which is why he continues write papers, give talks and encourage colleagues to pursue related ideas.
In recognition for his career achievements Professor Hammond was elected as a Fellow of the British Academy in 2009, and is now an Emeritus Professor at Stanford and a part-time Professor at Warwick.