Traditional economics failed to predict the credit crunch – but supercomputers might stand a better chance, according to the EU
The European Union has announced the finalisation of a project designed to use supercomputing technology to create advanced financial models to help predict and prevent future financial crises.
In a statement released this week, the European authority announced that it financed the supercomputing project to the tune of €2.5 million. The EURACE project has been in development for three years and is based on agent technology – one of the applications of which is computer generated effects in the film industry – known as Flexible Large-scale Agent Modelling Environment (FLAME).
According to the EU, the software effectively simulates the interactions between “different economic actors” such as households and companies, banks and borrowers or employers and job-seekers. “This first class European research can help us make the move from the economics of pen and paper to the economics of super-computers,” said Viviane Reding, EU Commissioner for Information Society and Media.
“The results of this research project, will complement traditional economic statistics and assumptions about how economic actors react by enabling better testing of a policy’s effects on people, while still on the drawing board.”
The EU also explained that conventional approaches to financial modelling largely failed to predict the impact of the financial crisis but the agent-based approach could create more realistic simulations that would prove more useful to preventing future recessions.
“Traditional economics failed to predict the scale of the knock-on effect of the credit crunch on the world economy,” the EU stated. “The new software shows how banks react in different ways by looking at a wide range of factors like how much reserves they must keep compared to investments, their savers’ consumption/investment and saving patterns, and psychological factors like confidence in the market.”
Agent-based technology is an emerging brach of Articifical Intelligence technology according to the EU which allows for more sophisticated modelling. “Agent based modelling is one of the emergent branches of AI which best demonstrates complex, social behaviour of different communities living together in real world scenarios,” the EU states. “The ideology allows agents, representing individuals or groups, to be put into a simulated environment where their individual interactions can then be studied more closely.”
The specific FLAME agent technology was developed by the UK’s University of Sheffield and has been involved in a variety of projects including the Epitheliome Project for creating various models of the skin and the SUMO project for modelling the behaviour of E-Coli bacterial cells living in organisms.