Banks and brokerage houses employ a lot of physicists - it was a big fad of the last decade because the industry discovered that physicists knew how to make models and economists did not.

But physicists don't understand economics and the behavior of people any more than economists do and it hasn't worked.  No rational model predicts that some trader will just go crazy and lose billions and automated schemes have been a disaster. Yet there is hope.

Gerolamo Cardano, Italian Renaissance mathematician, was known to the outside world for his achievements in algebra, but locally Gardano was known as a notorious gambler.  Naturally, he tried to science that up 400 years ago much the same way banks are trying now and his extensive research into probability theory made him one of the founders of the field, writing the "Book on Games of Chance". The book is still relevant today, if you want to read probability theory, but so is the downside - he lost a lot of money.

The EU's Crisis (Complexity Research Initiative for Systemic InstabilitieS) project wants to produce a much smarter and  even more sophisticated method of modeling and understanding the European economy. Using complex systems-based tools, they are developing a bottom-up 'agent-based' simulation that fully accounts for the heterogeneity of the financial actions of households, companies and government actors, incorporating the latest advances from behavioral economics. This will be used to create serious games to predict more accurately the effects of different macroeconomic and financial events. 

You see two problems right away.  'Fully accounts' is not possible and behavioral economics is essentially applied psychology - it is fine for knowing how many people on average will buy a box of soap but, in economics, where an outlier result can wreck a company, it is a dangerous feel-good fallacy.

'[Current] models may provide "good enough" answers during normal times, during the crisis they proved highly inadequate. The crisis was characterised by behaviours that did not fit the "perfectly rational" models, markets that failed to clear severe economic imbalances that were far from equilibrium, and micro-level features of the system and network structures of interconnection between institutions had major systemic impacts,' the Crisis project partners explain. 'The Crisis project will seek to address these limitations by building a next-generation macroeconomic and financial system policymaking model.'

In a similar vein, researchers at the Consiglio Nazionale delle Ricerche in Italy are also trying to improve economists and policy-makers' ability to predict systemic risk and global financial instabilities. In the FOC-II (
Forecasting Financial Crises II) project, an interdisciplinary consortium of computer scientists, physicists, economists and policy-makers are developing a collaborative platform for monitoring systemic fragility and the propagation of financial distress across institutions and markets around the world. 

While Cardano might be impressed by technology that offers a model of future risk and probabilities, he likely knows probability theory doesn't work in macro-economics when it doesn't even work at dice.