The Mathematical Finance Group in Oxford has long been a leader in research on financial mathematics. In recent years a number of research areas have become key focal points within the Group, notably behavioural finance and financial big data, robust pricing and machine learning. In particular, research has focused on financial stability, an area that became critical after the Financial Crisis in 2008.
Prof. Doyne Farmer and some of his students are working on modelling complex financial systems and the ways that losses and financial distress can spread through such systems. In particular, the Group is looking at how to design a model that accurately captures the losses the banking system, shadow banking system and real economy could incur in another crisis. These models are called 'stress tests' and are nowadays conducted by most major banks. However, the current methodology of the stress tests does not capture the network of interbank connections, although this has been the major driver of losses in a crisis. Prof. Doyne Farmer's group is developing models that more accurately capture these interconnections.
Today, a letter published in the Financial Times, discusses exactly this: the methodology of existing stress tests and their limitations.