Our paper on credit risk simulation at UKSim conference

Our paper "Solution Space Size in Credit Risk Simulation" has been accepted for presentation at the 15th International Conference on Modelling and Simulation UKSim, to be held in Cambridge, in April, 2013. The evaluation of credit risk is a major problem for lenders. Borrowers occasionally default and may cause huge losses on the lender's portfolio, due to the correlation existing among the borrowers, which all suffer from the same economic cycle. The evaluation problem is often dealt with through simulation (read here our previous paper on the simulation of correlated defaults), which actually samples the set of states (i.e., the combinations of the individual defaulting states of the borrowers) where the portfolio loss exceeds a given threshold. But sometimes we may wish to know all the situations where the portfolio loss is very large, rather than just a sample of it. This would allow us to identify the most critical contributors to overall large losses. It turns out that the number may be very large even for portfolio composed of a  few tens of borrowers. We are tackling the problem and start to see how we can compute the number of such critical states and enumerate all of them.