Quick grid to cope with counterparty risks

摘要

This paper describes, analyzes and evaluates algorithms and implementations for calculating potential future credit exposures borne by financial institutions. As potential future exposure belongs among the most important measures of counterparty risk, it needs to be calculated precisely and on a daily basis. This creates a very high demand on computer power where single computers can not cope with the high number of calculations within the given time. This paper focuses on a grid-type computer solution, the optimal number of computers connected to the grid, the speed gain of connecting an individual computer to the grid and the order in which the subtasks are assigned to the grid. The results indicate that using a grid for simulating potential future exposure profiles is a feasible solution for credit risk management practitioners.