In the aftermath of the 2007/2008 global financial crisis, financial institutions have drastically shifted their focus to more actively manage the risks associated with over-the-counter (OTC) derivatives. This typically involves calculating credit exposure metrics in a computationally expensive Monte-Carlo simulation. The exposure to the counterparty or the bank itself is calculated at a set of future time steps for thousands of scenarios. Various metrics can be derived from these exposures, such as expected exposure (EE), potential future exposure (PFE), and many more. These metrics form the basis for computing capital charges (e.g. SA-CCR, SA-CVA, RWA) and various valuation adjustments (XVAs).
Banks struggle to cope with the computational complexity involved as the simulation requires huge volumes of data and billions of valuations. It is therefore essential that banks employ a combination of algorithmic optimisation and software code modernisation techniques. This paper gives an overview of the complexities involved in calculating credit exposure metrics and details a range of optimisation techniques to cope with the challenge.