Introduction

Abstract

The new curves introduced in Part I reflect the segmentation of the interbank derivatives market, in particular following the crisis after the default of Lehman Brothers in 2008. These new curves allow us to consistently price perfectly collateralized vanilla interest rate swaps, tenor basis swaps and cross-currency swaps. However, we have already seen in Section 5.1 that collateralization is very often far from being perfect, and that we cannot capture all sources of imperfection in the discount curve. We therefore have to ask ourselves, what do we do if collateral is not exchanged “continuously” or at least daily? What if collateral is posted only beyond certain thresholds, or asymmetrically (i.e. by one party)? What if collateral is posted with delay due to operational problems or disputes? It is clear that credit risk is not mitigated entirely in these cases, and the question is to which degree this imperfection affects a derivative’s price. The amount by which the perfectly collateralized derivative price deviates from the credit risk-adjusted derivative price is the Credit Value Adjustment (CVA). This depends in principle on the CSA details and complexities as described in Section 5, and computing it for a large portfolio can be a serious computational challenge. Since 2013 this is a challenge faced by all organizations reporting under the International Financial Reporting Standard (IFRS) — which comprises e.g. all listed companies in Europe — with IFRS 13 taking effect: CVA has to be reflected in derivatives’ valuations for accounting purposes. It is not solely for “internal information” anymore. This not only affects banks but also corporates, insurance companies and pension funds that use derivatives, and the subject has gained attention among audit firms, not least due to IFRS 13. At the time of writing, a major part of the industry is moving or has to move towards applying prudent CVA analytics regularly, at least to check CVA materiality.