Framework for Enhanced Modelling of Collateralised Lending

Quantifying the risks associated with collateralised lending has a particular relevance for central banks, notably in their role as a lender of last resort. Central banks are susceptible to losses resulting from double default events where a reverse repo counterparty and the collateral default simultaneously. While there is a broad range of theoretical models for assessing and quantifying portfolio credit risk and double default, this paper aims to propose a practical framework that uses a portfolio of notional CDS contracts with dummy counterparties to represent collateral for loan exposures. A key advantage of this approach is that it can be implemented within CreditManager – a widely used risk IT application. Extensions to enable modelling of additional heterogeneous pools of collateral and to account for the effect of overcollateralisation are also outlined. The preliminary results indicate that applying this framework generates loss distributions that reflect the inherent credit risks associated with collateralised lending.