Extended Viewer

Technical Note - Backtesting Counterparty Credit Risk Models - October 2014

In this Technical Note - following the Basel Committee's recommendations - we present our methodology for backtesting the Counterparty Credit Risk models in MSCI's RiskManager product. We test risk factor simulation models and our margining framework separately by comparing the realized risk factor values and the netted exposures to their forecast distributions at multiple horizons. We use the Cram'r-von Mises test to check if the forecasts were statistically correct during the observation period. In the second part of the document, we discuss the practical questions related to the implementation of the methodology. Finally, we recommend visualization techniques and address the typical challenges in Counterparty Credit Risk backtesting.