Lifetime PD Analytics for Credit Portfolios: A Survey

Abstract

The recent publication of the IFRS 9 norms has emphasized the fact that a shared and comprehensive methodology for PD analytics on credit portfolios was still lacking. Credit risk assessment is often static and short term because the industry has focused on assessing risk over a one year horizon, pushed forward this way by common practices and by the regulatory framework. Dynamic aspects are crucial though. IFRS 9 requirements raise new issues regarding dynamic and long term risk assessment. Plenty of information is available for calibrating PD curves (scores, risk classes, risk class migrations, observed defaults, delinquencies...) and there is a large set of statistical methods at hand as well. This paper surveys the main the available models for PD analytics and focuses on retail portfolios analytics because, contrary to the case of wholesale portfolios, no consensus has emerged yet on the way to calibrate lifetime PDs for retail exposures.