Conditional Loss Estimation Using a South African Global Error Correcting Macroeconometric Model

by Albert H. De Wet of FirstRand Bank, South Africa,
Reneé Van Eyden of the University of Pretoria, and
Rangan Gupta of the University of Pretoria

July 2008

Abstract: Active credit portfolio management is becoming a central part of capital and credit management within the banking industry. Stimulated by the Basel II capital accord the estimation of risk sensitive credit and capital management is central to success in an increasingly competitive environment. If any risk mitigation or value-enhancing activity is to be pursued, a credit portfolio manager must be able to identify the interdependencies between exposures in a portfolio, but more importantly, be able to relate credit risk to tangible portfolio effects on which specific actionable items can be taken.

This analysis draws on the macroeconometric vector error correcting model (VECM) developed by De Wet and Van Eyden (2007) and applies the proposed methodology of PSTW (2006) to a fictitious portfolio of corporate bank loans within the South African economy. It illustrates that it is not only possible to link macroeconomic factors to a South African specific credit portfolio, but that scenario and sensitivity analysis can also be performed within the credit portfolio model. These results can be used in credit portfolio management or standalone credit risk analytics which is ideal for practical credit portfolio management applications.

JEL Classification: G32, E17.

Keywords: Credit portfolio modelling, macroeconometric correlation model, economic capital, scenario analysis, default threshold.

Books Referenced in this paper:  (what is this?)

Download paper (287K PDF) 32 pages