HSBC to  use Strategy Management
Why do organizations need decision analytics?
A driving force behind the Spanish economy
Russia's automotive market sees extensive growth
SM roll-out at RBS in Germany
European Telecom Forum 06
White paper: Fraud in the telecommunication industry
White paper: Modelling personal bankruptcy in the UK
New Solution: ACQUIRE SM
Faster & faster
Incorporating Macroeconomic Dynamics into Credit Risk models
The Expert's column

Incorporating Macroeconomic Dynamics into Credit Risk Models
 

Traditionally, little attention has been devoted to the impact of economic developments on operational strategies such as limit management, pricing and collections management. Experian Decision Analytics has developed a pragmatic approach that shows the impact of different scenarios in a risk model.

Incorporating relevant aspects of the economy into traditional credit risk measurement models is difficult. Historical information on credit risk performance is typically limited and may lack sufficient variation in aspects of interest, whilst macroeconomic data relevant to credit risk modelling may not be readily available. Furthermore, risk models need to recognise path-dependence in credit risk performance and correlation among economic variables, and recognise both the direct and indirect impact of economic events on credit risk. To tackle this problem, Experian Decision Analytics has developed a pragmatic approach that specifically recognises both the limitations in the data and the requirements for comprehensive modelling of relevant macroeconomic factors. It does so in two stages: the first stage involves the development of a new, granular leading indicator of credit risk, underpinned by a full structural econometric model of the economy. This allows the formulation of default rate predictions based on central case macroeconomic forecasts as well as specific hypothetical or historical macroeconomic scenarios. The second stage includes a methodology for linking the leading risk indicator to traditional scores in order to update risk prediction based purely on historical experience and to produce predictions under stress scenarios. The impact of a recession or a stress scenario appears to vary greatly across risk segments, thus confirming that aggregate, portfolio-level modelling may fail to provide the type of information required to make appropriate operational decisions. The practicality of the approach should assist financial services organisations' efforts to build expected economic developments into competitive lending propositions, and to bring scenario planning and stress testing into all their retail credit risk management practices.  

Paul Russell Analytics Centre of Excellence Director, Experian Decision Analytics

Top