Stability of a "through-the-cycle" rating system during a financial crisis
The aim of this paper is to research the construction of a "through-the-cycle" (TTC) rating system to assess credit risk in a developing country that has faced a major economic crisis. The country analysed was involved in a severe macroeconomic crisis that spread throughout the financial system.
Constructing a rating system involves estimating a credit scoring model and using the estimated scores to construct risk categories. Furthermore, a TTC rating system uses specific and dynamic information on obligors to assess the credit quality of borrowers. It remains relatively stable in business cycles as stressed scenarios have been considered. But in the developing country analysed, the macroeconomic crisis influenced obligor payment behaviour and therefore affected the rating system. Some problems were encountered in building a stable TTC rating system. The unstable conditions of this particular developing country pose difficulties for constructing rating systems that would be shared by other emerging economies.
The scoring model was constructed estimating a panel data model using public credit registry information on the country's corporate debtors. The database used for the panel regression followed the same debtors in the financial system for a fiveyear period. The scores obtained from the panel regression were the inputs of the rating system that grouped obligors in different risk categories. Achieving stable risk categories was quite problematic, as the obligors' performance in crisis years was completely different from the one observed in stable ones. As a matter of fact, the annual frequency of defaults increased from 13% to 30% during the year of the crisis.
Traditionally, credit institutions decided whether or not to grant credit to a particular individual based on human judgment and historical experience about the default risk. However, sophisticated statistical credit scoring models were recently developed to aid the credit granting decision. They are used to estimate the probability of default (PD) using predictor variables taking into account the characteristics and financial situation of applicants. The decision to accept or reject candidates can be taken after comparing estimated PDs with a suitable threshold. These models are also used in the "ongoing" process of the loan to estimate its likelihood of default.
In June 2004, the Basel Committee on Banking Supervision published the International convergence of capitalmeasurement and capital standards: a revised framework (Basel II). One of its main objectives is to promote the adoption of stronger risk management practices by the banking industry. An important innovation of the Revised Framework is the possibility of using internal rating systems as inputs for capital calculations after they have met minimum requirements set out in the document.
The Revised Framework considers that human judgment should be used in the decision to grant loans but highlights the necessity of establishing a formal methodology to rate obligors and to estimate the associated PDs. Thus, it describes methodologies for banks to construct their internal ratingsbased (IRB) systems. Banks may use IRB systems to calculate regulatory capital requirements but also as the basis for internal risk measures; so they will use these risk measures for pricing, managing portfolio exposures and establishing reserves. It is important that IRB systems should accurately discriminate between bad and good obligors, those that have higher and lower PD. The accuracy of the estimated PDs and the structure of the rating system would influence capital requirements. This is the reason for focusing on the estimation of a credit scoring model and the construction of a rating system in this document.
The risk measures used to calculate capital requirements are the probability of default (PD), loss-given-default (LGD), exposure at default (EAD) and effective maturity (M). There are two IRB approaches, foundation and advanced. Under both approaches, banks have to provide their own estimates of PD subject to minimum requirements. The Revised Framework specifies that all banks using IRB approaches must estimate a PD for each risk category of the rating system distinguishing between corporate, sovereign and bank exposures.
The Revised Framework highlights that estimated PDs must be a long-run average of one-year PDs for borrowers in each category of the rating system. The recent document published by the Basel Committee on Banking Supervision2 describes different types of rating systems. The rating system can be calculated with information of one period (one year) as a "point-in-time" (PIT) rating system or, in line with the Revised Framework, it can be calculated with information of a longer period, that is, a "through-the-cycle" (TTC) rating system. The latter rating system would consider long-run estimations of the PDs. The Revised Framework explicitly points out that the length of the underlying historical observation period used to calculate PDs must be at least five years.
The present exercise is an empirical research that constructs a rating system for corporate obligors in the entire financial system registered at the Public Credit Registry of Borrowers (PCRB). We are aware that this exercise is different from an IRB system constructed by a bank that has access to detailed information on borrowers. However, the intention is to construct a broader rating system with information on obligors in a crisis to highlight the problems that would be confronted in the validation process of such an extremely stressed situation.
The paper proceeds as follows. Section 1 briefly describes differences between a PIT and a TTC rating system. Section 2 describes the database used in the empirical estimation and presents the macroeconomic variables to be considered. Section 3 calculates the panel model and the score of the rating system that is validated in Section 4. Section 5 presents the rating system for two years that have a similar percentage of default rates and Section 6 highlights the importance of having a database with the financial history of obligors. Finally, the conclusions show problems and suggestions discovered during the research.
1 The views expressed herein are solely those of the author and do not necessarily represent official policies, statements or views of the Central Bank of Argentina. I would like to thank Cristina Pailhé for her helpful comments and Matías Gutiérrez for his useful advice.
2 BCBS (2005).