A joint IACPM/ISDA study evaluated the degree of convergence of economic credit capital estimates as implemented by a large number of financial institutions. The results of this study can be helpful not only to the participating firms in providing benchmarks of economic capital modeling methodologies, but also to supervisors in their evaluation of practices under a Basel II perspective.
Significant progress has been made over the last eight years in developing economic credit capital models. Economic capital determination is a function of both the input credit risk parameters and the way in which the models combine these parameters to estimate credit capital. While new regulatory credit capital requirements under Basel II have been promulgated, these allow only firms following an advanced approach to submit their own estimates of key parameters as input into a single regulatory formula under Pillar I. The Basel Committee has expressed willingness to explore the use of bank internal models; although regulators have not yet accepted their use, a review of the soundness of economic credit capital models is part of the supervisory process under Pillar II. In furthering their objective of fostering research on credit risk capital modeling, the International Association of Credit Portfolio Managers (IACPM) and the International Swaps and Derivatives Association (ISDA) undertook a two-year project in 2004 to explore the convergence of economic credit capital models in use by their member firms.
The principal objectives of this study (1) were twofold:
* To provide banks participating in this project with comparisons of the capital measures generated by different credit capital models--i.e., expected loss for the portfolio and the amount of economic capital needed to support the credit risk of the portfolio at a specified confidence level. In addition to making portfolio-level comparisons and sensitivities to changes in key parameters, the study also sought to explore the assessment of credit capital allocated to individual exposures.
* To provide to external audiences, such as the Basel Committee and other interested parties, objective, verifiable, and reproducible comparisons of internal credit capital models for their review in assessing the appropriateness of using such models for regulatory purposes.
To this end, a representative portfolio of transactions was assembled with pre-specified data assumptions regarding their risk characteristics. The intent was to allow for modeling differences but not to allow for data assumptions to be the controlling variables, though some modeling differences may have resulted from data interpretation. Throughout this project an overriding consideration was to develop an understanding of the various model assumptions that would reconcile differences in capital estimates.
The overall conclusion from this study is that economic capital models employed by firms can, for the most part, be shown to converge in their estimates of portfolio-level capital requirements, given the same data assumptions. Where differences arise, a road map of the modeling assumptions can be used to reconcile these differences. Sensitivity analysis of portfolio composition indicates that implementation of correlation assumptions may still contribute to some dispersion of results. As expected, while firm-level capital estimates have been shown to converge, how firms choose to allocate capital to individual transactions indicates significant dispersion, reflecting both the diverse purposes to which these estimates are used as well as the risk management practices specific to individual firms.
Project Participants and Governance
Members of IACPM and ISDA were invited to participate in this study. Twenty-eight large financial institutions elected to participate under the guidance of a Steering Committee. (2) An initial survey of the participants …