ارزیابی مقایسه‌ای مدل‌های ریسک اعتباری برای محاسبه زیان مورد انتظار: دلالت‌هایی برای ثبات بانکی

نویسندگان
دانشکده اقتصاد دانشگاه تهران
چکیده
رویکردهای گذشته‌نگر در شناسایی زیان‌ها پس از وقوع آنها، یکی از دلایل اصلی کاهش تاب‌آوری بانکی است. این مطالعه با تأکید بر محاسبه زیان مورد انتظار پرتفوی اعتباری توسط مدل‌های مختلف، بر ارزیابی دلالت‌های مدل‌های ریسک اعتباری برای ثبات و ارتقای تاب‌آوری بانکی تمرکز دارد. با توجه به محدودیت‌های داده‌ای در ایران و عدم ثبت دقیق مبتنی بر استاندارد، با استفاده از روش تولید مصنوعی داده‌های استاندارد ازداده‌های موجود شامل پرتفوی اعتباری با ۱۰۰۰ تسهیلات به رتبه‌بندی اعتباری بر اساس توزیع فراوانی تجربی و احتمال نکول با استفاده از توزیع بتا-دوجمله‌ای و شبیه‌سازی تسهیلات با استفاده از توزیع پاره‌تو مقید اقدام شده است. تولید داده‌های استاندارد مصنوعی از داده‌های موجود بر پایه شبیه‌سازی مونت کارلو با یک میلیون تکرار استوار است. نتایج نشان‌دهنده آن است که مدل‌ واسیچک نسبت به مدل‌های ترکیبی در برآورد زیان مورد انتظار محافظه‌کارانه‌تر عمل می‌نمایند، ولی تغییرپذیری نتایج آنها نسبت به همبستگی نکول‌ها، بیشتر است. با توجه به یافته‌ها، تحلیلگران ریسک اعتباری با یک بده بستان بین سطح محافظه‌کاری و ثبات نتایج مدل‌ها مواجه هستند و تمرکز نهاد ناظر، بانک مرکزی، بر آستانه همبستگی می‌تواند در کاهش احتمال بحران‌های بانکی موثر‌تر بوده و تاب‌آوری نظام بانکی را افزایش دهد.
کلیدواژه‌ها

عنوان مقاله English

Comparative Evaluation of Credit Risk Models for Expected Loss Estimation: Implications for Banking Stability

نویسندگان English

Vahid Majed
Mohsen mehrara
Mustafa Shadab Sadabad
Faculty of Economics, University of Tehran
چکیده English

Backward-looking approaches to loss recognition are among the main causes of banking crisis. This study, emphasizing the calculation of expected credit portfolio losses, focuses on the implications of credit risk models for banking stability. Given data limitations in Iran, a synthetic dataset consistent with IFRS 9 was generated from existing data. The dataset consists of a credit portfolio with 1,000 loans that were assigned credit ratings based on the empirical frequency distribution, probabilities of default estimated using the beta-binomial distribution, and loan exposures simulated through the truncated Pareto distribution. The generation of standardized synthetic data from available information was based on Monte Carlo simulation with one million iterations. The results indicate that the Vasicek model yields more conservative estimates of expected loss compared with Mixture models, yet its outcomes are more sensitive to changes in default correlation. Credit risk analysts face a trade-off between conservatism and stability. Regulatory focus on setting correlation thresholds can more effectively reduce the likelihood of banking crises and enhance the resilience of the banking system.

کلیدواژه‌ها English

Credit risk
International Financial Reporting Standard 9 (IFRS 9)
Expected Credit Loss
Banking Stability
Threshold and Hybrid Models
Artigas, C. T. (2004). A Review of Credit Registers and Thier Use for Basel II. Bank for International Settlements.
Assefa, S., Dervovic, D., Mahfouz, M., Balch, T., Reddy, P., & Veloso, M. (2019). Generating synthetic data in finance: Opportunities, challenges, and pitfalls. In Proceedings of the 33rd Conference on Neural Information Processing Systems (NeurIPS 2019), Workshop on AI in Financial Services: Data, Fairness, Explainability, Trustworthiness, and Privacy. Vancouver, Canada.
Basle Committee on Banking Supervision. (1999). Credit risk modelling: Current practices and applications. Basle Committee on Banking Supervision.
Bellini, T. (2019). How to model and validate expected credit losses for IFRS 9 and CECL: A practical guide with examples worked in R and SAS. Academic Press.
Berglund, J. (2016). Estimating expected lifetime of revolving credit facilities in an IFRS 9 framework (Master’s thesis). Faculty of Engineering, Lund University, Centre for Mathematical Sciences.
Bolder, D. J. (2019). Credit-risk modelling: Theoretical foundations, diagnostic tools, practical examples, and numerical recipes in Python. Springer.
Chatterjee, S. (2015). Modelling credit risk. Bank of England, Centre for Central Banking Studies.
Credit Suisse Financial Products. (1998). Credit Risk+: A credit risk management framework (Technical report). Credit Suisse.
Diaz, D., & Gemmill, G. (2002). A systematic comparison of two approaches to measuring credit risk: CreditMetrics versus CreditRisk+. [Unpublished manuscript].
Eriotis, N., Kounadeas, T., & Vasiliou, D. (2019). From IAS 39 to IFRS 9: Literature review of studies on the implementation of IFRS in the European banking sector. International Journal of Finance, Insurance and Risk Management, 9(3-4), 29–51.
European Central Bank. (2019a). AnaCredit reporting manual: Part I – General methodology. European Central Bank
European Central Bank. (2019b). AnaCredit reporting manual: Part II – Datasets and data attributes. European Central Bank
European Central Bank. (2019c). AnaCredit reporting manual: Part III – Case Studies. European Cental Bank
Gordy, M. B. (1998). A comparative anatomy of credit risk models. Board of Governors of the Federal Reserve System.
Gundlach, M., & Lehrbass, F. (Eds.). (2004). CreditRisk+ in the banking industry. Springer.
Izzi, L., Oricchio, G., & Vitale, L. (2012). Basel III credit rating systems: An applied guide to quantitative and qualitative models. Springer.
Jankowitsch, R., Pichler, S., & Schwaiger, W. S. A. (2007). Modelling the economic value of credit rating systems. Journal of Banking & Finance, 31(1), 181–198.
Jarrow, R. A. (2009). Credit risk models. Annual Review of Financial Economics, 1(1), 37–68.
Kollár, B., & Gondžárová, B. (2015). Comparison of current credit risk models. Procedia Economics and Finance, 23, 341–347.
Kund, A.-G., & Rugilo, D. (2023). Does IFRS 9 increase banks’ resilience? (Working Paper No. 2792). European Central Bank.
Moraux, F. (2008). Sensitivity analysis of credit risk measures in the beta-binomial framework. The Journal of Fixed Income, 19(3), 66–77
Merton, R. C. (1974). On the Pricing of Corporate Debt: The Risk Structure of Interest Rates. The Journal of Finance, 29(2), 449–470.
Řezáč, M. (2013). Beta distributed credit score - Estimation of its J-divergence. In The 7th International Days of Statistics and Economics (pp. 1182–1191). Prague, Czech Republic.
S&P Global Ratings. (2023). Default, transition, and recovery: 2023 annual global corporate default and rating transition study. S&P Global Ratings.
Schuermann, T., & Hanson, S. (2004). Estimating probabilities of default (Staff Report No. 190). Federal Reserve Bank of New York.
Shi, B., Meng, B., & Wang, J. (2016). An optimal decision assessment model based on the acceptable maximum LGD of commercial banks and its application. Scientific Programming, 2016, Article 9751243, 9 pages.
Trueck, S., & Rachev, S. T. (2009). Rating based modeling of credit risk: Theory and application of migration matrices. Academic Press.
Vasicek, O. (2002). The distribution of loan portfolio value. Risk, 15(12), 160-162.
World Bank. (2016). Credit registry. In Global Financial Development Report 2016: Background. Retrieved January 19, 2025