自從 2004 年 6 月底新巴塞爾資本協定正式定案後，整個金融體系莫不致力於其相關規範的研究，其中又以信用風險最受到矚目，而信用評等是其中極重要的元素。在新協定中，鼓勵各金融業可於符合規範下，自行建立對客戶之內部信用評等系統，希望能使銀行的業務活動和曝險程度具有更高的敏感度；但相對而言，對此評等機制的控管及驗證也顯得更加重要。標竿化是目前受到較少限制的一種驗證方法，其給予銀行自行選擇標竿的彈性空間，進而比較內部評等和其外部評等或任一市場資訊之差異，標竿的選擇和評等對照 (mapping) 在標竿化的過程中扮演很重要的角色，也因此需要一個動態的方法去進行標竿化的比較，推論評等系統的特性，以確保評等對照過程的穩定。

本研究簡化 Bardos, Foulcher 和 Oung 三位學者所提出之進行動態標竿化的 state space 模型，將新巴塞爾資本協定的資本需求概念和評等的動態做連結，並以此模型作為判斷所選取之標竿是否適當和評等對照是否穩定的依據。我們分別比較國內某商業銀行的內部評等和台灣企業信用風險指標（TCRI）以及台灣企業信用風險指標和國外評等系統 Standard & Poor's 的一致性，結果發現兩者的一致性皆不存在，故建議其尋找更合適的評等系統作為標竿。

Since the Basel Ⅱ framework was imposed, the banking sector has kept paying much attention to the internal-ratings based (IRB) approaches for credit risk. As a result, validation of risk parameters and the underlying rating system plays an increasingly important role on the supervisory review process. Benchmarking and mapping are parts of the whole process of producing internally estimates and validation. However, most of the validation process still focuses on a static approach. But mapping should stress the need to infer the dynamic behavior of rating transitions. Thus, we propose a dynamic benchmarking model to infer default model or rating system consistency.

Following the model proposed by Bardos, Foulcher and Oung, we use a state space model to do dynamic backtesting and link the dynamics of capital requirements and credit ratings. In our empirical study, the compositions of two domestic rating systems or a domestic rating system and a foreign rating system do not follow the same dynamics. As a consequent, the default models are still inconsistent.

1. Basel Committee on banking supervision (BCBS) (2003), “Consultative Document -the New Basel Capital Accord”.
2. Basel Committee on Banking Supervision (BCBS) (2005), “Studies on the Validation of Internal Rating Systems”, Revised version (May, 2005)
3. Basel Committee on Banking Supervision (BCBS) (2006), “International Convergence of Capital Measurement and Capital Standards: A Revised Framework” (June, 2006).
4. Carey, M. and M. Hrycay (2001), “Parameterizing credit risk models with rating data”, Journal of Banking and Finance 25, 197-270.
5. Hamilton, J. D.(1994), Time Series Analysis, Princeton University Press, 372-408.
6. Harvey, A. C. (1989), Forecasting, Structural Time Series Models and the Kalman Filter. Cambridge University Press, 101-166.
7. Heitfield, E. (2004), “Rating System Dynamics and Bank-Reported Default Probabilities under the New Basel Capital Accord”, Board of Governors of the Federal Reserve System.
8. Jacobson, T., J. Lindé, and K. Roszbach (2003), “Internal Ratings Systems, Implied Credit Risk and the Consistency of Banks’ Risk Classification Policies”, Working Paper Series from Sveriges Riksbank 155.
9. Koopman, S. J., N. Shephard, and J. A. Doornik (1999), “Statistical algorithms for models in state space using SsfPack 2.2“, Econometrics Journal, Volume 2, 113–166.
10. Oung, V. (2003), “Exigences de capital et cycle économiques: une étude empirique sur les données françaises.“ Bulletin de la Commission bancaire 28 (April, 2003).
11. Sun, M. Y. and S. F. Wang (JCIC Risk Research Team) (2005), “Validation of Credit Rating Models - A Preliminary Look at Methodology and Literature. Review”, Joint Credit Information Center (JCIC) Column.
12. Treacy, W. F. and M. S. Carey (2000), “Credit risk rating at large U.S. banks”, Journal of Banking and Finance 24, 167-201.