銀行在信用風險的管理上除了將預期損失認列為損失準備外，還需經濟資本(風險值減預期損失)做為緩衝，而資產組合如何影響經濟資本的大小將為本文的研究重點。此外，本文亦將銀行的信用風險納入銀行間傳染風險及流動性風險，建構一個更完整的損失模型。
本文首先以CreditPortfolioView模型為基底，考量總體經濟環境和資產組合違約率的關係，以全台灣 35 間銀行的 9 種放款作為資產組合，利用蒙地卡羅模擬法建立資產組合之信用損失分配，探討相關性與風險值的關係，並計算個別資產組合的風險貢獻，與相關性做比較。此外，我們進行更多層面的探討，研究資產組合長期和短期的相關性，並分析長期與短期信用損失分配的差異。再者，以模擬出的信用損失分配為基礎，考量傳染風險及流動性風險對信用損失分配造成的影響，模擬出全台灣銀行整體的損失分配，再計算各個銀行對應的風險貢獻。
結果顯示相關性無法完全反映資產組合風險貢獻的大小排序，風險貢獻除了受相關性影響外，亦受到違約率、曝險額的影響。長期的相關性較短期相關性高，長期信用損失分配的經濟資本也比短期信用損失分配的經濟資本大。此外，流動性風險對於整體損失分配的影響程度大於信用風險，對於各銀行風險貢獻的影響程度也很大。

Banks manage credit risk using expected loss to set the loss provisions and using unexpected loss (economic capital) to set the capital buffer. How the asset correlations of the loan portfolios affect economic capital will be the main focus of this paper. In addition, we consider contagion and liquidity risk in conjunction with credit risk to build a complete model.
Based on CreditPortfolioView model, we consider the relationship between macroeconomic variables and the probability of default. Monte Carlo simulation is applied to generate the loss distribution of 9 loans of 35 Taiwanese banks, and we analyze the impact of the correlations on VaR. We then compute the risk contribution for each loan portfolio and analyze the relationship between risk contribution and corresponding correlation. In addition, we also investigate the correlations in the long run and short run, and analyze the differences between corresponding credit loss distributions. Then, based on the simulated credit loss distribution, we consider the impact of contagion risk and liquidity risk. Loss distribution of the Taiwanese banking system is generated and we compute the risk contribution from it for each bank.
The results show that correlations do not have relevant impact on risk contributions, which may also depend on the probability of default and exposure at default. The correlations in the long run are higher than those in the short run and so is the corresponding economic capital Moreover, liquidity risk has more impact on the entire loss distribution than credit risk, and it also plays a critical role in the risk contribution of banks.

Bluhm, Christian, Overbeck, Ludger, and Wagner, Christoph (2003), An Introduction to Credit Risk Modeling, CHAPMAN \& HALL.

Kurth, A., Tasche, D. (2003),"Contributions to credit risk," Risk 16(3) (March 2003), 84-88.

Pesaran, M. H., T. Schuermann, B. Treutler, and Weiner S.M. (2006), "Macroeconomic Dynamics and Credit Risk: A Global Perspective," Journal of Money, Credit and Banking, 38, 1211-1261.

Virolainen, K. (2004),"Marco Stress Testing with a Macroeconomic Credit Risk Model for Finland,"Bank of Finland Discussion Papers, Finland.

Wilson, T. (1997a),"Portpolio Credit Risk, Part I,"Risk, 10, 113-117.

Wilson, T. (1997b),"Portpolio Credit Risk, Part II,"Risk, 10, 56-60.

Yamai, Y., Yoshiba, T. (2005),"Value-at-risk versus expected shortfall: A practical perspective," Journal of Banking \& Finance, 29, 997-1015.

鍾經樊(2010)，我國銀行信用損失評估之研究， 中央銀行季刊第三十二卷第二期。

鍾經樊(2011)，涵蓋信用風險、銀行間傳染風險、與流動性風險的台灣金融系統風險量化模型， 中央銀行季刊第三十三卷第二期。