2011-08-012024-05-16https://scholars.lib.ntu.edu.tw/handle/123456789/666491摘要：本計畫將探討系統風相關議題。本計畫分成兩部分。在第一部分中，我將建構一理論模型，由財務網路的觀點來探討金融機構彼此連結程度與金融系統脆弱性之間的關係。在模型中，金融機構可以透過與其他金融機構交換資產風險來進行風險分散。然而，雖然資產風險交換有讓金融機構分散風險之優點，但其也使金融機構之間的連結更為緊密。而當金融機構之財務危機有傳染性時，金融機構之間更緊密的關係可能會提高金融機構倒閉的風險。本計畫的模型將用於探討下列議題。(1)金融機構倒閉機率與其彼此連結程度（亦為金融機構資產風險分散程度）之間的關係。本計畫預期此關係可能不是單調的遞增或遞減。(2)當系統風險的考慮重要時，政府應如何設計銀行自有資本管制。本計畫主張：就風險管理來看，銀行風險分散（或避險）與自有資本之間並非一定是替代品的關係，而有可能是互補品；也就是說，只有當銀行具有充足的資本時，風險分散或避險交易才能有效降低銀行財務危機發生機率。(3)財務網路的結構如何影響社會福利。(4)政府與金融機構可以採取哪些措施降低系統性金融風暴發生的可能性。 在計畫的第二部分中，我將進行實證研究，整合兩種系統風險衡量指標來預測金融機構在金融風暴發生時的損失，也就是預測金融機構的系統脆弱性。第一種指標是Acharya et al. (2010) 所發展的邊際期望損失(Marginal Expected Shortfall)，第二種則是Billio et al. (2010)所提到、利用Granger causality所建構出的指標。這兩種指標均試圖捕捉個別金融機構的系統脆弱性，但其建構方式相當不同，故兩者應包含不同的系統脆弱性資訊。因此，整合兩者應可發展出更有效率的系統風險指標。<br> Abstract: This project will study issues related to systemic risk. It contains two parts. In the first part, I will develop a theoretical model of financial network to study the relationship between the interconnectedness among financial institutions (FIs thereafter) and the fragility of the financial system. In the model, an FI can diversify asset risk by swapping risk with other FIs. Although FIs become better diversified when they are more closely interconnected through risk swap, an increase in the degree of interconnectedness may make FIs more likely to fail due to the fact that failures of FIs are contagious. The model will be applied to study the following issues. (1) The relationship between the bankruptcy probability of an FI and the interconnectedness among FIs (that is, the degree to which the FIs’ assets are diversified). The model predicts that the relationship may not be monotonic. (2) How bank capital regulation should be designed when systemic risk is an important concern. The model proposes that asset diversification and capital may be complements rather than substitutes for banks’ risk management. (3) How the structure of the financial network will affect social welfare. (4) The roles played by the government and the private sector in preventing systemic financial crises. In the second part of the project, I will integrate two different types of systemic risk measures to predict the losses of individual FIs during the 2007-2009 financial crisis. The first is the marginal expected shortfall developed in Acharya et al. (2010), and the second is the measures based on the Granger causality relationships between the FIs’ stock returns, which are developed in Billio et al. (2010). Both types of measures try to capture how bad an FI will look like when a financial crisis occurs, but the ways they are constructed are very different. It is likely that they contain different but valuable information about the systemic fragility of individual FIs. Therefore, it is reasonable to integrate these two types of systemic risk measures to come up with more efficient ones.系統風險銀行自有資本管制財務網路風險分散Granger causalitysystemic riskbank capital regulationfinancial networkdiversificationGranger causality