| DC 欄位 |
值 |
語言 |
| DC.contributor | 統計研究所 | zh_TW |
| DC.creator | 吳柏寬 | zh_TW |
| DC.creator | Bo-Kuan Wu | en_US |
| dc.date.accessioned | 2018-7-23T07:39:07Z | |
| dc.date.available | 2018-7-23T07:39:07Z | |
| dc.date.issued | 2018 | |
| dc.identifier.uri | http://ir.lib.ncu.edu.tw:88/thesis/view_etd.asp?URN=105225010 | |
| dc.contributor.department | 統計研究所 | zh_TW |
| DC.description | 國立中央大學 | zh_TW |
| DC.description | National Central University | en_US |
| dc.description.abstract | 隨著2007 年金融危機的過去，信用風險管理在各大金融機構中逐漸成為一個重要的議題。Duffie (2011) 提出了相關違約具有三種特性，分別是共同影響性，傳染性以及違約後重要資訊的釋出。Fuh and Kao (2017) 使用因子模型去捕捉三種特性並且給出相對應的多維違約距離。為了研究公司之間的相關性，本研究使用Fuh and Kao (2017) 所提出的因子模型去捕捉公司之間的相關性並且計算出相關性矩陣。同時也使用美國市場資料計算出相關性矩陣以及多維違約距離。為了觀察違約前後的相關性是否有變化，本研究把市場資料依照時間分成兩塊。
從實際資料所算出的相關性矩陣可以發現公司規模較大的公司不容易受到其他公司違約的影響; 反之，公司規模較小的公司較容易受到其他公司違約的影響。此外我們可以從相關性矩陣明顯的強弱區分公司規模較大的公司與公司規模較小的公司。最後，我們利用美國市場資料計算出多維違約距離和違約機率。從多維違約距離來看，我們發現相關違約中的傳染性確實會影響其他公司，然而公司的財務狀況以及市場波動也有一定的影響力存在。 | zh_TW |
| dc.description.abstract | After financial crisis in 2007, the credit risk management has became one of the important
issues around the financial institutions. In Duffie (2011), there are several features of correlated default between each firm, namely co-movement, contagion, and information
release. Fuh and Kao (2017) use the commonly factor model to capture those features and proposed three types of multi-name Distance-to-Default. In order to study the correlation between each firm, we use factor model which propose by Fuh and Kao (2017) to capture the correlation and calculate the variance-covariance matrix. Using the US financial data, we calculate the variance-covariance matrix and multi-name Distance-to-Default. We also separate our data to observe the different of the correlation of between each firm before and after firm default.
From the correlation matrix based on empirical data, we can find that the firm with high firm-value is not easily influenced from other firm default and the firm with low firmvalue is easily influenced from other firm default. In addition, we also observe that we can easily distinguish the firm with high and low firm-value from the color of correlation matrix. Based on multi-name Distance-to-Default, we can discover that the default event have an impact on other firms in same industry sector. However, the default probability of other firms are also affected from the financial state of itself or the fluctuation of financial market. | en_US |
| DC.subject | 信用風險 | zh_TW |
| DC.subject | 因子模型 | zh_TW |
| DC.subject | 違約距離 | zh_TW |
| DC.subject | Credit Risk | en_US |
| DC.subject | Factor Model | en_US |
| DC.subject | Distance-to-Default | en_US |
| DC.title | Empirical Evidences for Correlated Defaults | en_US |
| dc.language.iso | en_US | en_US |
| DC.type | 博碩士論文 | zh_TW |
| DC.type | thesis | en_US |
| DC.publisher | National Central University | en_US |