【摘要】：隨著金融全球化及金融創(chuàng)新步伐不斷加快，金融機(jī)構(gòu)間的聯(lián)系逐漸變得更加緊密。在我國，銀行業(yè)是金融系統(tǒng)的核心。銀行與銀行之間通過業(yè)務(wù)往來等具有相互關(guān)聯(lián)性的行為聯(lián)系變得更加緊密。如何刻畫銀行間的相關(guān)性，以及如何測度銀行間的風(fēng)險溢出強(qiáng)度，是一個不容忽視的問題。 Copula理論在分析變量間的相關(guān)結(jié)構(gòu)時具有很多優(yōu)點，可以較好地刻畫變量間非線性、非對稱的尾部相關(guān)關(guān)系，同時變結(jié)構(gòu)Copula可以精確找到變量間相關(guān)結(jié)構(gòu)的變點，為研究風(fēng)險溢出提供依據(jù)。本文選取我國上市銀行中十家銀行作為樣本，實證分析了上市銀行的相關(guān)性特別是時變相關(guān)性和風(fēng)險溢出效應(yīng)。在實證研究過程中，首先，金融時間序列普遍存在尖峰、厚尾等現(xiàn)象，研究結(jié)果顯示采用GARCH(1,1)-t模型對擬合我國上市銀行間收益率序列的邊緣分布是合適的。其次，基于常相關(guān)Copula模型和時變相關(guān)Copula模型對我國上市銀行間的相關(guān)性進(jìn)行了詳細(xì)的研究，結(jié)果表明我國銀行間常相關(guān)性很強(qiáng)，而銀行間的相關(guān)系數(shù)是不斷變化的，而且圍繞著某一固定值上下波動，走勢非常相似。再次，在時變相關(guān)的研究基礎(chǔ)上，利用Z檢驗法對上市銀行間的相關(guān)結(jié)構(gòu)進(jìn)行變點檢驗，結(jié)果表明在2008年9月18日，大多數(shù)銀行間的相關(guān)結(jié)構(gòu)發(fā)生了變化。因此，本文以2008年9月18日作為分水嶺，基于CoVaR結(jié)合分位數(shù)回歸技術(shù)研究美國金融危機(jī)前后我國國有銀行對股份制商業(yè)銀行的風(fēng)險溢出強(qiáng)度的變化。研究結(jié)果表明：在q=0.05的情況下，危機(jī)后，我國國有銀行對大部分股份制商業(yè)銀行的風(fēng)險溢出強(qiáng)度增大，特別是對民生銀行，，風(fēng)險溢出強(qiáng)度達(dá)40%以上。 當(dāng)前，監(jiān)管當(dāng)局在確定系統(tǒng)重要性銀行時提出不僅要考慮銀行的規(guī)模因素，更要考慮由于銀行間的相互關(guān)聯(lián)性，考慮單個銀行陷入困境對其他銀行的風(fēng)險溢出及風(fēng)險溢出強(qiáng)度。本文對我國上市銀行間的相關(guān)性研究及量化風(fēng)險溢出強(qiáng)度首先有助于金融監(jiān)管部門及時捕獲銀行間的風(fēng)險強(qiáng)度，進(jìn)而對高風(fēng)險溢出機(jī)構(gòu)進(jìn)行監(jiān)測和管理，從而維護(hù)金融市場的穩(wěn)定；其次為微觀金融主體進(jìn)行投資分析和投資組合提供依據(jù)。
[Abstract]:With the acceleration of financial globalization and financial innovation, the relationship between financial institutions has gradually become closer. In China, banking is the core of the financial system. Banks and banks become more closely connected through interrelated behaviors such as business transactions. How to depict the correlation between banks and how to measure the risk spillovers between banks is a problem that can not be ignored. Copula theory has many advantages in analyzing the correlation structure between variables, which can describe the nonlinear and asymmetric tail correlation between variables. At the same time, the variable structure Copula can accurately find the variation point of the correlation structure between variables. To provide the basis for the study of risk spillover. In this paper, ten banks of listed banks in China are selected as samples, and the correlation of listed banks, especially the time-varying correlation and risk spillover effect, is analyzed empirically. In the process of empirical research, first of all, financial time series generally exist peak and thick tail phenomena. The results show that the GARCH (1 ~ 1) -t model is suitable to fit the marginal distribution of the return series between listed banks in China. Secondly, based on the regular correlation Copula model and the time-varying correlation Copula model, the correlation among the listed banks in China is studied in detail. The results show that the correlation coefficient between the banks in our country is very strong, and the correlation coefficient between banks is constantly changing. And around a fixed value up and down, the trend is very similar. Thirdly, based on the research of time-varying correlation, we use Z test method to test the correlation structure of listed banks. The results show that most of the interbank correlation structures have changed on September 18, 2008. Therefore, this paper takes September 18, 2008 as the watershed, based on CoVaR and quantile regression technology, to study the change of risk spillover intensity between Chinese state-owned banks and joint-stock commercial banks before and after the American financial crisis. The results show that after the crisis, the risk spillover intensity of state-owned banks to most joint-stock commercial banks increases, especially to Minsheng Bank, and the risk spillover intensity is more than 40%. At present, when determining systemically important banks, regulators should consider not only the scale factors of banks, but also the risk spillover and risk spillover intensity of a single bank into a dilemma to other banks due to the interrelationship between banks. This paper studies the correlation between listed banks in China and quantifies the risk spillover intensity. Firstly, the financial supervision department can catch the risk intensity among banks in time, and then monitor and manage the high risk spillover institutions. In order to maintain the stability of the financial market; Secondly, it provides the basis for investment analysis and portfolio analysis.
【學(xué)位授予單位】：湖南大學(xué)
【學(xué)位級別】：碩士
【學(xué)位授予年份】：2012
【分類號】：F832.3;F224

【參考文獻(xiàn)】

相關(guān)期刊論文 前10條

1 陸靜;張佳;;中國上市銀行系統(tǒng)重要性評估[J];金融論壇;2011年09期

2 張杰;劉偉;;基于時變Copula的股票市場相關(guān)性分析[J];商業(yè)經(jīng)濟(jì);2010年07期

3 郭晨;;我國銀行同業(yè)拆借市場交易特征及風(fēng)險傳染研究[J];經(jīng)濟(jì)研究導(dǎo)刊;2010年02期

4 馬君潞;范小云;曹元濤;;中國銀行間市場雙邊傳染的風(fēng)險估測及其系統(tǒng)性特征分析[J];經(jīng)濟(jì)研究;2007年01期

5 宋群英;;基于Copula函數(shù)的系統(tǒng)重要性銀行的傳染性研究[J];金融與經(jīng)濟(jì);2011年10期

6 趙學(xué)雷;艾永芳;;基于Copula-GARCH的金融市場時變相關(guān)性分析[J];科學(xué)決策;2010年06期

7 謝福座;;基于CoVaR方法的金融風(fēng)險溢出效應(yīng)研究[J];金融發(fā)展研究;2010年06期

8 譚治國;蔡乙萍;;分位數(shù)回歸在風(fēng)險管理中的應(yīng)用[J];統(tǒng)計與決策;2006年17期

9 任仙玲;張世英;;基于Copula函數(shù)的金融市場尾部相關(guān)性分析[J];統(tǒng)計與信息論壇;2008年06期

10 李守偉;何建敏;龔晨;;銀行風(fēng)險傳染隨機(jī)模型研究[J];統(tǒng)計與信息論壇;2010年12期

相關(guān)博士學(xué)位論文 前1條

1 萬陽松;銀行間市場風(fēng)險傳染機(jī)制與免疫策略研究[D];上海交通大學(xué);2007年

相關(guān)碩士學(xué)位論文 前2條

1 路婷;基于風(fēng)險傳染角度下銀行系統(tǒng)性風(fēng)險測度研究[D];蘇州大學(xué);2009年

2 高雙雙;基于變結(jié)構(gòu)Copula函數(shù)的上證綜指波動溢出效應(yīng)研究[D];青島大學(xué);2010年

本文編號：2304601