我國股市尾部風險度量及尾部相關性研究
本文選題:尾部風險 + 尾部相關; 參考:《暨南大學》2013年碩士論文
【摘要】:極端事件對金融市場的影響巨大,我國股市受08年金融危機所累,至今未能走出發(fā)展的泥潭。極端事件引起的尾部風險越發(fā)受到廣大學者和金融監(jiān)管部門的興趣和關注。如何對尾部風險進行測度,是時下學者們重點研究的課題。 風險價值(Value at risk)是現(xiàn)行風險度量的國際標準。不加區(qū)別的使用VaR對尾部風險進行測量,將造成風險低估的嚴重后果。隨著金融風險度量理論的發(fā)展,,滿足一致性四公理被認為是良好風險測度指標的必要條件。滿足一致性準則的尾部風險度量工具包括CVaR、ER、ES、TCE和TM等指標,本文在簡述,分析各指標的優(yōu)劣后,選擇基于極值理論的VaR和ES作為尾部風險度量工具。 度量和分析股市尾部相關性,可以更加全面深刻的了解股市尾部風險。Copula函數(shù)由于善于捕捉變量間非線性、非對稱及尾部相關性特性,因此常被用來測量資產(chǎn)或股市間的尾部相關性。 本文的實證分兩部分進行,首先是基于改進閾值選取法的POT模型計算VaR和ES,返回檢驗結(jié)果表明不加改進的POT-VaR效果不如前者。實證第二部分采用Copula理論對上證指數(shù)、恒生指數(shù)、納斯達克指數(shù)做兩兩尾相關性度量,結(jié)果表明我國股市與境外成熟股市尾相關性明顯,呈非對稱性,且上尾相關性小于下尾相關性。
[Abstract]:The extreme events have great influence on the financial market. The stock market of our country has not been able to get out of the quagmire of development so far because of the financial crisis of 2008. The tail risk caused by extreme events has attracted more and more attention from scholars and financial regulators. How to measure tail risk is an important research topic for scholars. Value at riskis the current international standard of risk measurement. Using VaR indiscriminately to measure tail risk will result in serious consequences of risk underestimation. With the development of financial risk measurement theory, satisfying the four axioms of consistency is considered to be a necessary condition for good risk measurement. The tail risk measurement tools that satisfy the consistency criterion include CVaRGERE ESTCE and TM. After analyzing the merits and demerits of each index, this paper chooses VaR and es based on extreme value theory as tail risk measurement tools. By measuring and analyzing the tail correlation of the stock market, we can understand the stock market tail risk. Copula function is good at capturing the nonlinear, asymmetric and tail correlation characteristics of the stock market. Therefore, it is often used to measure the tail correlation between assets or stock markets. The empirical results of this paper are divided into two parts. Firstly, the VaR and ES-based POT-VaR are calculated based on the improved threshold selection method. The results of return test show that the effect of POT-VaR without improved POT-VaR is not as good as that of the former. In the second part, we use Copula theory to measure the correlation of Shanghai Stock Exchange Index, Hang Seng Index and Nasdaq Index. The results show that the correlation between Chinese stock market and overseas mature stock market is obvious and asymmetric. The correlation between upper tail and lower tail is smaller than that of lower tail.
【學位授予單位】:暨南大學
【學位級別】:碩士
【學位授予年份】:2013
【分類號】:F224;F832.51
【參考文獻】
相關期刊論文 前10條
1 潘家柱,丁美春;GP分布模型與股票收益率分析[J];北京大學學報(自然科學版);2000年03期
2 高瑩;周鑫;金秀;;GARCH-EVT模型在動態(tài)VaR中的應用[J];東北大學學報(自然科學版);2008年04期
3 楊青;薛宇寧;蔣科;;極端金融風險度量模型述評——基于一致性原理的VaR改進方法[J];復旦學報(自然科學版);2009年06期
4 李悅;程希駿;;上證指數(shù)和恒生指數(shù)的copula尾部相關性分析[J];系統(tǒng)工程;2006年05期
5 花擁軍;張宗益;;極值BMM與POT模型對滬深股市極端風險的比較研究[J];管理工程學報;2009年04期
6 劉宇飛;VaR模型及其在金融監(jiān)管中的應用[J];經(jīng)濟科學;1999年01期
7 陳守東,俞世典;基于GARCH模型的VaR方法對中國股市的分析[J];吉林大學社會科學學報;2002年04期
8 高岳;朱憲辰;;基于極值理論的同業(yè)拆借利率風險度量——基于AR-GARCH-POT方法的VaR值比較研究[J];數(shù)量經(jīng)濟技術經(jīng)濟研究;2009年08期
9 陳希孺;變點統(tǒng)計分析簡介[J];數(shù)理統(tǒng)計與管理;1991年01期
10 歐陽資生;趙霞;;基于指數(shù)回歸模型的極值指數(shù)估計的門限選擇[J];數(shù)理統(tǒng)計與管理;2006年06期
相關碩士學位論文 前2條
1 晏海兵;中國股市波動溢出效應的實證研究[D];重慶大學;2004年
2 李艷杰;變點分析及其在中國證券市場脆弱性研究中的應用[D];西南交通大學;2005年
本文編號:2029998
本文鏈接:http://www.wukwdryxk.cn/guanlilunwen/zhqtouz/2029998.html