【作者】 陈雷；

【导师】 陈守东；

【作者基本信息】 吉林大学 ， 数量经济学， 2004， 硕士

【摘要】 目前我国债券市场已经形成了银行间债券市场为主体、交易所市场和商业银行柜台市场并存、场内市场和场外市场并存的市场格局。交易主体、交易品种以及托管方式等方面的差异是造成我国债券市场分割局面的主要原因。本文主要研究在这种分割的状态下，我国债券市场内部各子市场之间如何相互影响,并且从实证角度分析了我国债券市场的波动状况及债券市场波动的影响因素。第一章对我国债券市场的发展现状进行了阐述，从理论上分析了交易主体、交易品种以及托管方式等方面的差异是造成我国债券市场分割局面的主要原因，并且论述了这种分割状态所带来的诸多问题。本章的后部分结合我国债券市场2002年的走势分析影响债券市场波动的因素。第二章主要介绍参数变量的选取、定义、研究范围及资料来源以及在文章中所用到的模型：GARCH类模型、VaR模型、协整理论、ECM模型、Granger因果检验模型以及波动溢出效应模型。第三章的第一部分首先对债券总指数进行GARCH建模，估计出债券市场总指数的条件异方差，然后根据基于GARCH模型的计算VaR的方差协方差模型： 来计算债券市场收益率的VaR。接下来我们用同样方法计算同时期的股票市场收益率的VaR值，并且将债券市场和股票市场的收益和风险进行对比分析，发现在这段时间内股票市场的风险大约是债券市场的十倍左右，而股票市场的收益远远小于债券市场的收益。第三章的第二部分对债券市场的内部结构进行了实证分析。首先，对交易所国债指数和银行间国债指数进行协整分析，得到具有标准化协整系数的协整关系估计为： <WP=46>从协整关系中我们可以看出交易所国债指数和银行间国债指数之间存在同方向变动关系。银行间债券市场和交易所债券市场指数之间显著的协整关系说明两个市场的从长期来看是可以整合的，存在共同的长期趋势。然后建立ECM模型，结果表明银行间债券市场和交易所债券市场的收益率存在长期协整趋势，并且银行间债券市场收益率变化的“示范效应”显著，同时他们的短期波动过程存在相异的调整模式。这一结论在波动溢出效应模型中也得到了验证。从脉冲响应图中我们看到：两个市场处于行政割裂的状态下,两市场的信号传导机制并不通畅,银行间市场债券收益率调整的信号传导到交易所市场需要很长一段时间。两市场信号传导机制不畅导致的收益率调整时滞必然会增大债券市场的效率损失。第三章的第三部分对债券市场外部影响因素进行实证分析。首先对债券总指数和上证综指进行协整分析，得到具有标准化协整系数的协整关系估计为：从协整关系我们可以看出债券总指数和上证综指之间存在反方向变动关系，这种相互影响主要是通过资金量在两个市场的流动来实现的。然后建立ECM模型，ECM模型表明债券市场和股票市场的收益率存在长期协整趋势，并且股票市场收益率变化的“示范效应”显著，同时他们的短期波动过程存在相异的调整模式。这种“示范效应”，在后面的“Granger 因果检验”和“溢出效应”的分析中得到确证。通过建立波动溢出效应模型，我们看到上证综指收益率的波动对债券<WP=47>总指数收益率的波动存在显著影响，而债券总指数收益率波动对上海综指收益率的波动影响不显著。这意味着股票市场对债券市场的波动溢出效应显著存在。这种溢出效应的非对称性表明，波动性存在单方向的影响。第四章为实证结果分析以及政策建议。我国债券市场分割的现状依然存在，比如通过实证分析我们发现两个市场处于行政割裂的状态下,两市场的信号传导机制并不通畅,银行间市场债券收益率调整的信号传导到交易所市场需要很长一段时间。但是目前中国债券市场统一的条件已经逐渐成熟，这主要表现在(1)不同市场间收益率差距缩小(2)目前银行间债券市场和交易所债券市场指数的关联性增强(3)跨市场交易品种增多。 在这样的条件下，我们认为在促进债市统一进程深化发展方面，我们应该(1)进一步提高市场主体的交叉程度(2)增加可进行跨市交易的债券品种(3)建立统一的托管清算系统(4)统一监管部门。更多还原

【Abstract】 The bond market in China has already formed the structure that the interbank bond market has been the main body, stock exchange bond markets and commercial bank counter markets are compatible at present. The differences in trade subject, the variety of trade and way of trusteeship, etc. are the main reasons that cause the segmentation situation of bond market. We mainly do some research on the inner structure of the bond market under the segmentation situation, that is , how is the relationship between each sub markets and how the markets influence each other. Then we analyse the volatilities state of the bond market and the influence factors of the bond market volatilities.Chapter one of this text has explained the current situation of China bond market, and analysed the main reasons that cause the segmentation situation of bond market. Discuss a great deal of questions that the state brings. Then we analyse the influence factors of the bond market combined the tendency of bond market in 2002.Chapter two introduced choosing, defines of the parameter variables, research range , materials source and model used in the article: GARCH model , VaR model , cointegration theory , ECM model , Granger causality test , spillover effects model.First part of chapter three at first set up Garch model on bond index, estimate the conditional heteroscedasticity, then according to we calculate VaR of the yield of bond market. And then we calculate VaR of the stock market in the same period with the same method, and compare the yield and risk of the bond market to that of stock market, we find that the risk of the stock market <WP=49>is probably about ten times of the bond market within this period of time, and the yield of the stock market is far smaller than the yield of the bond market.Second part of chapter three have carried on analysis to the inner structure of the bond market. First of all, analyse the co-integration relation of stock exchange treasury bond index and interbank treasury bond index, we receive standardized co-integration relation to be.We can find out stock exchange treasury bond index and interbank treasury bond index have the same direction change relation. The remarkable co-integration relation prove that two market can be combined on long terms, and the common secular trend exists.Then set up ECM model, the result indicates that there is long-term co-integration trend in the yield of the interbank bond market and the stock exchange bond market, and " demonstration effect " of the interbank bond market is remarkable, at the same time their volatilities in a short time has different adjustment modes. This conclusion is proved in the volatility spillover effect model. We see from the picture of pulse respond that under segmentation situation of bond market , conduction mechanism of the signals of the two markets is not unobstructed. Third part of chapter three have carried on analysis to the external factors of the bond market. First of all, analyse the co-integration relation of bond index and Shanghai exchange stock index, we receive standardized co-integration relation to be. <WP=50>We can find out the yield of bond index and the yield of index of Shanghai Stock Exchange have reverse direction change relation, this kind of influence is realized by fund flow.Then set up ECM model, the result indicates that there is long-term co-integration trend in the yield of the bond market and the stock market, and " demonstration effect " of the stock market is remarkable, at the same time their volatilities in a short time has different adjustment modes. Through the spillover effect model, we see volatility of Shanghai Stock Exchange index has influence to volatility of bond market index. This means that spillover effect of stock market volatility is significant. The asymmetry of spillover effect indicates that volatility has one direction influence.Chapter four is results analysis and policy recommendations. The segmentation situation of bond market still exists,But the unified condition更多还原