我国企业债券信用价差影响因素实证研究
发布时间：2018-07-13 16:19
【摘要】：债券融资是近年我国资本市场发展极为迅速的融资形式,信用价差能够反映债券的风险水平,是债券定价和投资策略的重要考量。影响信用价差的因素是多方面的,目前对这些因素的实证研究还没有统一的结论。本文以探究影响中国企业债券信用价差的因素为主要研究目的,在对国外重要成熟理论模型和实证结果进行梳理的基础上,结合我国企业债券市场发展的客观情况和目前已有的国内研究,提出了比较全面的信用价差影响因素,并在定性分析的基础上给出了定量分析的实证研究。 最早开始研究企业债券信用价差的是莫顿,他在1974年以布莱克-斯科尔斯的期权定价模型为基础,提出了针对企业债券进行估值的定价模型——结构化模型,此模型假定企业债券的违约发生在发债企业的价值下降到某一临界点以下。鉴于莫顿模型的假设条件众多而且要求严格,之后的多位学者在其基础上拓展了莫顿模型。Black and Cox (1976)模型便是其中比较有代表性的模型。该模型的进步体现在：一是考虑了违约风险和利率风险,二是考虑了严格绝对的优先受偿权的误差。现实中,债券的信用价差是连续的,而结构化模型假设信用价差在债券到期时等于零,这与现实不同。基于这个考虑,首达时间模型、跳跃扩散模型和随机利率模型对结构模型进行了拓展。第二类著名的信用风险定价模型是简化模型。该模型假设违约不可预测并服从跳跃过程,从违约概率和偿付率作为研究角度,这方面重要的模型是Jarrow, Lando and Turnbull (1997)模型和Duffie and Singleton(1999)模型。随着研究的进一步推进,越来越多的学者发现了“信用价差之谜”,即企业债券实际的信用价差远远大于预期违约损失,代表性的研究分为两类：信用价差分解理论和信用风险分解困境理论。综合来看,前者认为信用价差是由信用价差是税收、风险溢价、流动性溢价、系统风险、企业特定风险等造成的,后者认为系统性风险不可能分散掉,能够完全分散掉违约风险的投资组合在现实中很难找到,所以非预期损失难以避免。 对信用价差影响因素的实证研究也有很多,都是针对某个或某类因素展开实证分析。本文总结归纳了这些因素并将其分为两类：第一类,宏观因素,包括短期利率、国债利率差、股票收益率、股市波动率、经济增长率、居民消费价格指数的变动、货币供应量的变动等；第二类包括微观因素,即企业自身因素,比如：信用等级、杠杆率、企业资产价值、剩余期限等。理论上讲,宏观因素与信用价差的相关性表现在：短期利率和信用价差是负相关的,利率上升说明经济发展良好,企业未来盈利能力增强,信用价差减小；股票收益率与信用价差是负相关的；国债利率差和信用价差是负相关的,因为国债收益率曲线斜率增加说明有经济增长趋势,信用价差减小；经济增长率和信用价差负相关；居民消费价格指数的变动和信用价差正相关相关；股市波动率和信用价差负相关；货币供应量的变化与信用价差负相关。微观因素与企业债券信用价差的相关性表现在：信用评级和信用价差负相关；杠杆率与信用价差正相关；剩余期限与信用价差正相关；企业资产价值与信用价差负相关。 从国内外对这些因素的实证研究结果看,对于某些因素(如短期利率等)的影响没有绝对统一的结论,这与学者们各自研究的假设有重要关系,因此,得到基于中国企业债市场实际情况下信用价差与这些因素的相关关系是本文实证研究要解决的问题,并在此基础上辨别出因素对信用价差解释程度的结构。 在确定本文要研究的影响因素之后,本文开始分别建立时间序列的向量自回归模型和横截面数据的微观的模型。首先本文对被解释变量和解释变量进行设置。笔者发现很多研究直接使用了企业债券和国债的到期收益率而不是即期收益率来计算信用价差,这样就没有考虑到利率期限结构的影响,严格来说是不准确的。本文宏观影响因素的研究区间的选择将本次金融危机前后都考虑进去,宏观影响因素的变量使用月度数据,使用了5年和10年剩余期限的企业债券信用价差作为被解释变量,分别建立了VAR模型分析在不同剩余期限的情况下,因素对信用价差的影响有何变化,还对模型进行了脉冲响应分析与方差分解。微观模型选取了2012年6月29日的上市公司企业债券,通过国泰安系列研究数据库给出的财务数据、评级情况及剩余期限,计算出各自的即期收益率,并通过插值法计算出对应的国债即期收益率得到对应的信用价差,进而建立多元回归模型。这两个模型的建立并未直接套用国外流行的结构化模型,而是考虑了我国债市的具体情况,借鉴了罗斯套利定价模型的思想,建立多元回归方程。宏观模型的实证结果显示：两种不同剩余期限的信用价差对宏观因素的影响存在7个月的滞后,但是各自的影响显著的因素发生了变化。对于剩余期限较长的企业债券而言,信用价差与短期利率是正相关的关系,这可以从研究的时间区间内的经济具体运行情况得到现实依据。对于股市波动率、股票收益率、经济增长率的影响,得到的结论都与理论相符。微观模型的实证结果显示：模型的拟合程度很高,信用等级和剩余期限对信用价差的影响最为显著,并且结论与理论假设一致。 本文的创新之处表现在：第一,时间区间跨度长,涵盖了金融危机和债券市场发展最为迅速的几年；第二,对信用价差影响因素的探索更加全面；第三,被解释变量使用了即期收益率,考虑到了利率期限结构的影响,计算更为准确,并且使用交易所即期收益率曲线,能更好的代表整个企业债券市场；第四,实证结果与国外已有研究有不同之处,并得到合理的解释。 论文框架结构如下： 第一部分：绪论。第一节简单介绍了我国企业债券市场的发展现状和制度建设情况,列出了目前我国债券市场的规模和债券品种,陈述了我国债券市场存在的问题,针对目前的经济环境展望了下一步企业债券的发展。第二节阐述了本文的研究对信用价差理论和现实的意义之处。第三节给出了本文研究对象的相关定义的界定,接着在第四节阐述了本文的研究思路和使用的模型及方法,第五节给出了研究的创新和不足。 第二部分：企业债券信用价差利率综述。该部分是对本文研究问题的理论综述,首先根据信用价差的组成和国外国内的研究成果,总结了目前该问题的研究现状,归纳和对比了相同结论和存在的差异及争议。第二节重点陈列和描述了信用风险定价模型：结构化模型和简化模型,比较了各个模型进步之处,也指出了其中的不足。第三节交待了对前面模型存在的质疑,并主要对比了“信用价差之谜”的两种重要理论。 第三部分：影响因素分析。这部分既是对前面文献综述的一个归纳,也是对我国企业债券市场进行的适应性分析,提出了宏观层面8个影响因素和企业个体方面4个影响我国企业债信用价差的因素,并探讨了各因素可能对信用价差产生影响的方向。为后面建立模型提供了依据。 第四部分：实证研究设计。分别基于VAR模型和多元线性回归实证分析了宏观和微观影响因素。宏观研究以时间序列数据为基础,确定了滞后阶数得出模型的方程组,并通过脉冲响应分析给出了各个内生变量一个标准差冲击影响信用价差预测标准差的力度和滞后期,通过方差分解得到各个内生变量对残差冲击影响信用价差预测标准差的解释结构。微观研究以横截面数据为样本,选取了2012年6月29日17只公司债券的数据,建立多元回归方程,最后给出了各自的实证结论。 第五部分：总结与建议是结合实证研究结论,给出了政策层面的建议。 第六部分：未来研究展望。阐述了本文研究的不足之处,并给出了改进的方法。
[Abstract]:Bond financing is a very rapid financing form in China's capital market development in recent years. The credit spread can reflect the risk level of the bond. It is an important consideration of the bond pricing and investment strategy. There are many factors affecting the credit spread. At present, there is no unified conclusion on the empirical study of these factors. This paper is to explore the influence of Chinese enterprises. The factor of the credit spread of the bond is the main purpose. On the basis of combing the important mature theoretical model and the empirical results abroad, combining the objective situation of the development of the corporate bond market in China and the existing domestic research, this paper puts forward a relatively comprehensive factor of the credit spread, and gives the qualitative analysis on the basis of the qualitative analysis. An empirical study of quantitative analysis.
In 1974, on the basis of Black Scholes's option pricing model, he proposed a pricing model based on the option pricing model of Black - Scholes, a structured model, which assumes that the default of corporate bonds falls below a certain threshold. In view of the many assumptions and strict requirements of the moton model, many scholars later expanded the moton model.Black and Cox (1976) model to be a representative model. The progress of the model is as follows: first, considering the risk of breach of contract and interest rate risk, and two considering the strict and absolute priority of compensation. In reality, the credit spread of the bond is continuous, and the structured model assumes that the credit spreads are equal to zero when the bonds expire. This is different from the reality. Based on this consideration, the first time model, the jump diffusion model and the random interest rate model are extended to the structural model. The second famous credit risk pricing models are simplified. The model assumes that default is unpredictable and obeys the jumping process, from the probability of default and the rate of payment as the research angle. The important model in this respect is Jarrow, Lando and Turnbull (1997) model and Duffie and Singleton (1999) model. With the further advance of the research, more and more scholars have discovered the "riddle of credit spread", that is, The actual credit spread of corporate bonds is far greater than the expected loss of default. The representative study is divided into two categories: the theory of credit spread decomposition and the theory of credit risk decomposition dilemma. It is considered that systematic risk can not be dispersed. It is difficult to find a portfolio that can completely disperse the risk of default. Therefore, it is difficult to avoid unanticipated losses.
There are a lot of empirical studies on the factors affecting the credit spread, which are aimed at an empirical analysis of some or some factors. This paper sums up these factors and divides them into two categories: the first category, the macro factor, including the short-term interest rate, the interest rate difference of national debt, the stock return rate, the stock market volatility, the economic growth rate, and the consumer price index. Changes, changes in the amount of money supply, and so on; the second category includes micro factors, namely, enterprise own factors, such as credit rating, leverage ratio, enterprise asset value, and residual maturity. In theory, the correlation between macro factors and credit spreads is manifested in the negative correlation between the short-term interest rate and the credit spread, and the rise in interest rates shows that the economic development is good. The future profitability of the enterprise is enhanced and the credit spread is reduced; the stock return is negatively related to the credit spread; the interest rate difference between the national debt and the credit spreads is negatively related, because the increase in the slope of the yield curve shows the trend of economic growth, the decrease of the credit spread, the negative correlation between the economic growth rate and the credit spread, and the consumer price index. The change of the number is positively related to the credit spread; the stock market volatility is negatively related to the credit spread; the change of the money supply is negatively related to the credit spread. The correlation between the micro factors and the credit spread of the corporate bonds is negative related to the credit rating and the credit spread; the leverage ratio is positively related to the credit spread, and the remaining period and the credit price. Positive correlation; negative correlation between corporate asset value and credit spreads.
From the empirical results of these factors at home and abroad, there is no absolute unified conclusion on some factors (such as short-term interest rate), which is important to the hypothesis of the scholars' respective studies. Therefore, the correlation between the credit price difference and these factors is an empirical study based on the actual situation of the Chinese enterprise debt market. To solve these problems, we can identify the structure of factors explaining the credit spreads.
After determining the influencing factors of this paper, this paper begins with the establishment of the vector autoregressive model of the time series and the microscopic model of the cross section data. First, this paper sets up the explanatory variables and the explanatory variables. I find that many studies have directly used the maturity yield of the corporate bonds and the Treasury bonds instead of the immediate collection. It does not take into account the impact of the term structure of interest rates, which is strictly inaccurate. The choice of the research interval of this macro impact factor will be taken into account before and after the financial crisis, the variables of the macroeconomic factors use the monthly data, and the corporate bond credit of 5 and 10 years is used. As an explanatory variable, the VAR model is established to analyze the changes in the impact of the factors on the credit spread and the analysis of the impulse response and the variance decomposition. The micro model selects the corporate bond of the listed company in June 29, 2012, and gives the data of the Tai'an series research database. The financial data, the rating situation and the remaining time limit are calculated, and the corresponding credit spreads are calculated by the interpolation method, and then the multiple regression model is established. The establishment of the two models does not directly apply the structured model of foreign flows, but consider the debt market of our country. In particular, we use the idea of Ross's arbitrage pricing model to establish a multiple regression equation. The empirical results of the macro model show that the impact of the credit spreads on the macro factors of the two different remaining periods has 7 months lag, but the significant factors have changed. For the longer remaining period of corporate bonds, The relationship between the credit spread and the short-term interest rate is a positive correlation, which can be based on the actual operation of the economy in the time interval of study. The results of the stock market volatility, stock return and economic growth rate are consistent with the theory. The empirical results of the micro model show that the model has a high degree of fitting. The effect of grade and residual period on credit spreads is the most significant, and the conclusion is consistent with the theoretical hypothesis.
The innovation of this paper is as follows: first, the length of the time interval is long, covering the financial crisis and the most rapid development of the bond market; second, the exploration of the factors affecting the credit spreads is more comprehensive; third, the explanatory variables use the immediate return rate, taking into account the effect of the term structure of the interest rate, and the calculation is more accurate, and The use of the exchange rate curve of the exchange can better represent the whole enterprise bond market. Fourth, the empirical results are different from those in foreign countries and have a reasonable explanation.
The framework of the paper is as follows:
The first section is introduction. The first section briefly introduces the development status and system construction of our country's enterprise bond market, lists the scale of the current bond market and the bond variety, states the problems existing in the bond market in our country, and looks forward to the development of the next enterprise bond in view of the current economic environment. The second section expounds this article. The research on the significance of the theory and reality of the credit spread. The third section gives the definition of the relevant definition of the object of this paper, then the fourth section of the research ideas and the use of the model and method, the fifth section gives the innovation and deficiency of the research.
The second part: a summary of the interest rate of the credit spread of corporate bonds. This part is a theoretical summary of the study of this paper. Firstly, according to the composition of the credit spread and the research results from foreign countries, the research status of the present problem is summarized and the differences and disputes of the same conclusion and existence are summarized and compared. The second section focuses on the display and description of the letter. The risk pricing model, structured model and simplified model, compares the progress of each model and points out its shortcomings. The third section deals with the question of the previous model, and compares the two important theories of the "riddle of the credit spread".
The third part: the analysis of influencing factors. This part is not only a summary of the previous literature review, but also an adaptability analysis of the enterprise bond market in China. It also puts forward 8 factors on the macro level and 4 factors affecting the credit spread of corporate bonds in our country, and probes into the possible difference between the factors and the credit spreads. The direction of the influence provides a basis for building a model later.
The fourth part: empirical research design. Based on the VAR model and multiple linear regression analysis, the macro and micro factors are analyzed respectively. Based on the time series data, the macroscopic study determines the equations of the lagging order and gives a standard difference impact credit price of each endogenous variable by the impulse response analysis. The intensity and stagnation of the difference prediction standard deviation, through the variance decomposition, the interpretation structure of the standard deviation of the prediction standard difference affecting the residual impact on the credit spread is obtained by the variance decomposition. In the microscopic study, the data of the cross section are taken as the sample, and the data of 17 corporate bonds in June 29, 2012 are selected and the multiple regression equation is set up. Finally, the empirical conclusions are given.
The fifth part: the conclusion and suggestion are combined with the conclusions of empirical research, giving policy recommendations.
The sixth part: the prospect of future research. The shortcomings of this study are expounded, and the improvement method is given.
【学位授予单位】：西南财经大学
【学位级别】：硕士
【学位授予年份】：2013
【分类号】：F832.51;F224
本文编号：2120018
[Abstract]:Bond financing is a very rapid financing form in China's capital market development in recent years. The credit spread can reflect the risk level of the bond. It is an important consideration of the bond pricing and investment strategy. There are many factors affecting the credit spread. At present, there is no unified conclusion on the empirical study of these factors. This paper is to explore the influence of Chinese enterprises. The factor of the credit spread of the bond is the main purpose. On the basis of combing the important mature theoretical model and the empirical results abroad, combining the objective situation of the development of the corporate bond market in China and the existing domestic research, this paper puts forward a relatively comprehensive factor of the credit spread, and gives the qualitative analysis on the basis of the qualitative analysis. An empirical study of quantitative analysis.
In 1974, on the basis of Black Scholes's option pricing model, he proposed a pricing model based on the option pricing model of Black - Scholes, a structured model, which assumes that the default of corporate bonds falls below a certain threshold. In view of the many assumptions and strict requirements of the moton model, many scholars later expanded the moton model.Black and Cox (1976) model to be a representative model. The progress of the model is as follows: first, considering the risk of breach of contract and interest rate risk, and two considering the strict and absolute priority of compensation. In reality, the credit spread of the bond is continuous, and the structured model assumes that the credit spreads are equal to zero when the bonds expire. This is different from the reality. Based on this consideration, the first time model, the jump diffusion model and the random interest rate model are extended to the structural model. The second famous credit risk pricing models are simplified. The model assumes that default is unpredictable and obeys the jumping process, from the probability of default and the rate of payment as the research angle. The important model in this respect is Jarrow, Lando and Turnbull (1997) model and Duffie and Singleton (1999) model. With the further advance of the research, more and more scholars have discovered the "riddle of credit spread", that is, The actual credit spread of corporate bonds is far greater than the expected loss of default. The representative study is divided into two categories: the theory of credit spread decomposition and the theory of credit risk decomposition dilemma. It is considered that systematic risk can not be dispersed. It is difficult to find a portfolio that can completely disperse the risk of default. Therefore, it is difficult to avoid unanticipated losses.
There are a lot of empirical studies on the factors affecting the credit spread, which are aimed at an empirical analysis of some or some factors. This paper sums up these factors and divides them into two categories: the first category, the macro factor, including the short-term interest rate, the interest rate difference of national debt, the stock return rate, the stock market volatility, the economic growth rate, and the consumer price index. Changes, changes in the amount of money supply, and so on; the second category includes micro factors, namely, enterprise own factors, such as credit rating, leverage ratio, enterprise asset value, and residual maturity. In theory, the correlation between macro factors and credit spreads is manifested in the negative correlation between the short-term interest rate and the credit spread, and the rise in interest rates shows that the economic development is good. The future profitability of the enterprise is enhanced and the credit spread is reduced; the stock return is negatively related to the credit spread; the interest rate difference between the national debt and the credit spreads is negatively related, because the increase in the slope of the yield curve shows the trend of economic growth, the decrease of the credit spread, the negative correlation between the economic growth rate and the credit spread, and the consumer price index. The change of the number is positively related to the credit spread; the stock market volatility is negatively related to the credit spread; the change of the money supply is negatively related to the credit spread. The correlation between the micro factors and the credit spread of the corporate bonds is negative related to the credit rating and the credit spread; the leverage ratio is positively related to the credit spread, and the remaining period and the credit price. Positive correlation; negative correlation between corporate asset value and credit spreads.
From the empirical results of these factors at home and abroad, there is no absolute unified conclusion on some factors (such as short-term interest rate), which is important to the hypothesis of the scholars' respective studies. Therefore, the correlation between the credit price difference and these factors is an empirical study based on the actual situation of the Chinese enterprise debt market. To solve these problems, we can identify the structure of factors explaining the credit spreads.
After determining the influencing factors of this paper, this paper begins with the establishment of the vector autoregressive model of the time series and the microscopic model of the cross section data. First, this paper sets up the explanatory variables and the explanatory variables. I find that many studies have directly used the maturity yield of the corporate bonds and the Treasury bonds instead of the immediate collection. It does not take into account the impact of the term structure of interest rates, which is strictly inaccurate. The choice of the research interval of this macro impact factor will be taken into account before and after the financial crisis, the variables of the macroeconomic factors use the monthly data, and the corporate bond credit of 5 and 10 years is used. As an explanatory variable, the VAR model is established to analyze the changes in the impact of the factors on the credit spread and the analysis of the impulse response and the variance decomposition. The micro model selects the corporate bond of the listed company in June 29, 2012, and gives the data of the Tai'an series research database. The financial data, the rating situation and the remaining time limit are calculated, and the corresponding credit spreads are calculated by the interpolation method, and then the multiple regression model is established. The establishment of the two models does not directly apply the structured model of foreign flows, but consider the debt market of our country. In particular, we use the idea of Ross's arbitrage pricing model to establish a multiple regression equation. The empirical results of the macro model show that the impact of the credit spreads on the macro factors of the two different remaining periods has 7 months lag, but the significant factors have changed. For the longer remaining period of corporate bonds, The relationship between the credit spread and the short-term interest rate is a positive correlation, which can be based on the actual operation of the economy in the time interval of study. The results of the stock market volatility, stock return and economic growth rate are consistent with the theory. The empirical results of the micro model show that the model has a high degree of fitting. The effect of grade and residual period on credit spreads is the most significant, and the conclusion is consistent with the theoretical hypothesis.
The innovation of this paper is as follows: first, the length of the time interval is long, covering the financial crisis and the most rapid development of the bond market; second, the exploration of the factors affecting the credit spreads is more comprehensive; third, the explanatory variables use the immediate return rate, taking into account the effect of the term structure of the interest rate, and the calculation is more accurate, and The use of the exchange rate curve of the exchange can better represent the whole enterprise bond market. Fourth, the empirical results are different from those in foreign countries and have a reasonable explanation.
The framework of the paper is as follows:
The first section is introduction. The first section briefly introduces the development status and system construction of our country's enterprise bond market, lists the scale of the current bond market and the bond variety, states the problems existing in the bond market in our country, and looks forward to the development of the next enterprise bond in view of the current economic environment. The second section expounds this article. The research on the significance of the theory and reality of the credit spread. The third section gives the definition of the relevant definition of the object of this paper, then the fourth section of the research ideas and the use of the model and method, the fifth section gives the innovation and deficiency of the research.
The second part: a summary of the interest rate of the credit spread of corporate bonds. This part is a theoretical summary of the study of this paper. Firstly, according to the composition of the credit spread and the research results from foreign countries, the research status of the present problem is summarized and the differences and disputes of the same conclusion and existence are summarized and compared. The second section focuses on the display and description of the letter. The risk pricing model, structured model and simplified model, compares the progress of each model and points out its shortcomings. The third section deals with the question of the previous model, and compares the two important theories of the "riddle of the credit spread".
The third part: the analysis of influencing factors. This part is not only a summary of the previous literature review, but also an adaptability analysis of the enterprise bond market in China. It also puts forward 8 factors on the macro level and 4 factors affecting the credit spread of corporate bonds in our country, and probes into the possible difference between the factors and the credit spreads. The direction of the influence provides a basis for building a model later.
The fourth part: empirical research design. Based on the VAR model and multiple linear regression analysis, the macro and micro factors are analyzed respectively. Based on the time series data, the macroscopic study determines the equations of the lagging order and gives a standard difference impact credit price of each endogenous variable by the impulse response analysis. The intensity and stagnation of the difference prediction standard deviation, through the variance decomposition, the interpretation structure of the standard deviation of the prediction standard difference affecting the residual impact on the credit spread is obtained by the variance decomposition. In the microscopic study, the data of the cross section are taken as the sample, and the data of 17 corporate bonds in June 29, 2012 are selected and the multiple regression equation is set up. Finally, the empirical conclusions are given.
The fifth part: the conclusion and suggestion are combined with the conclusions of empirical research, giving policy recommendations.
The sixth part: the prospect of future research. The shortcomings of this study are expounded, and the improvement method is given.
【学位授予单位】：西南财经大学
【学位级别】：硕士
【学位授予年份】：2013
【分类号】：F832.51;F224
【参考文献】
相关期刊论文 前7条
1 张燃;;信用价差变化的决定因素——一个宏观视角[J];当代财经;2008年09期
2 王一鸣,李剑峰;我国债券市场收益率曲线影响因素的实证分析[J];金融研究;2005年01期
3 程文卫;;我国交易所上市企业主体债券利差的影响因素研究[J];生产力研究;2009年08期
4 刘国光,王慧敏;公司债券信用价差和国债收益率动态关系研究[J];山西财经大学学报;2005年05期
5 赵娜;;企业债券信用利差研究综述[J];商业时代;2006年33期
6 赵银寅;田存志;;我国企业债券信用利差的宏观影响因素分析[J];商业时代;2010年34期
7 任兆璋;李鹏;;流动性风险对可违约债券信用利差期限结构的影响[J];系统工程理论方法应用;2006年03期
，本文编号：2120018
本文链接：https://www.wllwen.com/guanlilunwen/zhqtouz/2120018.html
最近更新
教材专著
