中国银行间债券市场企业债信用利差影响因素研究

发布时间：2018-07-23 19:23

【摘要】：企业债是有价证券市场上重要的信用债品种。其不仅是涉及公司资本结构来源的必要融资工具,而且其在二级市场上价值的波动,构成了信用风险定价的基础。企业债券市场蓬勃发展的同时,其价值确定和信用风险管理也对投资者提出了更高的要求。尤其2008年8月全面爆发的美国次贷危机(Subprime Crisis),生动的阐释了信用风险管理的重要性和复杂性。而企业债券自身收益风险非对称性的特点,使信用风险难以充分分散化,为投资者投资管理提出了较高的要求,同时也使其价格的估计较为复杂。而信用利‘差作为企业债的定价的核心更是引起国内理论界和具体实务界的重视。在理论界国内外学者研究多集中于研究信用利差的来源和具体影响因素；在实务界操作中,信用债和利率债的多空组合使投资者更加关注信用利差的变化。随着我国银行间债券市场的蓬勃发展,企业债亦发展迅速,信用利差为理论界和实务界所关注。而对于信用利差究竟是完全来自信用违约风险,还是承载其它因素并不十分清晰。本文根据我国银行间债券市场情况,以Merton(1974)结构化模型为理论基础对我国银行间债券市场企业债信用利差的影响因素来源和大小进行了定性和定量的探索。在研究方法上,本文首先通过对信用利差基础理论分析,根据Merton结构化模型从理论上对信用利差公式进行推导,得出信用利差是资产波动率、利率等变量的函数。但Merton结构化模型是在完美市场假定下给出的一个解析解,为更好的从国银行间企业债市场交投方面考虑,深入挖掘影响企业债信用利差的因素,本文又从我国企业债发行方式、企业债托管量发展趋势、发债主体结构以及二级市场企业债的投资机构结构和交易方式等几个方面对我国企业债的现实背景进行了分析。综合理论基础和我国现实背景,本文提出把银行间企业债信用利差影响因素分为信用风险因素和非信用风险因素两大方面。其中信用风险因素方面又分为两个方向,一是微观层面债券发行主体自身因素,主要包括资本结构、财务杠杆以及发行主体企业性质等因素；二是宏观层面系统性风险,主要包括无风险利率、通货膨胀以及反映经济周期的工业投资增加值、采购经理指数(PMI)等因素。借鉴国内外文献实证分析方法,本文从指数层面和银行间个券层面分别构建时序模型和面板数据模型,对信用利差进行量化分析。综合定性和定量分析,本文有如下分析结论： 1、影响信用利差的信用风险因素方面,微观层面看,反映企业财务杠杆的资产负债率变化对企业债信用利差影响有限。在第三章分析了企业债发行主体构成,以国有企业背景为主,且在企业债交投中未曾发生过实质性的违约事件,这与政府的“隐性担保”分不开,因而投资者更愿关注分析发债主体的背景性质,从而“揣测”其信用风险,这在国内投资经济结构环境中更为实用。 2、影响信用利差的宏观层面看,结构化变量利率因素是影响企业债信用利差的重要因素。不管从指数层面还是个券层面的实证分析中,以10年期国债收益率表示的无风险利率变化和国债收益率期限结构的变化对信用利差负相关。说明二者很好的反映了经济环境的变化,从而投资者预期企业盈利的变化而反映到信用利差的变化上。从加入的10年国债收益率变化的平方看,其系数较为显著,亦即意味着国债收益率变化对信用利差的影响是非线性的。 3、从反映宏观经济价格指标环比通货膨胀看,总体看通货膨胀环比和各评级企业债信用利差正相关,即通货膨胀环比上升或预期上升,各期限企业债信用利差趋于上升。从个券层面分析的通胀指标和信用利差正相关表明信用债对通胀预期较国债更为敏感。 4、沪深300指数波动率和沪深300指数收益率在指数层面和个券层面有较大不同。从指数层面来看,二者变动对信用利差影响微乎其微。不管从期限维度还是评级维度实证分析发现,沪深300指数波动率和沪深300指数收益率的回归系数均趋近于零且不显著。但从个券层面看,沪深300指数波动率和沪深‘300指数收益率对信用利差负相关且显著,虽然相关系数较小。本文分析认为个券层面包含更多的个性化信息,更多的刻画了二者的关系。从这一结果中另一方面也可以看到沪深300指数走势反映了经济周期的变化,至少是企业库存周期的变化趋势。 5、流动性对信用利差有负的影响,即流动性越强,信用利差将会越窄,但从实证结上看这种效果并不十分明显。本文采用基于交易频率的流动性变化,刻画了企业债二级市场流动性的变化,从其变动趋势看,显示较强的品种依赖和时间变化(time-varying)特征。由于银行间债券市场由于“代持”、“过券”等交易手段,虽然成交量放大,但并不能反映流动性引起的信用利差的变化。而从根本上,我国银行间企业债的机构投资者仍以商业银行、保险机构等为主,而此类一般以配置型为主,相较不是特别活跃,从而在实证分析结果上看,流动性对信用利差的影响不是特别大。 6、本文在指数层面实证模型解释信用利差变化的35%左右,个券层面面板数据模型只能解释信用利差变化的20%左右。这与我国债券银行间债券市场发展现状有很密切的关系,我国银行间债券市场目前仍是单边市场,机构投资者同质化严重,以“代持”、“过券”为代表的虚假成交较多,模型并无法规避此中影响,相应解释力较弱。另一方面,从按企业债债项评级的实证分析结果的拟合程度看,随着信用资质的降低,模型的解释力增强,这也与国外实证结果一致。通过指数层面和个券实证结果证结果总体对比发现,二者基本相容。这一方面说明了本文实证结果的稳健性,另一方面也表明了中债登估值数据整体的有效性。
[Abstract]:Corporate debt is an important credit loan in the stock market. It is not only a necessary financing tool that involves the source of the capital structure of the company, but also the fluctuation of its value in the two level market, which constitutes the basis of the pricing of credit risk. The higher requirements, especially the Subprime Crisis, which broke out in August 2008, vividly illustrate the importance and complexity of credit risk management, and the asymmetric characteristics of the corporate bond's own income risk make it difficult to divide the credit risk fully and put forward higher requirements for the investment management of the investors, as well as the investment management. The estimation of the price is more complex, and the core of the credit interest 'bad as the pricing of the enterprise debt is more important to the domestic theorists and the concrete practice circles. Investors pay more attention to the change of credit spreads. With the rapid development of the interbank bond market in China, the corporate debt is developing rapidly. The credit spreads are concerned in the theoretical and practical circles. However, it is not very clear whether the credit spreads are entirely from the credit default risk or the other factors. This paper is based on the interbank debt in China. With the Merton (1974) structural model as the theoretical basis, this paper makes a qualitative and quantitative exploration of the source and size of the factors affecting the credit spreads of corporate bonds in the interbank bond market.
In the research method, this paper first analyses the basic theory of credit spreads and deduces the formula of credit difference based on Merton structural model, and concludes that the credit difference is the function of asset volatility, interest rate and other variables. But the Merton structured model is an analytical solution given under the assumption of perfect market. In this paper, the factors that affect the credit spreads of corporate bonds are deeply excavated in the consideration of the exchange and investment of the corporate bond market between China's banks. This paper also comes into the realistic background of China's enterprise debt from several aspects, such as the way of issuing corporate debt, the development trend of corporate debt trust, the main structure of the debt issuance, the structure of the investment institutions and the trading mode of the two level market enterprise debt. On the basis of the comprehensive theory and the realistic background of our country, this paper divides the factors of credit spreads between banks into two aspects: credit risk factors and non credit risk factors. Among them, the credit risk factors are divided into two directions, one is the main factors of the bond issuing subject in the micro level, mainly including the capital structure, Financial leverage and the nature of the enterprise of the main issue; two is the systemic risk at the macro level, mainly including the risk free interest rate, inflation and the value of the industrial investment that reflect the economic cycle, the purchasing manager index (PMI) and other factors. Do not build time series model and panel data model, carry out quantitative analysis of credit spreads.
1, in terms of credit risk factors, the micro level shows that the change of the asset liability rate of corporate financial leverage has limited influence on the credit spreads of corporate bonds. In the third chapter, the main body of the corporate debt issuance is analyzed, which is dominated by the background of the state-owned enterprises, and that the substantive breach of contract has not occurred in the enterprise debt surrender. The "implicit guarantee" of the government can not be separated, so investors are more willing to pay attention to the analysis of the background nature of the subject of debt issuance, so as to "guess" its credit risk, which is more practical in the domestic investment economic structure environment.
2, on the macro level of the credit spreads, the interest rate factor of structured variables is an important factor affecting the credit spreads of corporate bonds. No matter from the index level or the positive analysis of the voucher level, the risk free interest rate change and the change of the bond yield limit structure, expressed by the 10 year Treasury yield, are negatively related to the credit spreads. Two It is a good reflection of the changes in the economic environment, so that the investors expect the change in the profit of the enterprise to reflect the change of credit spreads. From the square of the 10 year return on the yield of national debt, the coefficient is more significant, that is, the change of the yield of the Treasury bond is nonlinear.
3, from the view of the macroeconomic price index, the overall view of the inflation ring ratio is positively related to the credit spreads of the rating enterprise debt, that is, the inflation circle is rising or the expected rise, the credit spreads of the corporate bonds tend to rise in each period. The positive correlation between the inflation index and the credit margin analysis from the level of the voucher level shows that the credit debt is inflationary. It is expected to be more sensitive than Treasury bonds.
4, the Shanghai and Shenzhen 300 index volatility and the Shanghai and Shenzhen 300 index yield are very different between the index level and the voucher level. From the index level, the two changes have little influence on the credit spreads. No matter from the term dimension or the rating dimension empirical analysis, the Shanghai, Shenzhen 300 index volatility and the Shanghai and Shenzhen 300 index return are both convergence. At the level of voucher, the Shanghai and Shenzhen 300 index volatility and the Shanghai and Shenzhen '300 index return are negatively related to the credit spreads, although the correlation coefficient is small. This paper holds that the voucher level contains more personalized information and more portrays the relationship between the two. The CSI 300 index reflects the change of the business cycle, at least the changing trend of the enterprise stock cycle.
5, liquidity has a negative impact on the credit spreads, that is, the stronger the liquidity, the narrower the credit spread will be, but the effect is not very obvious from the empirical analysis. This paper uses the liquidity change based on the trading frequency to depict the changes in the liquidity of the two level market of enterprise debt, showing the strong variety dependence and time change from the trend of its change. Time-varying characteristics. Since the interbank bond market, due to the "holding", "voucher" and other trading means, although the volume of turnover is magnified, it does not reflect the change of credit spreads caused by liquidity. Fundamentally, the institutional investors of the interbank debt in our country are still mainly commercial banks, insurance institutions and so on. As a result, the impact of liquidity on credit spreads is not particularly significant.
6, in this paper, the empirical model of the index level explains about 35% of the change of credit difference, and the voucher level panel data model can only explain about 20% of the change of credit difference, which is closely related to the current situation of bond market development between China's bond banks. The market of the inter-bank bond market in China is still a unilateral market, and the institutional investors are homogeneous and strict. There are more false transactions represented by "holding" and "passing vouchers". The model can not avoid this effect, and the corresponding explanatory power is weak. On the other hand, from the fitting degree of the empirical analysis results of the debt rating of enterprise debt, with the reduction of credit qualification, the release force of the model is enhanced, which is also in agreement with the foreign empirical results. Through the index, the index is also in accordance with the index. The overall comparison of the empirical results between the level and the voucher shows that the two are basically compatible. This aspect shows the robustness of the empirical results, and on the other hand it shows the effectiveness of the overall valuation data.
【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】