中国公司债信用价差影响因素研究

发布时间：2018-03-22 12:17

本文选题：公司债　切入点：信用价差　出处：《华侨大学》2017年硕士论文　论文类型：学位论文

【摘要】：公司债因其低融资成本的优势已经成为直接融资的一个重要方式,它的风险中性以及收益稳定性备受机构投资者的青睐。而信用价差则是公司债的定价基础,能有效的衡量公司债的相对价值和风险收益水平。基于我国公司债发行实情,受其发行时间晚、早期发行数量少等因素的影响,国内关于公司债信用价差的研究相对较少,信用风险的度量和控制技术相对滞后。因此,在借鉴国外成熟债券市场研究成果的基础上,结合我国公司债市场的特点来研究公司债信用价差影响因素,对公司债的定价以及信用风险的控制至关重要。本文结合中国公司债券市场发展实际,从“信用价差之谜”出发,以信用价差分解理论为基础,从违约风险、流动性风险、宏观经济指标以及市场风险四方面来研究公司债券信用价差。通过层层筛选,最终从沪深交易所选取91支样本公司债券,截取2013年1月至2015年12月36个月度交易数据。在实证分析部分,为了研究违约风险、流动性风险、宏观经济指标以及市场风险对公司债券信用价差的影响,本文使用面板数据分析方法,经过两个步骤的模型筛选,最终选定固定效应模型来进行分析。同时对样本数据从行业、期限和信用等级三个维度进行分组来研究各因素对信用价差的影响程度。在对样本进行整体和分组面板数据固定效应模型分析后,利用固定效应变换法对各因素进行分位数回归分析,从各个分位点上更为细致深入的分析各因素对信用价差的影响过程。由实证分析结果可得到如下结论:(1)在固定效应模型分析中,假设1H、2H、4H、5H、6H、7H、8H均得到了实证数据结果支持,其中假设2H中杠杆比率与信用价差显著负相关,这与Merton模型预期不一样。假设3H中流动性指标换手率和零交易天数比率与信用价差显著正相关,这与预期假设相反,而非流动性指标Amihud与信用价差相关性不显著。(2)在样本分组回归结果中发现,不同行业、不同期限和不同信用等级信用价差的影响因素有差异。(3)在分位数回归分析中,发现有些因素在整体回归中与信用价差显著相关,但在分位数回归中这些因素并非在所有分位点上都与价差相关性显著。同时除无风险利率和杠杆比率外,其他变量与信用价差之间的关系与整体回归模型中一致,说明这些变量对公司债信用价差的影响是稳健的。
[Abstract]:Corporate debt has become an important way of direct financing because of its advantages of low financing cost. Its risk neutrality and income stability are favored by institutional investors. Credit spreads are the pricing basis of corporate bonds. The relative value and risk return level of corporate bonds can be effectively measured. Based on the fact of corporate bond issuance in China, due to the factors such as late issuance time and low amount of early issuance, the domestic research on credit spreads of corporate bonds is relatively rare. The measurement and control technology of credit risk is lagging behind. Therefore, based on the research results of foreign mature bond market and the characteristics of our country's corporate bond market, this paper studies the influencing factors of corporate bond credit spread. The pricing of corporate bonds and the control of credit risk are very important. This paper, based on the theory of credit spread decomposition, combines with the development of Chinese corporate bond market, starts from the "mystery of credit spread", from default risk, liquidity risk, etc. Macroeconomic indicators and market risks are four aspects to study the credit spreads of corporate bonds. Through layers of screening, 91 sample corporate bonds are selected from the Shanghai and Shenzhen Stock Exchange. From January 2013 to December 2015, 36 monthly trading data were intercepted. In the empirical analysis, the effects of default risk, liquidity risk, macroeconomic indicators and market risk on the credit spreads of corporate bonds were studied. This paper uses panel data analysis method, after two steps of model screening, finally select a fixed effect model to analyze. At the same time, the sample data from the industry, Three dimensions of duration and credit rating were grouped to study the impact of various factors on credit spreads. The quantile regression analysis of each factor was carried out by using the fixed effect transformation method. The effects of various factors on credit spreads are analyzed more carefully and deeply from each locus. From the empirical results, the following conclusions can be drawn: 1) in the fixed effect model analysis, we assume that 1H ~ (2) H ~ (2) H ~ (2) H ~ (4) H ~ (5) H ~ (5) H ~ (6) H ~ (6) H ~ (6) H ~ (7) H ~ (8) H is supported by the empirical data. It is assumed that the leverage ratio in 2H is significantly negatively correlated with the credit spread, which is different from the Merton model. Assuming that the liquidity index turnover rate and the zero trading days ratio in 3H are significantly positively correlated with the credit spread, this is contrary to the expected assumption. However, the correlation between Amihud and credit spread was not significant. (2) in the sample grouping regression results, there were differences in the influencing factors of credit spreads between different industries, different periods and different credit grades. 3) in the quantile regression analysis, there were significant differences in the credit spreads between different industries, different periods and different credit grades. It was found that some factors were significantly correlated with credit spreads in overall regression, but not at all loci in quantile regression. In addition to risk-free interest rates and leverage ratios, The relationship between other variables and credit spreads is consistent with the overall regression model, which shows that the influence of these variables on corporate bond credit spreads is robust.
【学位授予单位】：华侨大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：F832.51

【参考文献】