房地产行业债券信用利差影响因素分析

发布时间：2018-03-25 10:56

本文选题：信用利差　切入点：结构化模型　出处：《上海交通大学》2013年硕士论文

【摘要】：本文以116只房地产行业信用债券为样本，参考国外成熟模型，采用横截面数据和时间序列数据回归的方法，对结构化模型变量和其他变量对债券信用利差的影响进行了实证研究。横截面数据的回归结果表明，，债项评级、债券提前偿还条款和地方政府财政实力对房地产行业，尤其是城投类债券的信用利差有显著的影响，模型的解释力达到了49.36%，各个变量均显著。我们还对公司层面的财务数据行了研究，以公司杠杆比率、速动比率、和是否设置担保作为变量进行了回归，结果并不显著。这表明债券投资者对地方政府的财政实力更为关注，而对公司层面财务数据关注较少，主要是公司财务数据的披露较为滞后所导致的。时间序列日频率数据的回归结果表明，无风险利率和收益率曲线结构的变化对信用利差有显著的影响，股票市场的收益率和波动率变化并没有显著影响，在大多数发行人均为非上市公司的债券市场，股市的波动对债市的影响并不明显。模型整体的解释力达到了14.62%，与中外文献的结果类似，说明结构化变量对中国债券市场的信用利差有一定的解释力，但是并没有从根本上找到信用利差的主要影响因素。时间序列月频率数据由于数据点过少，模型整体的解释力受到影响和削弱，总的来说宏观变量的变化对信用利差的影响与我们预期的结果相反，可能是受到近期金融危机和信用债违约事件冲击的影响。
[Abstract]:In this paper, 116 real estate industry credit bonds are used as samples, referring to foreign mature models, and the method of cross section data and time series data regression is used in this paper. The effects of structural model variables and other variables on bond credit spreads are studied empirically. The regression results of cross-sectional data show that debt rating, bond repayment terms and local government financial strength have a significant impact on the real estate industry, especially the city investment bond credit spreads. The explanatory power of the model reached 49.36%, and the variables were significant. We also studied the financial data at the corporate level, using the corporate leverage ratio, the speed ratio, and whether or not to set guarantees as variables. The results are not significant. This suggests that bond investors are more concerned about the fiscal strength of local governments than about corporate financial data, mainly due to the lag in the disclosure of corporate financial data. The regression results of daily frequency data of time series show that the structure of risk-free interest rate and yield curve has a significant impact on credit spreads, but the change of return and volatility in stock market has no significant effect. In the bond market where most issuers are non-listed companies, the impact of stock market fluctuations on the bond market is not obvious. The overall explanatory power of the model has reached 14.62, which is similar to the results of Chinese and foreign literature. It shows that the structured variable has a certain explanatory power on the credit spread of China's bond market, but it does not fundamentally find the main influencing factors of the credit spread. Because the monthly frequency data of time series are too few, the overall explanatory power of the model is affected and weakened. In general, the effect of the change of macro variables on credit spreads is contrary to the expected results. May be affected by the recent financial crisis and credit default events impact.
【学位授予单位】：上海交通大学
【学位级别】：硕士
【学位授予年份】：2013
【分类号】：F299.23;F832.51;F224

【参考文献】