The Empirical Analysis Of The Urban Construction Bonds’ Issuance Spread

Since the tax system reform in 1994, the growth of local economy and financial revenue has been largely promoted. But at the same time, it also causes a new phenomenon that the local governments bear more responsibility but have less financial power which results in the local governments’ increasing funding gaps. Local governments can not issue government debts to raise money because of regulatory constraints, so they start to construct financial platforms (Urban Construction Investment Company. "UCIC") for financing. Since the first urban construction bond (UCB) was issued by Shanghai Urban Construction Investment and Development Corporation in 1993. the UCBs have finally embarked on the stage of history and aroused great interests of government, enterprises as well as investors after two decades" development. However, the explosive growth of the UCBs market also brought tough issues in the past several years, such as different aptitude urban construction companies, low profit, unreliable financial data and government’ various kinds of guarantee. People are more and more concerned about UCBs’ credit risk and therefore, the study of UCBs’ credit spread determinants is of great significance at this point.UCBs are seldom researched by domestic scholars because of their relative short history. Most relative journals are simple empirical analysis with small size of sample, usually of about 100 bonds, which is not very convincing. While in this paper,I use multiple linear regression models with 1402 corporate bonds issued by the UCICs as sample to study their primary market credit spread. The data range from i Jan.2006 to 31 Dec.2013 and the influencing factors are divided into non-financial factors as well as financial factors, which are further processed using the Principle Component Analysis. Then I study the relationship between the two kinds of factors and credit spread respectively before regressing them together in one model. After the multiple linear regression with all factors, I get an adjusted R-square of 0.74, which indicates that the whole model is quite convincing.After careful analysis, I mainly get the following empirical results:the non-financial indicators are mostly significant in deciding the credit spread, including the macro-indicators such as GDP growth, CPI growth, M2 growth and debt market index; the local indicators such as local government deficit ratio and administrative level of the issuer; the bond-related indicators such as credit rating and issue amount of the bond. Also, they have a great explanation power as a group with the adjusted R-square reaching 0.73. However, only two of the six financial principle components are significant in the whole factor model and when we add the financial principle components into the non-financial factors model, the adjusted R-square only increases by 0.01, which shows the financial factors’poor explanation power.