Risk Analysis of Infrastructure Asset-backed Securitization under Big Data Environment

Rating downgrade is an important signal of default, so this paper takes rating downgrade as an important basis for assessing the risk of asset-backed securitization (ABS) products. After analyzing the current situation of China's ABS market, we found that the rating downgrade concentrates in infrastructure enterprise asset-backed securitization (EABS) products. Therefore, this paper provides further risk analysis on infrastructure ABS products with big data mining and analysis technology. Firstly, we defined 9 independent variables. The dependent variable is a binary variable, which is recorded as 1 (Y=1) for downgrade_yes and 0 (Y=0) for downgrade_no. Secondly, we collected sample data from CNABS website and carry out necessary pre-process. Finally, we performed cross-tabulation analysis on the sample data and analyzed each independent variable's influence on rating downgrade with information gain (IG) method. It is found that spread, original-rating, year, region, and amount have a relatively large impact on rating downgrade, while industry, method, maturity and frequency have a smaller impact on rating downgrade. Based on the results, suggestions are put forward to effectively reduce rating downgrades. The issuer should optimize the securitization design, the investors should choose the appropriate products based on their preferences, and the regulators should strength supervision over rating downgraded products and reduce the default risk, which will help promote the sustainable development of infrastructure ABS. This study can also provide a reference for the promotion of PPP project asset-backed securitization (PPP-ABS) in the infrastructure field.