地方政府债务风险预警系统设计与模拟分析—以S省地方政府为模拟对象

Design and Simulated Analysis of Local Government Debt Risk Warning System—Example for S Province

摘要: 对预警系统流程进行整体设计，并从内部风险、相关风险和外部风险三方面构建指标集出发，构建基于聚类分析和因子分析的地方政府债务风险预警系统。以S省为例，选取该省2007年至2012年的反映地方政府债务风险的108个指标数据，运用多元统计中的聚类分析和因子分析进行实证分析。结果发现，S省现阶段地方政府债务风险处于中度风险状态，且该省地方政府债务风险呈逐年上升趋势；从提取的某个主因子的解释看出，控制S省政府债务风险的根本出发点和落脚点是降低债务总量和或有债务占比，从而降低结构风险和整体风险；另外S省的模拟也证明因子分析法可以很好地对地方政府债务风险预警系统进行分析，达到了预警的初衷。

Abstract: Designing overall process for warning system and starting from three aspects of internal risk, ex-ternal risk and related risk to build the index set, local government debt risk warning system was built based on cluster analysis and factor analysis. Taking S Province as an example, 108 metric data which reflect the risk of this province’s local government debt between 2007 and 2012 were selected to do empirical analysis using cluster analysis and factor analysis of multivariate statistic. It was found that S Province’s local government debt risk is in a moderate status at this stage and the province’s local government debt risk is increasing yearly; from the explanation of some ex-tracted primary factor, we can find that the fundamental starting point and end point to control S Province’s government debt risk are to reduce the total amount of debt and the percent of contin-gent liabilities, which can reduce structural risk and overall risk; furthermore, the simulation of S Province can also prove that factor analysis can analyze local government debt risk warning system well, achieving the goal of early warning.