中国地方政府债务风险动态监测预警研究——基于马尔科夫区制转换回归模型

The Study of Monitoring and Early Warning of China’s Local Governments’ Debt Risk——Based on the Markov zone transformation regression model

新常态下地方经济中低速增长,房住不炒式的调控使土地拍卖收入锐减,这些都促使我们需要进一步做好地方政府债务风险动态预警管理。本文收集1998—2016年我国31个省市地方政府债务经济数据,用熵值TOPSIS法,综合评价地方政府债务风险,构建马尔科夫区制转换回归模型,估算地方政府债务风险的预期持续期、区制转换概率与稳态概率,研究地方政府债务风险区制转移并做出动态监测预警。

Under the new normal economic situation,the local economy has a low-speed growth.Besides,the regulation of non-speculation investment on house has drastically reduced the revenue of land auction of local governments.These have prompted us to explore more in early dynamic warning management about local governments’debt risk.Based on the data of 31 provinces in China from 1998 to 2016,this paper uses the entropy TOPSIS method to comprehensively evaluate the local governments’debt risk,and constructs the Markov regional conversion regression model to estimate the expected duration,Conversion probability and steady-state probability.At last,we make a research about the inflection point of local governments’debt risk and make dynamic monitoring and early warning.