论文独创性声明
本人声明：本人所呈交的学位论文是在导师的指导下,独立进行研究工作所取得的成
果。除论文中已经注明引用的内容外，对论文的研究做出重要贡献的个人和集体，均已
在文中以明确方式标明。本论文中不包含任何未加明确注明的其他个人或集体已经公开
发表的成果。
本声明的法律责任由本人承担。
论文作者签名：
2022年6月12日
论文知识产权权属声明
本人在导师指导下所完成的论文及相关的职务作品，知识产权归属学校。学校享
有以任何方式发表、复制、公开阅览、借阅以及申请专利等权利。本人离校后发表或使
用学位论文或与该论文直接相关的学术论文或成果时，署名单位仍然为长安大学。
（涉密论文在解密后应遵守此规定）
论文作者签名：
导师签名：
2022年6月12日
2022年6月15日
摘
要
近年来，政府债务管控已成为重要话题，与国外的研究相比，由于我国特殊的国情和独特的
体制制度，中国与其他国家相比，目前对于政府债务的研究起步较晚，缺乏统一的共识，很多专
家在研究方法、指标选取等诸多问题上还没有达成一致。但随着经济社会发展，我国地方性债务
风险逐渐浮现，政府债务多次被写进重要文件，如何防范化解政府债务风险已成为当前的重要课
题。因此，关于政府债务风险的度量以及风险预警的研究无论对学界还是社会都具有相当重要的
作用。
本文通过梳理政府债务相关研究成果，结合相关欧亿·体育（中国）有限公司，首先概括总结了中国地方政府债务风
险发展、债务规模和债务结构等，其次综合我国实际情况，将地方政府债务按照明确的地方债务
风险—显性债务和隐藏的地方债务风险—隐性债务进行分类，分别对债务形成机理进行深入分析。
本文利用向量自回归 VAR模型基于 1994-2021年 31个省份的地方政府财政收入对 31个省份 2022
年的地方财政收入进行预测，之后通过 KMV模型计算出违约率。同时，本文使用熵值法来确定
预警模型中每个指标的权重，并基于 TOPSIS法编制了风险评估指标，成功构建地方政府债务风
险预警模型，评估各省地方政府债务风险，最后从显性债务和隐性债务两个维度提出地方政府债
务风险管理对策。
本文主要获得两点实证发现，一是我国各省地方政府债务整体状况不容乐观；二是我国地方
政府债务的风险存在区域差异性，风险高的多为西部省份。
关键词：政府债务，债务风险，KMV模型，预警体系
I
Abstract
In recent years, government debt control has become an important topic. Compared with foreign
research, due to Chinese special national conditions and unique system, compared with other countries,
China's current research on government debt has started relatively late and lacks a unified consensus.
Many experts have not reached an agreement on research methods, index selection and many other issues.
However, with the development of economy and society, debt risks of Chinese local government
gradually emerged, and government debts have been written into important documents many times. How
to prevent and resolve debt risks of Chinese local government has become an important issue at present.
Therefore, research on the measurement of government debt risk and risk early warning plays a
significantly important role in both academia and society.
By sorting out the relevant research results of government debt and combining relevant data, this
paper first summarizes the development, debt scale and debt structure of debt risks of Chinese local
government. Debt and hidden local debt risk-hidden debt are classified, and in-depth analysis of the debt
formation mechanism is carried out respectively. Based on the VAR Model, this paper predicts the local
government revenue of 31 provinces in 2022 based on the local government revenue of 31 provinces
from 1994 to 2021, and then calculates the default rate through the KMV Model. At the same time, this
paper uses the entropy method to determine the weight of each indicator in the early warning model, and
compiles risk assessment indicators based on the TOPSIS method, and successfully builds a local
government debt risk early warning model to assess the local government debt risk in each province.
From the two dimensions of implicit debt, it proposes local government debt riskmanagement
countermeasures.
This paper mainly obtains two empirical findings. First, the overall situation of local government
debt in Chinese provinces is not optimistic; second, there are regional differences in the risk of debt risks
of Chinese local government, and the highest risks are mostly in the western provinces.
Key words: Government debt, Debt risk, the KMV Model, Early warning system
II