The Research Of Local Government Debt In Measurement Index System And Risk Early-warning

Local government debt plays a decisive role in promoting regional economic development. But, with the expansion of debt, the hidden danger in local fiscal risk gradually growing, and has become a threat to the local economy and social stability. We should have correctly understanding of the relation between local government debt and local economic development, and establish standards of local government debt management system and scientific effectively debt risk early-warning system. In the same time, based on local economic development level of science and debt bearing threshold, we should optimize the local government debt management, guarantee the economic security and social stability, and the great significance is promoting social and economic sustainable development.In recent years, the local debt is rising rapidly, but, as the lack of solvency and fund application is inappropriate, some local governments have to excessive increasing debt, and thus form a vicious cycle of debt. Many regional government debt deficit has the serious influence to the development of local economy. So, how to determine the appropriate level of debt to promote the sustained and rapid development of the economy? Which measure-model is appropriate to determine the quantitative relationship between them to achieve mutual promotion, complementary effect? The problem is urgent need to solve.The research object is local debt, and the study focus on local government debt paying ability, capital regeneration ability, the stock of debt, the using of debt money and local GDP growth and so on. We establish a full diameter index evaluation system and using the combination of qualitative analysis and quantitative analysis method to analyze the debt risk situation. We take the concept of debt using in to the index system, and make deep study in the using of debt money, making up for the deficiency of the previous research.Using clustering analysis on the index, and through clustering analysis, we eliminate duplicate index and get effective index system, to achieve the principle of independence and the least. Then, we use factor analysis to get three main factors, and use the three factors to evaluation local government debt position in the past years, so we can evaluation the BP system as well.At the same time, the research introduces the BP neural network which uses the result of factor analysis as the training set to early-warning on local government debt. BP neural network is a reverse propagation model and it has multiple input and multiple output which is suitable for the complex influence factors object analysis and forecast. The input set of BP neural network is divided into training set and test set The training set, in the study, is the results of factor analysis, and we make the latest issue of data as the test set. Through the processing of early warning model, we can further determine the risk interval of local government debt. The paper puts forward the targeted policy recommendations.The research use Shandong province as a case study analysis and forecast early-warning in 2016 local government debt risk at moderate risk range. Through the whole research, it is not hard to find that local government debt risk is mainly insufficient solvency, excessive stock of debt and unreasonable debt using. The central government and local governments need to strengthen the regulation and risk assessment in raising capital debt and debt fund using aspects.