| Moderate financialization is conducive to alleviating the financing difficulties of the real estate industry and accelerating industrial expansion.However,excessive accumulation of financial resources will make the level of real estate financialization too high,aggravate real estate financial risks and even endanger the financial system.Judging and controlling the development status of real estate financialization has become the key to the healthy development of the real estate industry.The level of real estate financialization is not clear,and policy regulation is difficult to " suit the remedy to the case," which can easily make the real estate financialization fall into two extreme levels of " too low " and " too high," which is not conducive to the healthy and sustainable development of the real estate industry.Based on this,this thesis attempts to build a comprehensive evaluation model of the level of real estate financialization,in order to accurately evaluate and judge the level of real estate financialization,so that the level of real estate financialization remains within a healthy and reasonable range.Firstly,on the basis of analyzing the mechanism of real estate financialization,this thesis constructs a comprehensive evaluation model of real estate financialization based on neural network.The model construction steps are as follows : 1)Screening of evaluation indicators.Combined with the existing index system and expert scoring,the evaluation indexes of real estate financialization are compared,analyzed and screened.From the five dimensions of real estate market,real estate enterprises,residential sector,financial institutions and government departments,10 important evaluation indexes are designed,such as the proportion of real estate investment,the growth rate of housing price,the proportion of real estate loans and the asset-liability ratio of enterprises.2)Establishment of evaluation criteria.Based on the 3sigma principle,the five-level evaluation standard of real estate financialization is established.When the evaluation results are 0.6,0.8,1,1.2 and 1.4,the corresponding real estate financialization level is too low,low,normal,high and too high.3)Training neural network.The neural network structure is determined and trained.The root mean square error and other indicators are used to evaluate the network performance.After the evaluation is qualified,the neural network for the evaluation of the real estate financialization level is obtained.4)MIV analysis is used to analyze the variables,and the evaluation indexes which have great influence on the comprehensive evaluation results of real estate are selected.After that,an empirical study was conducted on the level of real estate financialization in Wuhan.The results showed that the comprehensive evaluation values of the level of real estate financialization in Wuhan from 2016 to 2020 were 0.9894,1.0966,1.2029,1.0480,and 1.0984,respectively.The indicators that have a greater impact on the comprehensive evaluation results of real estate in Wuhan are the proportion of real estate investment,the ratio of house price to income,the proportion of real estate loans,and the proportion of state-owned land use right transfer fees.Through this study,we hope to provide some reference for the evaluation of real estate financialization in large and medium-sized cities in China,help government departments to control the development status of urban real estate financialization,and promote the healthy and sustainable development of urban real estate industry. |
