The Early Warning Research On The Financial Crisis Of The Real Estate Enterprises

For the real estate enterprises and its stakeholders, it is critical important that seekthe appropriate financial crisis warning analysis model and method for China’s realestate enterprises, as well as precisely predict the proper financial difficulties faced bythe real estate enterprises which in order to take effective measures as early as possibleto prevent and dissolve the financial crisis for the real estate enterprises. In the processof the operation for real estate enterprises, it is utmost important for consolidatingenterprise competitive position and developing stably and healthily that how tooptimize the enterprise’s financial structure, control risk scientifically and reasonablyand solve financial crisis. Therefore, effective financial crisis warning is the inevitablechoice of system innovation for enterprises’ financial management, especially underthe background of more severe macro control policies for real estate market made bycurrent state, the survival and development of enterprises and maintaining the rightsand interests of related enterprises’ beneficiaries have extreme theoretical and socialpractical significance.Based on the characteristics of the real estate industry and developed existingcondition analysis of the real estate enterprises, this paper constructed a targetedfinancial crisis warning index system of real estate enterprises and used BP neuralnetwork method to establish financial crisis warning model for real estate enterprisesso as to provide decision-making method and reference frame for the managers andother stakeholders of the real estate enterprises. The main aspects are demonstrated asfollowed:First of all, based on the characteristics of the real estate industry anddeveloped existing condition of the real estate enterprises, this paper analyzed thevarious inducements of the real estate enterprises financial crisis. The real estateindustry has many kinds of characteristic, such as guided, basic, comprehensive, relevant, capital-intensive, risky, spreads, regional, multifactor of real estate pricedecision-making and influenced by national macroeconomic policy. In the period offast growth of real estate enterprises, there are also many financial problems exposed.For instance, inadequacy of the enterprises’ own funds, unreasonable capital structure,high asset-liability ratio, overuse of financial leverage; single financing method,significant proportion of bank loans and high current debt ratio; imbalance ofenterprise cash inflow term, common phenomenon of short loan long investment;keen to land hoarding, serious fund precipitation; blind diversified developmentwhich can erode main operating revenue. Secondly, it established the warning indexsystem of financial crisis for real estate enterprises. In order to fully embody thethoughts of characteristics for real estate enterprises, based on the cause analysis offinancial crisis, from the two aspects of financial index and non financial index, wemade preliminary selection of62index among8kinds, and initially establishedfinancial crisis warning index system of real estate enterprises. For the sake ofreducing the correlation between the variables, decreasing variables, screening andoptimizing the index system through the Kruskal-walis H significant test and principalcomponent analysis, we get8public factors which determined the modelingparameters and the input variables of the model.Thirdly, it established the dynamic BPneural network model of financial crisis warning for the real estate enterprises. Thispaper designed and finally set up a three-layer neural network with input node8times,output node3times. By comparing the training effect ideal choice of trainim methodas training function so as to construct neural network model, through the training andtesting, we get the financial crisis warning model of real estate enterprises. Thecomprehensive accuracy rate of testing results of the testing sample tested by themodel is92.38%, and the ratio of misjudged financial crisis enterprises to healthyenterprises is zero, thereby demonstrating there is a better recognition ability of thewarning model for the financial crisis. To apply such model into true real estateenterprises, the results showed that the distinguish accuracy rate is comparatively highand using effect is better, which are able to provide a kind of effective method for realestate enterprises financial crisis prediction.