Financial Distress Of Chinese Listed Real Estate Companies Based On Bp Neural Network Research

After several decades development, the research of company's financial distress prediction has become a widespreadly concerned issue by scholars both at home and abroad. Not only has high academic value, but also it has great value to provide useful information for enterprise managers and stakeholders to do some decision-makings. Retrospect to the research about company's financial distress prediction at home and abroad, it mainly concentrated in two aspects, including the selection of explanatory variables and prediction modles' construction. However the construction of explanatory variables of financial distress, as well as prediction models, is much more universal in nature for all the companies in all industries. While a prediction model for a special industry has become more and more important than before.Up to now, after a period of rapid development, China's real estate industry has become one of the pillar industries in the national economy. Along with the rapid development of China's securities market, the number of listed real estate companies and scale are constantly expanding, and a growing number of real estate investors concerned about the performance and development of listed real estate companies. Thus it's very important to construct financial distress prediction model for listed real estate companies.Through the retrospect of the relevant literatures of company's financial distress prediction, as well as the characteristics of China's real estate industry, the paper proposals an approach, including the definition of financial distress and the states of the financial distress, the selection of samples and explanatory variables, as well as the construction of financial distress prediction model aimed at listed Chinese real estate company, with its prediction horizon: one year.In the third part, this paper gives an explanatory variable system targeted in listed Chinese real estate companies. On one hand, there are 44 listed real estate companies was selected as samples, with its 1998-2005 financial reports used to predit 1999-2006 financial situation. On the other hand it is based on a wide reading of literatures, there are 63 explanatory variables of 8 aspects are selected to fully reflect all aspects of a real estate company, including the solvency, profitability, business development capability, asset management capabilities, shareholders profitability, capital market information, as well as corporate governance.In the part of preliminary data analysis, Kruskal-walis H testing is used to test whether its values for each group of firms come from three ranked populations. And then the factor analysis is used to optimize the explanatory variable system for the financial distress prediction of Chinese listed real estate company.Then this paper gives three-state financial distress prediction model of China-listed real estate companies. Based on the explanatory variable system, the forth part of this paper gives a BP neural network method with prediction horizon of one year. This model is constructed with 212 original samples and then tested with 105 holdout samples. Test results show: the model has the correct rate of 92.38%. What's more, this model is very strong in telling the financial distress from financial stability. That's to say the model can avoid misjudgment of the financial crisis as financial stability.Finally, using the three-state financial distress prediction model of China-listed real estate companies built above, the 1998-2005 financial data of "Haitai development" is simulated to forecast the financial situations of "Haitai development" in 1999-2006. The result indicates that the forecasts for financial-distress year are entirely correct, only one financial-stability year is misjudged, proving the greater value of this three-state financial distress prediction model of China-listed real estate companies...