A Theoretical And Empirical Study On Financial Crisis Prediction For Listed Real Estate Companies

In view of the interaction of many factors which result the financial crises of enterprise, to establish a simple but efficient evaluation model involving multi-financial factors is not only significant to the control and defense of financial risks, but is also useful to the listed companies in their decision support.In this paper, we pay more attention to the specific circumstances of real estate companies in China. Based on financial indicators and corporate governance, two Logistic pre-warning models M1 and M2 for financial distress are constructed, in which the SPSS 17.0 is applied and the estimate data is chosen from 117 companies: using respectively their sample data of 2007 and the weighted average of that from 2007 to 2009. We have also tested our models by using the latest data from Annual Report of 2010, which shows us good prediction. We construct our model M1 in detail, but for the M2, we only sketch the main process, for the similarities in their constructions.Our paper consists of five chapters, and we summary that in the following.Firstly, we describe the background and significance of pre-warning for financial distress of real estate listed companies. After a brief review of the achievements already known in this field, we introduce the main contents and methods appearing in our paper.In the second chapter, we define financial crises conceptually and focus on two sides of financial distress of real estate listed companies:the causes and presentations.The third chapter is devoted to project design, in which we introduce the principle we choose samples and data, and that for the selection of indicators of pre-warning. We have tested the K-S normal distribution of indicators'data, applying SPSS 17.0. Furthermore, we have done the independent T data test for the indicators satisfying normal distribution, but for the other indicators non-parameter Mann-Whitney test is used. After that, we select principal components from significant financial indicators.Basing on chapter three, we establish the pre-warning Logistic models for financial crises in the fourth chapter. The prediction efficiency of the models is then verified, using the sample data of 2010, and an analysis of the result is contained.We come to a summary of our results in the last chapter, and we end our paper by some remarks on the relative parts that should be improved in near future.It's shown in the project design that the following 10 indicators turn out to be more important and stable for their contributions to differ the companies in financial distress:operating margin, assets ratio of retained earnings, liqudity ratio, cash ratio, total assets ratio of working capital, debt ratio, inventory turnover rate, Ln(total assets), net assets growth, operating leverage, audit opinions from non-financial indicators.Besides, the conclusion from model M1 tells that the principal components F1, F2, Fs and the corporate governance variable K12 are negatively related to the probability of companies'financial distress, where F1 is mainly responsible for profitability and solvency, F2 mainly for solvency and operation capability, F8 mainly for size and growth capacity, and where K12 tells that whether the companies are issued by unqualified opinion. From that, we can conclude:it is less likely that one company would encounter financial distress when it owns stronger abilities mentioned above. We could see this even more clearly, considering the relation between the company's issue by unqualified opinion and its possibility with financial trouble. This is basically compatible with our anticipation.After a comparison of the two models M1 and M2, it is clear that they present similar in prediction. However, M2 shows better in over-all prediction accuracy and in recognition of the well-done enterprise. In contrast, M1 is better at identifying those companies in financial trouble.We finally mention some specialties and innovations of our paper to end the abstract.(1) The usual one-by-one pair principle according to firm size and industry is not used in our selection of sample data. Instead, we take all the real estate listed companies as the potential choices, from which we finally select 117 companies, in accordance with the listing standards of life and main business. As a result, the ratio of companies in financial trouble to that in good situation is closer to the ratio in real, compared with that in other similar studys.(2) We construct one pre-warning model based on the data of last three years, but in fact we use its weighted average form, which gives us ideal result.(3) The Cut Value is usually chosen to be 0.5. However, we prefer a smaller one 0.4 in this paper, for the purpose of improving the efficiency of our model to identify companies with financial distress.