Empirical Research On The Prediction Model Of Financial Distress Based On Our Country's Construction Industry

In the increasingly fierce competition of the construction market, financial risk exists everywhere. As the construction industry is seriously influenced by politics and economy, it could easily fall into financial distress because of its big investment scale. So it is very important to build a prediction model of financial distress. In order to reduce the risk to the minimum, enterprise managers and staff members of finance departments should pay attentions to risk forecasting by establishing an effective prediction model of financial distress. It is useful to raise the alarm to the managers and stakeholders timely in the incubation period of risk. So business managers and staff members of finance departments can identify vulnerabilities and take remedial measures to improve the business operations and financial condition. Therefore, it has practical significances for companies of construction industry to build a prediction model of financial distress.Non-financial variables and financial variables can not only complement each other by making up for the weakness of financial variables, but also provide more important information for corporate to forecast financial crisis. Therefore, in order to prevent the financial crisis effectively, this paper tried to introduce non-financial indicators and built a prediction model of financial distress, combining with the characteristics of the construction industry. This paper used both the theoretical analysis and empirical research. It introduced the research status of financial risk warning at home and abroad in theory. In addition, it did research from the empirical aspects. This article firstly introduced the research background, purpose and significance, then did empirical research. At last, it drew conclusions and recommendations. The main contents of empirical research were as follows: this paper selected listed companies of construction industry as the research samples based on datas from the year of 2011 to 2014.From the point of rigor, it excluded B-share listed companies which lack datas, and sample companies that have listed under 5 years and the companies that had been delisted. At last, it took 60 listed companies of construction as the research objects and selected financial crisis enterprise samples by using cluster analysis method. In order to conduct this research, the model of financial distress regards whether one company takes financial risk or not as the dependent variable, regards a number of financial and non-financial indicators as independent variables in empirical research. Firstly, it analyzed the differences between 28 financial indicators and 16 non-financial indicators from the profitability, solvency, developing capability and governance capabilities. Secondly, in order to formulate logistic regression model, it tested all indicators by T-test, U-test and collinearity diagnostics, selecting indicators that have significant differences. Then we constructed logistic regression models and did back inspection.At last, it found that non-financial indicators have a certain effect on forecasting, and the predicting accuracy of integrated model is higher than the purely financial model. All of these show that model which adds non-financial ratios has a better ability to predict.