|
Since the beginning of the 21st century, with the globalization of the world economy, the listed companies as the main body of the capital market opportunities while also facing a variety of difficulties. In 2008, the global financial crisis triggered by the U.S. subprime mortgage crisis, to the people alert. Financial difficulties of the enterprise, not only to make themselves into a debt crisis, and more like a food chain of the ecosystem of the implicated numerous stakeholders and even a devastating impact. Up to 155 by the end of 2009, cumulative A shares in Shanghai and Shenzhen, China motherboard market listed companies special treatment, that China has a considerable part of the listed company is in financial distress, not only to bring their own operating pressure and enable investors to creditors and other interests are threatened. Therefore, the financial plight has become a significant problem constraints listed companies as well as capital market development to accurately predict the results favor the listed company in a timely manner to prevent and resolve financial difficulties, improve management of science, contribute to the healthy and stable operation of the company operating. , Although more financial distress prediction model, but there are some problems, mainly in the model to predict the reliability is not high, the applicability is not strong. The paper proposes using the the gray combination method (gray autocorrelation, gray cross-correlation and the gray clustering) screening a valid indicator variables into the prediction model; gray logistic regression model to predict the samples of the company's financial predicament. The article mainly include the following: (1) draw on existing research results at home and abroad and defined the concept of financial distress, inductive analysis of the theoretical basis of the financial distress prediction overview of the current domestic and international representative Corporate Financial Distress quantitative prediction model. ② advantages of using traditional logistic regression prediction model, as well as the gray modeling method to establish the small sample size, data exist under the conditions of certain grayscale gray logistic regression model and using the model to predict the financial distress of the listed companies . ③ article sample by company ST and non-ST (paired proportion of 1:3) composed of a total of 120 samples (64 in 2009, 56 in 2010). Years and by ST is defined as t, use year t previous 5 years (t-1, t-2, t-3, t-4, t-5), the sample data of the indicator variable filters; using t years ago 4 years (t-2, t-3, t-4, t-5) data for samples to build prediction models. ④ empirical process, use gray combination were screened on the the 9 categories system 54 indicator variables into the forecasting model variables. Using traditional logistic regression model and gray logistic regression model to forecast the 2009 sample companies in 2010, the probability of financial distress, and comparative analysis of the two prediction models sentenced positive rate. Traditional logistic regression model, the highest cumulative sentenced positive rate of 85.9%, the company will be the maximum penalty of ST's is 62.5%. The being the minimum cumulative sentenced gray logistic regression model to predict the rate was 89.3%, the company will be the minimum sentence of ST is 71.4% minimum goodness of fit of the model was 0.899.
|