The Prediction Research Of Corporate Distress Of Chinese Listed Companies From A View Of Intellectual Capital

With the increasingly complex business environment, more and more corporate focus on predicting and preventing various corporate distresses, but because of the uncertainty of the environment, the prediction work has become complex and difficult. In addition, the traditional financial indicators were limited when reflecting the external threats and competitive pressures that the modern corporate have to face, therefore, we added intellectual capital, the core competitiveness of the modern corporate, to predict corporate distress, while the combination with financial indicators can complete the prediction information and promote the accuracy of prediction; In addition, the research above can propose kinds of useful ideas for making efficient use of and maximize the potential of the intellectual capital of listed companies in China.First, this dissertation divided intellectual capital into human capital, structural capital and relation capital, and analysis the impact of listed companies' intellectual capital to the corporate distress from a theory angle, and discussion the relationship between intellectual capital and corporate distress under various environmental factors. Our research confirmed that:there is a significant positive relationship among the intellectual capital and listed companies corporation distress and the environment factors have positive moderating role on the relationship between listed companies corporation distress and intellectual capital.Then, we divided the samples into the growth, maturity, or recession period by applying the life cycle theory, we optimized the indicators of intellectual capital through analysising of the relationship among intellectual capital indicators and listed companies corporation distress at different life cycle stages. The research results showed that the influence of intellectual capital on the listed companies corporation distress are dissimilar under the different life cycle stages, in the growth stage, the human capital and structural capital have a negative correlation with listed companies distress, while relation capital has no significant relationship with the listed companies distress, therefore, we add only the human capital and structural capital to the prediction system when predicting the listed companies at growth period; In the maturity, or recession period, human capital and relation capital have a negative correlation with listed companies distress, while structual capital has no significant relationship with the listed companies distress, therefore, we add only the human capital and relation capital to the prediction system when predicting the listed companies at maturity, or recession period.After that, we constructed a "strong coupling" Rough neural networks based on variable precision rough set (Various Precision Rough set Neural Networks, VPRNN) model, and applied this method to predict the distress of listed companies. This model using the variable precision rough set to lower the interference of "noise" data, and extract, reduce the original data effectively with no intereference of critical information, meanwhile, guided neural network designing and operation. Variable precision rough set neural network model's greatest strength is to reduce the network structure and improve the training speed, it can apply the appraisal rules which obtained from the small-scale data to the larger scale prediction research, and therefore, it has a better anti "interference" ability and generalization ability.Finally, we used the selected intellectual capital indicators and financial indicators with the VPRNN model to predict the listed companies'distress of different stages of life cycle. When predicting the listed companies at the growth, maturity and recession period of the test sample, the accuracy of prediction are 88.571%,87.143%,90.000%; when adding the indicators of intellectual capital, the accuracy of prediction are up to 97.143%, 95.714%,98.571%.