An Model Study On The Financial Early-warning Of Manufacturing Listed Companies In China

As an important industry of the national economy in our country,manufacturing industry has greatly promoted the development of our country's economy since the start of reform and opening policies.But under the background of economic globalization,the manufacturing listed companies in our country face enormous challenges.In recent years,more and more manufacturing listed companies fall into the financial crisis among the intensified competition,which would not only destroy the order of securities market,also damage the interests of investors.Any company's financial crisis is not happening overnight.It is usually because that managers do not pay enough attention to financial risk monitoring and the crisis symptoms,and fail to take timely measures,so that the financial situation becomes worse,even led to the crisis.No matter from which position,monitoring financial early warning has very important practical significance.Therefore,we think it is necessary to study the financial early warning model of manufacturing industry listed companies in our country.In this paper,based on 146 samples which contain 73 listed companies Special Treated with abnormal financial position in nearly two years and the matching 73 listed companies with normal financial position,we select 23 financial indicators as the independent variable X to build financial early warning models.The sample data selected is relatively new,so that the research has timeliness.In the process of building models,we select 50 companies as the training samples separately from the two types of listed companies randomly.First of all,the traditional Logistic model is established with all 23 financial indicators.The second,we reduce the dimension of the data by using principal component analysis(PCA)method and partial least squares(PLS)method,then we extract the principal component factors and PLS factors.Next,we build PCA-Logistic model and PLS-Logistic model by using the extracted factors instead of the original financial indicators to regress with the dependent variable Y.For component factor and PLS factor can respectively be expressed as a linear combination of the original financial indicators,models built in this paper are all explicit expressions which can directly reflect the relation between the dependent variable and the original indicators.So we can use test samples to verify the applicability of the models.Finally,this article use three models to forecast the 46 samples respectively.Through analyzing and comparing the predicted results,we argue that the three models can be used in financial early warning of our country manufacturing industry listed company.In terms of early warning effect,the predicted results of improved PCA-Logistic model and PLS-Logistic model are superior to the traditional Logistic model,and the PLS-Logistic model is the best of all.This article's results can provide some reference for manufacturing listed companies to improve financial early warning capability.