高新技术制造企业的动态财务危机预警研究

发布时间：2018-12-16 00:12

【摘要】：高新技术制造业是制造业的主力军，具有高技术含量及高附加值两个特点，对高新技术制造企业进行财务预警研究，能够帮助高新技术制造企业避免财务危机，保障高新技术企业的良性经营，进而促进国民经济发展。本文回顾和总结了国内外财务预警研究文献，对比分析各类研究的优缺点，认为基于人工智能的组合模型是现代财务预警研究的有效方法。在界定财务危机概念时，结合我国上市公司实际情况，将公司被ST(特别处理)作为陷入财务危机的标志。以高新技术制造业上市公司为研究对象，根据预警指标体系的构建原则和高新技术制造企业的行业特点，构建了适用于高新技术制造企业的财务预警指标体系。由于企业财务危机的出现是一个连续的动态发展过程，本文从短期与长期两个角度出发对高新技术制造业上市公司进行财务危机预警研究。其中，短期动态预警是以季度为单位，将上市公司的季度财务面板数据引入综合灰色预测GM(1,1)和BP神经网络的动态模型中来判断公司财务状况，并以“思达高科”上市公司作为实例进行模型的应用，结果表明基于灰色-BP神经网络模型能有效反映公司财务状况的发展趋势，时效性较强；长期动态预警是以年度为单位，将发生危机前两年和前三年（T-2期和T-3期）的财务面板数据引入基于Logistic-BP神经网络模型中进行动态预警，将预测结果与一般Logistic回归分析和BP神经网络模型比较，证明Logistic-BP神经网络预警模型更能体现财务危机的发生机理，并以“新华制药”上市公司作为实例进行模型的应用，结果证明了模型的有效性，体现了模型较高的预警精度。本文的主要研究结论如下：一、根据不同行业的特点选取适当的财务预警指标，并对预警指标进行筛选和精简是建立有效预警模型的前提；二、本文针对高新技术制造业上市公司构建的短期财务预警模型和长期财务预警模型，均具有较高的预警精度，企业可以根据相应指标的变化及时了解财务状况，，做出合理的判断，最终通过理性决策来避免财务危机；三、通过对财务预警指标的时序数据进行分析，将短期与长期结合、静态与动态结合构建的财务预警模型可以充分挖掘企业财务信息，及时有效的反映财务状况的发展趋势，实现企业财务危机动态预警；四、合理集成各个单一预测方法的混合分析模型能够发挥各个方法的优势，提高模型的泛化能力，是未来创新研究的趋势。
[Abstract]:The high-tech manufacturing industry is the main force of the manufacturing industry, with the characteristics of high technology content and high added value. To study the financial early-warning of high-tech manufacturing enterprises can help high-tech manufacturing enterprises to avoid financial crisis. Safeguard the benign management of high-tech enterprises, and then promote the development of the national economy. This paper reviews and summarizes the domestic and foreign financial early warning research literature, compares and analyzes the advantages and disadvantages of all kinds of research, and thinks that the combination model based on artificial intelligence is an effective method of modern financial early warning research. When defining the concept of financial crisis, combined with the actual situation of listed companies in China, the company is regarded as the sign of financial crisis by ST (special treatment). Taking the listed high-tech manufacturing companies as the research object, according to the construction principle of early-warning index system and the industry characteristics of high-tech manufacturing enterprises, the financial early-warning index system suitable for high-tech manufacturing enterprises is constructed. Because the emergence of enterprise financial crisis is a continuous dynamic development process, this paper carries on the financial crisis early warning research to the high-tech manufacturing industry listed company from the short-term and the long-term angle. Among them, the short-term dynamic early warning is based on the quarterly financial panel data of the listed company, which is introduced into the dynamic model of comprehensive grey forecast GM (1Q1) and BP neural network to judge the financial situation of the company. The application of the model based on grey BP neural network model shows that the model can reflect the development trend of the company's financial situation effectively and has strong timeliness. Long-term dynamic early warning is based on the Logistic-BP neural network model, which introduces the financial panel data of the first two years and the first three years (T-2 and T-3) into the dynamic early warning system based on the Logistic-BP neural network model. Comparing the prediction results with general Logistic regression analysis and BP neural network model, it is proved that the early warning model of Logistic-BP neural network can better reflect the occurrence mechanism of financial crisis, and the application of the model is carried out with the listed company of Xinhua Pharmaceutical Company as an example. The results show the validity of the model and the high warning accuracy of the model. The main conclusions of this paper are as follows: first, selecting appropriate financial early-warning indicators according to the characteristics of different industries, and screening and streamlining the early-warning indicators is the premise of establishing an effective early-warning model; Second, the short-term financial early-warning model and the long-term financial early-warning model constructed by listed companies in high-tech manufacturing industries have high warning accuracy. Enterprises can understand the financial situation in time according to the changes of corresponding indicators. Make reasonable judgment and finally avoid financial crisis through rational decision; Thirdly, through the analysis of the time series data of the financial early-warning index, the financial early-warning model, which combines the short-term and long-term, static and dynamic combination, can fully excavate the financial information of the enterprise. Timely and effectively reflect the development trend of financial situation and realize the dynamic early warning of enterprise financial crisis; Fourthly, it is the trend of innovation research in the future that the hybrid analysis model with reasonable integration of each single prediction method can give play to the advantages of each method and improve the generalization ability of the model.
【学位授予单位】：江南大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：F406.7;F276.44

【参考文献】