我国制造业上市公司信用风险的测度研究

发布时间：2018-04-05 14:20

本文选题：信用风险测度　切入点：判别分忻　出处：《东华大学》2013年硕士论文

【摘要】：自2008年国际金融危机爆发以来,世界经济经历了急剧地动荡与衰退地考验,现在又进入到艰难复苏的后国际金融危机时期。受经济周期下行和国家产业政策调整等因素的影响,国内制造业、地产、船舶、钢铁贸易等被列入了高风险行业。其中,制造业作为国民经济的基础,是经济指数良好运行强有力地支撑。然而,由于我国近期宏观经济走势趋缓、出口疲软、用工成本上升、订单外流等因素的影响,使得我国制造业发展形势很不乐观,日常运营面临重大挑战,信用风险开始集中暴露。因此,如何有效测度制造业上市公司的信用风险亟待解决。基于此,文章采用定性和定量的方法对我国制造业上市公司信用风险的测度进行了理论分析和实证研究。首先,文章对国内外信用风险文献资料,按照信用风险测度的指标模型和非指标模型这一标准进行了梳理与学习,并从研究对象、研究视角、研究方法三个方面总结归纳了学者们的研究成果对作者所研究课题的启发。其次,文章对信用风险的理论知识进行了简要说明,并对信用风险的测度模型按照指标模型和非指标模型这一标准进行了解析,以加深对模型的认知,为后续信用风险测度模型的选择与运用奠定理论基础。同时,文章也对我国制造业上市公司信用风险的成因及特点进行了分析。通过上述信用风险文献综述、理论知识以及各种模型的解析,我们发现：现有的针对国内上市公司信用风险的测度分别从会计报表截面数据和资本市场时间序列数据两个单方面因素展开研究。而公司信用风险发生与否是一个长期积累的过程,且会计报表截面数据是信用主体的历史记载,具有“向后看”的特性；资本市场时间序列数据作为先行指标,具有“前瞻性”。为此,如果将两者有效结合,使其较为全面地涵盖公司未来违约的预测信息,那么其整体的预测效果是否比从单一方面考虑会更好是文章研究的口的所在。因此,文章结合实际情况以及相关模型的适用性,先运用财务比率指标作为会计报表截面数据,基于主成分和判别分析对制造业上市公司的信用风险进行初步测度；接着运用股票价格时间序列数据作为资本市场数据,基于KMV模型计算违约距离DD,然后将财务比率指标主成分因子与违约距离DD共同纳入到Logistic模型中,实现会计报表截面数据和资本市场时间序列数据的有效结合,以提高对我国制造业上市公司信用风险的测度能力。实证研究结果表明,会计报表截面数据和资本市场时间序列数据有效结合后,使得Logistic模型整体预测准确率达到85%,相比仅有财务比率指标数据的分析模型,其整体预测准确率提高了5%,充分说明财务比率指标与违约距离DD的结合对模型的预测是有效的。至此,文章从会计报表截面数据以及资本市场时间序列数据两个方面,通过运用不同模型和方法的优势互补性,试探性地提出我国制造业上市公司信用风险两阶段测度体系,从而更有效地对制造业的信用风险进行测度,并为其有效管理提供初步的理论依据。
[Abstract]:Since 2008 the international financial crisis, the world economy has experienced turbulence and recession test, and now into the difficult recovery after the international financial crisis. Affected by the downward economic cycle and the national industrial policy adjustments and other factors, the domestic real estate industry, shipbuilding, steel trade, etc. are included in the high risk industry the manufacturing industry is the foundation of the national economy, the economic index is good operation of the strong support. However, due to China's recent macroeconomic trends slowed, weak exports, rising labor costs, affect the outflow of orders and other factors, the development of China's manufacturing industry is very optimistic about the situation, the daily operation is facing major challenges, credit risk began to focus on the exposure. Therefore, how to effectively measure the credit risk of Listed Companies in the manufacturing industry to be solved. Based on this, the dissertation uses the method of qualitative and quantitative of China's manufacturing industry. The measure of the credit risk of the city company has been analyzed theoretically and empirically.
First of all, the article of domestic and international credit risk literature, according to the index model of credit risk measurement and non index model which is a standard of the carding and study, and from the research object, research perspective, research methods inspired three aspects summarizes the research achievements of scholars research on them.
Secondly, the credit risk of the theoretical knowledge of this paper are briefly described, and the measure model of credit risk in accordance with the index model and non index model of this standard are analyzed, in order to deepen the cognition model, lay the theoretical foundation for the follow-up and use of credit risk measurement models. At the same time, the article also on credit risk China's manufacturing industry listed companies causes and characteristics are analyzed.
The credit risk through literature review, theoretical knowledge and analytical models, we find that the existing credit risk of domestic listed companies to measure and carries out the research on accounting statements section data and time series data of capital market and two unilateral factors. The company credit risk occurs or not is a long-term process of accumulation, and accounting report section data is the main credit history, has the characteristic of "look back"; the capital market time series data as a leading indicator, are "forward-looking". Therefore, if the effective combination of the two, the forecast information comprehensively covers the company's future defaults, then whether the overall prediction results than considering one aspect of the paper is to be better in place.
Therefore, based on the actual situation and the applicability of the model, we use financial ratios as the accounting statements section data, principal component analysis and discriminant analysis for manufacturing the credit risk of listed companies based on the preliminary measure; then use the stock price time series data as the capital market data, the default distance calculation model based on KMV and DD. The financial ratio index of principal components and DD into the distance to default to the Logistic model, realize the effective combination of the accounting statements section data and time series data of the capital market, to improve the ability to measure the credit risk of China's Listed Companies in the manufacturing industry.
The empirical results show that the effective combination of the accounting statements section data and time series data of capital market, makes the Logistic model the overall prediction accuracy reached 85%, compared with the model only financial ratio index data, the overall prediction accuracy is increased by 5%, and shows the default ratios from DD to forecast model is combined with effective. Thus, this article from the two aspects of accounting statements and the section data of capital market time series data, through the complementary advantages of the use of different models and methods, we put forward some of China's manufacturing industry listed companies' credit risk measurement system of two stages, so as to more effectively to the manufacturing industry to measure credit risk, and to provide the theoretical basis for its effective management.

【学位授予单位】：东华大学
【学位级别】：硕士
【学位授予年份】：2013
【分类号】：F425;F832.51;F224

【参考文献】