论文部分内容阅读
本文以我国2008～2010年3年首次被ST的56家公司为样本研究对象,采用配对抽样的方法为56家ST样本公司选取56家非ST公司作为配对样本。以发生财务危机的当年作为T年.由于上市公司被特别处理前一年(T-1年)的年报和其是否被特别处理在时间上几乎一致,所以T-1年的数据预警价值不大。本文着眼于T-2年和T-3年的预警研究,重点以T-2年为例,详细阐述了Logistic预警模型及修复的预警模型建立过程。本文首先搜集了112家公司T-2年和T-3年的20项财务指标的数据,通过主成分分析法对财务指标进行筛选,使用Logistic回归对剩下的财务指标建立财务危机预警模型。得到T-2年整体预警准确度为85.7%,T-3年整体预警准确度为68.8%。接下来本文选取合适的投入产出指标计算各公司的SBM效率,并将其作为一项非财务指标重新进行Logistic回归得到修正的预警模型。结果表明修正后的模型T-2年整体预警准确度提高到89.3%,T-3年整体预警准确度提高到71.4%。
In this paper, the first time in our country from 2008 to 2010, 56 companies of ST were selected as the sample study. Fifty-six non-ST companies were chosen as the paired samples from 56 ST sample companies by means of paired sampling. The year of the financial crisis was taken as year T. Since the annual reports of the special year (year T-1) handled by the listed company were nearly identical in time to the special treatment, the estimated expenditures for the year T-1 were not significant . This article focuses on the T-2 and T-3 pilot war studies, focusing on T-2 years as an example, detailing the Logistic warning model and repair early warning model building process. This article first collected the data of 20 financial indicators from T-2 and T-3 of 112 companies, screened the financial indicators through principal component analysis, and used Logistic regression to establish the financial crisis early-warning model for the remaining financial indicators. The current average of T-2 estimates for the full-scale budget was 85.7%, while the T-3 estimates for the entire period were estimated at 68.8%. Next, we select the appropriate input-output indicators to calculate the SBM efficiency of each company, and use it as a non-financial indicator to re-logistic regression to obtain a revised early warning model. The results showed that the revised model T-2 year adjusted for the overall adjusted forecast reached 89.3%, T-3 year overall adjusted for the highest forecast 71.4%.