摘要
基于财务困境预测的研究大多局限于截面数据的静态计量,即以T-2,T-3(T代表被特殊处理的年份)的财务指标,运用主成分进行预测,而忽视了公司财务状况的变化过程,因此本文证实了建立基于时序立体数据表的全局主成分分析模型,比经典主成分分析建立的截面模型预测准确度要高。并利用Mann-Whitney U检验对财务指标进行筛选,以我国上市医药公司为研究样本建立预测模型,对其财务状况进行了良好的预测。
Most researches on prediction of financial distress mainly focus on static measurement of cross-sectional data, which use principal component analysis to predict financial distress on T-2＇s or T-3＇s （T representing the year receiving special treatment） financial indicators, ignoring the process of companies＇ financial changing. This paper confirms that the accuracy of the model based on cubic time series data （CTSD） and all-round principal component analysis （PCA） is better than the model based on classic principal component analysis from the cross＇sectional data. Also it screens the financial indicators by Mann-Whitney U＇s test and builds a model on the data from pharmaceutical companies and makes a good prediction.
出处
《科技与管理》
2010年第6期65-69,共5页
Science-Technology and Management
基金
上海理工大学博士启动基金
