考虑多维效率的上市公司财务困境预警研究数据集

已有的财务困境预警研究一般基于财务指标，或在财务指标基础上引入单一效率指标，而多维效率指标能够更加全面有效地反映不同行业、不同资产规模的上市公司整体状况，从而对上市公司财务困境产生更好的预警效果。课题组从沪深两市选取了148家上市公司进行实证研究，以验证所提出的模型对于上市公司财务困境预警的有效性。为了避免样本类别不均衡对预警效果的影响，选取74家被特别处理（ST）的公司作为财务困境公司样本，并按照1:1的比例选取与之所处行业、资产规模相近的非ST公司作为非财务困境公司样本。数据量为7.13 MB。

Existing studies on financial distress prediction generally rely on financial indicators alone, or add a single efficiency indicator based on financial indicators. By contrast, multi-dimensional efficiency indicators can comprehensively and effectively reflect the overall status of listed companies across different industries and asset sizes, thereby delivering better early warning performance for the financial distress of listed companies. The research team selected 148 listed companies from the Shanghai and Shenzhen Stock Exchanges to conduct an empirical study, aiming to verify the effectiveness of the proposed model in financial distress prediction for listed companies. To mitigate the impact of sample class imbalance on prediction performance, 74 Special Treatment (ST) companies were selected as the financial distress sample, and non-ST companies with matching industries and asset sizes were selected at a 1:1 ratio as the non-financial distress control samples. The total size of this dataset is 7.13 MB.

搜集汇总

背景与挑战

背景概述

该数据集专注于上市公司财务困境预警研究，通过引入多维效率指标来弥补传统财务指标的局限性，以更全面反映不同行业和资产规模公司的整体状况。数据集包含148家上市公司样本，其中74家为ST公司（财务困境），74家为非ST公司（非财务困境），按1:1比例匹配行业和资产规模以确保样本平衡，数据量为7.13 MB，支持数据包络分析、支持向量机等方法的应用。