Predicting distresses using deep learning of text segments in annual reports

作者：

Highlights：

• A convolutional recurrent neural network is used to predict firms’ default risk.

• Both structured and unstructured data of the firms’ annual reports are employed.

• The unstructured data improves prediction accuracy, especially auditors’ reports.

• Auditors’ reports of large firms are in particular informative.

摘要

•A convolutional recurrent neural network is used to predict firms’ default risk.•Both structured and unstructured data of the firms’ annual reports are employed.•The unstructured data improves prediction accuracy, especially auditors’ reports.•Auditors’ reports of large firms are in particular informative.

论文关键词：Corporate default prediction,Natural language processing,Convolutional neural networks,Recurrent neural networks

论文评审过程：Received 13 November 2018, Revised 30 April 2019, Accepted 30 April 2019, Available online 1 May 2019, Version of Record 11 May 2019.

论文官网地址：https://doi.org/10.1016/j.eswa.2019.04.071