Predicting the Financial Distress of Chinese Listed Companies Using a Novel Hybrid Model Framework with an Unbalanced Data Perspective


















































Predicting the Financial Distress of Chinese Listed Companies Using a New Hybrid Model Framework with an Imbalanced Data Perspective – Journal of Risk Model Validation





  • A new hybrid model framework is built to solve the problem of predicting the financial distress of listed Chinese companies using unbalanced data.
  • This novel hybrid model framework is developed based on logistic regression and backpropagation neural networks combined with a security level SMOTE.
  • We get 19 important characteristics for the prediction of financial distress, which cover six categories, including solvency, profitability, leverage, operational capacity, growth capacity and leaders.

When predicting financial hardship, an unbalanced business dataset can lead to overfitting of the majority class and lead to poor classifier performance. The problem of classification with unbalanced data is therefore a realistic and critical problem. In this paper, a new hybrid model framework is constructed to solve the problem of predicting the financial distress of listed Chinese companies using unbalanced data. This framework is developed based on logistic regression and backpropagation neural networks combined with the safe level synthetic minority oversampling technique. We validate the model on a dataset of Chinese listed companies and compare the proposed model with seven benchmark models. The results confirm that the proposed model has superior performance. In addition, we find 19 important characteristics that significantly influence the financial distress of Chinese listed companies.

Sorry, our subscription options are not loading at this time

Please try again later. Contact our customer service if this problem persists.

New to Risk.net? See our subscription options

If you already have an account, please login here.

Sarah J. Greer