Journal of Risk Model Validation

Risk.net

Predicting financial distress of Chinese listed companies using a novel hybrid model framework with an imbalanced-data perspective

Tong Zhang and Zhichong Zhao

  • A novel hybrid model framework is constructed to solve the problem of predicting the financial distress of Chinese listed companies using imbalanced data.
  • This novel hybrid model framework is developed on the basis of logistic regression and backpropagation neural networks combined with safe-level SMOTE.
  • We obtain 19 important features for financial distress prediction, which cover six categories including solvency, profitability, leverage, operational capacity, growth ability and executives.

When predicting financial distress, an imbalanced data set of company data may cause overfitting to the majority class and lead to bad performance of the classifiers. The problem of classification with imbalanced data is, therefore, a realistic and critical issue. In this paper a novel hybrid model framework is constructed to solve the problem of predicting the financial distress of Chinese listed companies using imbalanced data. This framework is developed on the basis of logistic regression and backpropagation neural networks combined with the safe-level synthetic minority oversampling technique. We validate the model on a data set of Chinese listed companies and compare the proposed model with seven baseline ones. The results confirm that the proposed model has superior performance. Further, we find 19 important features that significantly influence the financial distress of Chinese listed companies.

Sorry, our subscription options are not loading right now

Please try again later. Get in touch with our customer services team if this issue persists.

New to Risk.net? View our subscription options

If you already have an account, please sign in here.

You need to sign in to use this feature. If you don’t have a Risk.net account, please register for a trial.

Sign in
You are currently on corporate access.

To use this feature you will need an individual account. If you have one already please sign in.

Sign in.

Alternatively you can request an individual account here: