ABSTRACT
With the outbreak of many major default events in China's bond market, the problem of corporate bond default prediction has increasingly become the focus of academic and practical circles, and machine learning and deep learning algorithms have been widely used. By constructing a random forest-BP neural network model and a random forest model based on convolutional neural network, we make an empirical analysis of the financial data of listed companies whose bonds default or are due to be paid from 2014 to 2019, and compare the prediction results of the algorithms to test the effectiveness of the models. The empirical results show that the random forest model based on convolutional neural network is the best in the prediction of corporate credit default, and the combined model has better performance than the single model.
- E. Altman, N. Fargher, and E. Kalotay, “A simple empirical model of equity-implied probabilities of default,” The Journal of Fixed Income, vol. 20, pp. 71-85, Winter 2011.Google ScholarCross Ref
- R. L. Rosner, “Earnings manipulation in failing firms,” Contemporary Accounting Research, vol. 20, pp. 361-408, 2003.Google ScholarCross Ref
- M. Malekipirbazari, and V. Aksakalli, “Risk assessment in social lending via random forests,” Expert Systems with Applications, vol. 42, pp. 4621-4631, 2015.Google ScholarDigital Library
- X. Ma, J Sha., D. Wang, “Study on a prediction of P2P network loan default based on the machine learning LightGBM and XGboost algorithms according to different high dimensional data cleaning,” Electronic Commerce Research and Applications, vol.31, pp.24-39, 2018.Google ScholarCross Ref
- J. Duan, “Financial system modeling using deep neural networks (DNNs) for effective risk assessment and prediction,” Journal of the Franklin Institute, vol. 356, pp.4716-4731, 2019.Google ScholarCross Ref
- X. Man, and Y. Chen, “Research on early warning of default risk of corporate bonds based on neural network,” Economic & Social Development Research, vol. 27, pp.1, 2020.Google Scholar
Index Terms
- Prediction of Enterprise Credit Bond Default based on Random Forest and Neural Network
Recommendations
Informed Bond Trading, Corporate Yield Spreads, and Corporate Default Prediction
Taking advantage of recently augmented corporate bond transaction data, we examine the pricing implications of informed trading in corporate bonds and its ability to predict corporate defaults. We find that microstructure measures of information ...
Credit Default Swaps and Debt Overhang
We analyze the impact of credit default swaps (CDSs) trading on firm investment, long-term debt financing, and valuation. In our model, the firm is endowed with a real option to initiate a project and enhance its future growth. Its creditors have access ...
Corporate Bond Pricing Model with Interaction between Liquidity and Credit Risk
This study derives a liquidity and credit risk-adjusted capital asset pricing model and investigates the model using the data set in China's corporate bond market. Our research shows that the channels through which liquidity risk affects corporate bond ...
Comments