Study on predicting corporate credit ratings using CART†
Geunhee Lee1 · Kang-Min Kim2 · Hoi-Jeong Lim3
123Graduate School of Data science, Chonnam National University, Public Data Analytic Center
Correspondence to: † This work was partly supported by Innovative Human Resource Development for Local Intellectualization program through the Institute of Information & Communications Technology Planning & Evaluation (IITP) grant funded by the Korea government (MSIT) (IITP-2024-RS-2022-00156287) and by Innovative Human Resource Development for Local Intellectualization program through the Institute of Information & Communications Technology Planning & Evaluation (IITP) grant funded by the Korea government (MSIT) (IITP-2024-RS-2022-00156287).
1 Master student, Graduate School of Data Science, Chonnam National University, Public Data Analytics Center, Gwangju 61186, Korea.
2 Master student, Graduate School of Data Science, Chonnam National University, Public Data Analytics Center, Gwangju 61186, Korea.
3 Professor, Graduate School of Data Science, Chonnam National University, Public Data Analytic Center. Gwangju 61186, Korea. E-mail: hjlim@jnu.ac.kr
1 Master student, Graduate School of Data Science, Chonnam National University, Public Data Analytics Center, Gwangju 61186, Korea.
2 Master student, Graduate School of Data Science, Chonnam National University, Public Data Analytics Center, Gwangju 61186, Korea.
3 Professor, Graduate School of Data Science, Chonnam National University, Public Data Analytic Center. Gwangju 61186, Korea. E-mail: hjlim@jnu.ac.kr
Received July 19, 2024; Revised September 5, 2024; Accepted September 10, 2024.
This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/licenses/by-nc/3.0) which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited.
- Abstract
- This study utilizes machine learning techniques to predict the credit ratings of companies based on their financial statements. Five variables used in this study are Market capitalization, Total debt leverage, Market-to-book ratio, Operating margin, and Dividend payer. The target variable from AAA to D, encompassing 21 grades, were converted into numerical values. Prediction was conducted using Classification and Regression Trees, Support Vector Regression, and Multiple Linear Regression models. The experimental results showed that the model using all variables performed best in terms of both R2 and MSE. The evaluation metrics of the model including only three variables(Market capitalization, Total debt leverage, Market-to-book ratio) closely approximated those of the best-performing model. The final model with three variables offers simplicity and ease of interpretation. When market capitalization was below 10 billion KRW, below 50 billion KRW, and below 4.8 trillion KRW, companies typically receive credit ratings of B+, BBB, and A+, respectively. To improve credit ratings, increasing market capitalization and reducing debt leverage were found to be effective strategies.
- Keywords : CART, credit ratings, decision tree, financial statements, machine learning
Full Text(PDF) Free
