Subject. Gradient boosting algorithms for forecasting credit spreads of corporate bonds in the Russian debt instrument market. Objectives. To develop and empirically test a concept for building predictive models of credit spreads. Methods. Methods of multivariate statistical analysis, regression modelling and ensemble machine learning were applied. Gradient boosting models were used as the main toolkit. To assess the contribution and significance of predictors, the SHapley Additive exPlanations (SHAP) interpretation methodology was employed. Results. It has been established that the use of ensemble gradient boosting models provides a significant improvement in the accuracy of credit spread forecasting compared to traditional regression approaches. The developed models identify complex nonlinear interactions between factors at various levels and demonstrate stability during periods of market shocks. Structured approaches to integrating micro- and macroeconomic indicators into a unified forecasting model have been proposed. Conclusions. Gradient boosting, particularly the CatBoost model, is effective as a tool for analyzing and forecasting credit spreads. The applicability of the method in financial analytics, risk management and supporting investment decisions has been substantiated. Limiting factors include the dependence of forecast quality on the completeness and reliability of the input data, as well as the need for regular model updates. The proposed approaches could be implemented in systems for automated investment decision-making, credit scoring and stress testing, aimed at enhancing the transparency and stability of the Russian corporate debt market.
Kozlov S.V., Sedenkov S.A. [LSTM and GRU model analysis for time series forecasting]. International Journal of Open Information Technologies, 2024, vol. 12, no. 7, pp. 43–50. (In Russ.) EDN: DMKFKO
Karminsky A.M., Dyachkova N.F. [Empirical study of the relationship between credit cycles and changes in credit ratings]. Zhurnal Novoi ekonomicheskoi assotsiatsii, 2020, no. 4, pp. 138–161. (In Russ.) DOI: 10.31737/2221-2264-2020-48-4-6 EDN: LVYGXP
Kosnikov S.N., Demchenko T.S., Molchanov F.A. [Economic cycles and their impact on financial markets]. Estestvenno-gumanitarnye issledovaniya, 2025, no. 2, pp. 588–593. (In Russ.) EDN: ERHWUU
Erofeeva T.M. [Predicting the yield spreads on Russia’s debt market]. Vestnik Moskovskogo universiteta. Seriya 6: Ekonomika, 2021, no. 6, pp. 54–76. (In Russ.) DOI: 10.38050/0130010520216.3 EDN: MTFLWJ
Gu S., Kelly B., Xiu D. Empirical asset pricing via machine learning. Review of Financial Studies, 2020, vol. 33, iss. 5, pp. 2223–2273. DOI: 10.1093/rfs/hhaa009 EDN: DCWZEU
Wang S., Chen B. Credit card attrition: An overview of machine learning and deep learning techniques. Ekonomika. Upravlenie, 2023, vol. 2, no. 4, pp. 134–144. (In Russ.) DOI: 10.47813/2782-5280-2023-2-4-0134-0144 EDN: YQLKTW
Korostelkina I.A., Dedkova E.G., Korostelkin M.M., Mamatkulova O.I. [Application of methods of assessment and forecasting of credit activity in the risk management system in a commercial bank]. Finansovye rynki i banki, 2025, no. 5, pp. 255–262. (In Russ.) DOI: 10.24412/2658-3917-2025-5-255-262 EDN: NXQNXU
Samokhvalova K.V. [Peculiarities of the corporate bonds market development in modern Russia]. Mezhdunarodnyi zhurnal gumanitarnykh i estestvennykh nauk, 2020, no. 3-2, pp. 60–65. (In Russ.) DOI: 10.24411/2500-1000-2020-10262 EDN: JFUBHH
Machado M.R., Osterrieder J., Chen D. Forecasting Commercial Customers Credit Risk Through Early Warning Signals Data: A Machine Learning based Approach. SSRN Electronic Journal, 2024. DOI: 10.2139/ssrn.4754568
Galimnurov A.A., Ismagilova A.S. [The application of mathematical modeling for forecasting corporate bond spreads]. Inzhenernyi vestnik Dona, 2025, no. 2, pp. 721–737. (In Russ.) EDN: PDEGSR
Shiltsina E.A. [Application of machine learning for time series forecasting]. Vestnik nauki, 2025, vol. 2, no. 5, pp. 953–960. (In Russ.) EDN: GHFRIT
Aghayev N.B., Nazarli D.Sh. [Hybrid forecasting model of non-scheduled passenger air transportation]. Nauchnyi vestnik MGTU GA, 2025, vol. 28, no. 2, pp. 8–21. (In Russ.) DOI: 10.26467/2079-0619-2025-28-2-8-21 EDN: GMDLDM
Dzhunkeev U. Forecasting inflation in Russia using gradient boosting and neural networks. Russian Journal of Money and Finance, 2024, vol. 83, no. 1, pp. 53–76. EDN: IANEFK
Berg F., Kölbel J.F., Rigobon R. Aggregate confusion: The divergence of ESG ratings. Review of Finance, 2022, vol. 26, iss. 6, pp. 1315–1344. DOI: 10.1093/rof/rfac033 EDN: BRYCLT
Oukhouya H., Kadiri H., El Himdi Kh., Guerbaz R. Forecasting International Stock Market Trends: XGBoost, LSTM, LSTM-XGBoost, and Backtesting XGBoost Models. Statistics, Optimization and Information Computing, 2023, vol. 12, iss. 1, pp. 200–209. DOI: 10.19139/soic-2310-5070-1822 EDN: QVDXWM
Abbasov R. Revolutionizing risk management in banking: Implementation of AI/ML-based gradient boosting machines (GBM) and random forest models for credit risk management. International Journal of Research in Finance and Management, 2023, vol. 6, iss. 1, pp. 441–444. DOI: 10.33545/26175754.2023.v6.i1d.324 EDN: SBKDCJ
Konstantinov A.V. Deep gradient boosting for regression problems. Computing, Telecommunications and Control, 2021, vol. 14, iss. 3, pp. 7–19. DOI: 10.18721/JCSTCS.14301 EDN: CZBFQN
Peykani P., Peymany Foroushany M., Tanasescu C. et al. Evaluation of Cost-Sensitive Learning Models in Forecasting Business Failure of Capital Market Firms. Mathematics, 2025, vol. 13, iss. 3. DOI: 10.3390/math13030368 EDN: LDAOKC
