In this report supervised by OQAM, the analysts Niklas Gälldin, Pontus Neumann, and Oscar Näslund Cuesta explore the possibility of predicting public Nordic companies’ credit ratings using machine learning models. By utilizing historical quarterly financial data, stock data, and Moody’s historical ratings, the hypothesis is that changes in credit ratings can be predicted ahead of their publication, providing a competitive edge.
The developed model showed promising results and predictive abilities regarding a company’s credit risk. However, the limitation of the dataset to the Nordic market significantly impacted the predictive power of the model. While the hypothesis remained unchallenged by the investigation, the need for a wider scope of considered companies is evident.