Abstract:This paper addresses the critical challenge of detecting financial crises in their early stages given their profound economic and societal consequences. It investigates the efficacy of machine learning models by comparing the standard econometric model of Logistic Regression with k-nearest Neighbours, Random Forest, Extremely Randomised Trees, Support Vector Machine, and artificial Neural Network models in a cross-validation experiment. The study utilises financial crisis observations between 1870 and 2020 from the MacroHistory database and up to 14 early warning indicators from different theoretical backgrounds. The results demonstrate that advanced machine learning models, particularly Random Forest and Extremely Randomised Trees, outperform Logistic Regression, as measured by the area under the receiver operating characteristic curve (AUROC). This finding holds across different crisis data sources, early warning indicator sets, and model specifications. Furthermore, this study emphasises the importance of model explainability for both researchers and policymakers by employing Accumulated Local Effect (ALE) plots to unravel the complex predictive processes of advanced machine learning models and provide deeper insight into the underlying crisis dynamics. Nonlinear connections to financial crises were identified for several early warning indicators, which are difficult to capture by linear models. Several U-shaped relationships were observed, highlighting the disruptive role of strong economic changes for financial stability. The results emphasise the potential of advanced machine learning methods to limit the societal impact of financial crises by enabling earlier intervention by policymakers and public institutions.

Evaluation of International Monetary Policy Coordination: Evidence from Machine Learning Algorithms

International Monetary Policy Coordination through Communication: Chasing the Loch Ness Monster

Uncovering heterogeneous regional impacts of Chinese monetary policy

An Exploration to the Correlation Structure and Clustering of Macroeconomic Variables

Application of Time Series Analysis and Regression Algorithm in Assessing the Effectiveness of Monetary Policy in the Era of Digital Economy

Predicting financial crises: an evaluation of machine learning algorithms and model explainability for early warning systems

International monetary policy spillovers: Evidence from a time-varying parameter vector autoregression

Assessing machine learning for forecasting economic risk: Evidence from an expanded Chinese financial information set

Application of Machine Learning in Forecasting International Trade Trends

Navigating Uncharted Waters: The Transformation of the Bank of Korea's Monetary Policy in Response to Global Economic Uncertainty

Sovereign Debt and Currency Crises Prediction Models Using Machine Learning Techniques

Unveiling the Impact of Macroeconomic Policies: A Double Machine Learning Approach to Analyzing Interest Rate Effects on Financial Markets

Cross-border spillovers in G20 sovereign CDS markets: cluster analysis based on K-means machine learning algorithm and TVP–VAR models

Econometrics of Machine Learning Methods in Economic Forecasting

Opening the Black Box: Machine Learning Interpretability and Inference Tools with an Application to Economic Forecasting

Analysing the Influence of Macroeconomic Factors on Credit Risk in the UK Banking Sector

Forecasting Sovereign Risk in the Euro Area via Machine Learning

U.S. Monetary Policy Spillovers to Emerging Markets: Both Shocks and Vulnerabilities Matter

Modelling Sovereign Credit Ratings: Evaluating the Accuracy and Driving Factors using Machine Learning Techniques

Monetary policy, Twitter and financial markets: evidence from social media traffic

Confronting Machine Learning With Financial Research