At the renowned Finance Forum Liechtenstein, Prof. Dr. Michael Hanke and doctoral candidates Merlin Bartel and Sebastian Petric from the University of Liechtenstein presented their latest research findings on a highly topical issue: the use of artificial intelligence (AI) to forecast banking crises. The workshop attracted strong interest from both industry professionals and investors.

The research team used the U.S. regional banking crisis of 2023 as a case study to demonstrate how machine learning can be applied effectively to detect potential risks at an early stage. Using advanced AI models, the researchers analysed a wide range of macroeconomic and bank-specific data to identify warning signs that may indicate impending market disruptions.

What stood out in particular was the practical relevance of the results: the model not only has the capacity to detect crisis potential but can also generate actionable investment recommendations. For instance, it was shown that an investor following the model’s signals could exit the market in advance of negative developments—resulting not only in reduced downside risk but also in significantly improved portfolio performance, especially in terms of final wealth and the Sharpe ratio.

The results underline the growing importance of data-driven approaches in finance and demonstrate how academic research can help strengthen the resilience of financial markets. This project exemplifies the successful integration of theory and practice—and serves as yet another indicator of the innovative research being conducted at the University of Liechtenstein at the intersection of financial economics and artificial intelligence.