The stock market is full of surprises—rapid swings, extreme values, and unpredictable patterns that challenge even the most seasoned investors. A recent systematic literature review has shed light on how advanced methods could help reduce the risks investors face in this turbulent environment.
The study highlights the potential of combining Extreme Value Theory (EVT) with Machine Learning (ML) to create a more accurate and flexible model for predicting investment risk. While EVT is strong in detecting extreme fluctuations, it becomes complex when dealing with multiple variables. On the other hand, ML offers adaptability and strength in handling large, nonlinear, and high-frequency data. By merging the two, researchers propose a hybrid model that can provide a sharper and more reliable estimate of Value-at-Risk (VaR)—a key measure for financial risk.
From more than 1,100 scientific articles reviewed, only a handful explored the intersection of EVT and ML. This reveals a clear research gap, and the authors suggest that a hybrid framework could better identify extreme events and market shocks, ultimately helping investors make smarter decisions in volatile conditions.
This effort aligns with SDG 8: Decent Work and Economic Growth, since stable and reliable investment risk management can contribute to more sustainable financial systems and economic resilience.
#MachineLearning #ExtremeValueTheory #EconomicGrowth
Link to the paper: https://www.scopus.com/pages/publications/85184030978
21/Mat/2025
