This paper investigates the potential of AI-driven anomaly detection models to identify exploitable inefficiencies in stock markets. Using Isolation Forest for anomaly detection and CatBoost regression for predictive modeling, we analyze a subset of S&P 500 stocks to assess whether machine learning techniques can uncover arbitrage opportunities. Our findings suggest that AI-based strategies can generate excess returns testing the market efficiency hypothesis.
AI-Driven Anomaly Detection in Stock Markets: Testing Market Efficiency with Machine Learning
Vittorio Carlei
;Donatella Furia;Piera Cascioli;Alessandro Ceccarelli
2026-01-01
Abstract
This paper investigates the potential of AI-driven anomaly detection models to identify exploitable inefficiencies in stock markets. Using Isolation Forest for anomaly detection and CatBoost regression for predictive modeling, we analyze a subset of S&P 500 stocks to assess whether machine learning techniques can uncover arbitrage opportunities. Our findings suggest that AI-based strategies can generate excess returns testing the market efficiency hypothesis.File in questo prodotto:
Non ci sono file associati a questo prodotto.
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
