Benford’s law and intraday microstructure anomalies: Forecasting market movements with high-frequency data

Benford’s Law has been widely used to detect data irregularities across various domains, including finance. While prior work (Hamida et al., 2024) demonstrated its predictive relevance for daily stock returns, this study extends that framework to high-frequency intraday data, offering a new perspective on short-term market dynamics. Specifically, we apply Benford’s Law to the first-digit distributions of intraday stock returns, trading volume, and trade durations, two microstructural variables never explored in Benford-based financial research. Using data from twenty Euronext Paris-listed companies, we uncover significant deviations from Benford’s distribution that are particularly pronounced during anomalous market periods. These deviations are then incorporated into both linear and switching regime models to assess their predictive impact. Our findings reveal that Benford-derived indicators are conditionally informative, showing predictive value primarily during anomalous market periods, whether driven by intentional manipulation or by significant, non-fraudulent irregularities. These indicators function as effective early-warning signals for abnormal intraday dynamics. The study advances both forecasting and forensic finance by integrating a diagnostic tool into real-time predictive modeling in high-frequency trading environments.