Multi-Frequency Price Discovery in Emerging Market Equity Futures: Evidence from Explainable AI and Econometric Modelling

This study examines price discovery dynamics between spot and futures markets in emerging economies using the MSCI Emerging Markets index. While prior research highlights the informational role of futures markets, limited evidence exists on how these dynamics vary across time horizons and how they can be interpreted using explainable artificial intelligence (AI). A multi-frequency framework (daily, weekly, and monthly) is employed, combining econometric techniques with explainable machine learning. The results show that price discovery is strongly frequency-dependent, with significant cointegration and futures-led causality at higher frequencies that weaken over longer horizons. Spot markets adjust more slowly than futures, confirming their secondary role in information incorporation. SHAP-based analysis further reveals that futures returns are the dominant drivers of spot market movements. This study contributes by integrating econometric modelling with explainable AI to provide a transparent and robust framework for analysing price discovery. The findings have implications for risk management, market efficiency analysis, and the application of interpretable machine learning in financial markets.