Abstract
This study investigates the weak-form efficiency of South Africa’s summer grain spot markets, focusing on white maize, yellow maize, sunflower, and soybean. Using daily log return data from 2007 to 2025, we apply autocorrelation, Portmanteau (Q), and heteroskedasticity-robust Lo–MacKinlay variance ratio tests, along with Bai–Perron structural break analysis, Pesaran–Timmermann directional accuracy tests, and mean return per trade calculations. Results reveal significant short-term serial dependence and mean-reverting behaviour across all commodities, indicating partial predictability and deviations from weak-form efficiency. Structural break analysis identifies multiple regimes within the price series, showing that market dynamics are not constant over time. Directional accuracy and MRP results indicate that while some predictability exists, the economic gains from exploiting past prices are small and likely insufficient to overcome trading frictions. These findings suggest that price adjustments are gradual rather than instantaneous, reflecting structural and operational market frictions such as limited liquidity, low adoption of electronic trading, and constrained transparency. Enhancing digital trading platforms, improving real-time price reporting, and investing in storage and logistics could strengthen price discovery and reduce transaction costs. The study provides insights into emerging agricultural markets and highlights the importance of considering market structure when evaluating efficiency.