Time series analysis of factors influencing white maize price fluctuations in South Africa (1994-2024)

Abstract

This research examines the major economic and environmental factors that shape movements in South Africa’s white maize prices over the period 1994 to 2024. The analysis focuses on how supply dynamics, international trade activity, macroeconomic conditions and climate variability collectively influence price behaviour. Since white maize is a central component of the national food system, identifying the forces behind its price shifts is crucial for informed policy decisions, market oversight and long-term food security strategies. Annual time-series data are employed, and the study applies an ARDL modelling strategy, complemented by bounds testing, an error-correction formulation and Granger causality techniques to evaluate both short-term adjustments and long-run relationships. Preliminary unit root tests reveal that the variables are integrated at different orders, thereby supporting the use of the ARDL approach. The bounds testing procedure confirms the presence of a stable long run cointegration relationship between white maize prices and their determinants. The error-correction term is negative and highly significant (ECT = –1.1293; p < 0.01), indicating a strong pull toward long-run equilibrium, with roughly 112.9 percent of any previous deviation corrected within one year, reflecting a rapid adjustment process with slight overshooting. In the short run, import levels significantly raise maize prices (β = 0.3120; p < 0.01), and fuel costs also contribute positively to immediate price increases (β = 1.1202; p < 0.01). However, the lagged effects of fuel prices (β = –1.1752; p < 0.01) and rainfall (β = –2.2745; p < 0.01) exert downward pressure on prices. In the long-run estimates, only imports and fuel prices remain significant influences, while maize production, exports, the exchange rate and rainfall do not display statistically meaningful long-term effects. Structural break tests reveal notable disruptions in the early 2000s, 2012 and 2021, aligning with periods of climatic stress, macroeconomic instability and global market disturbances. Granger causality outcomes indicate that imports and fuel prices Granger-cause maize prices, while the remaining variables show limited or no predictive influence. Model diagnostics confirm that key statistical assumptions including normality, and homoscedasticity satisfied, supporting the reliability of the estimated ARDL model. The study concludes that white maize prices in South Africa are predominantly shaped by trade-related pressures, energy costs and climatic variation. Policy recommendations emphasise strengthening climate-risk adaptation, managing transport-related cost burdens, improving information flows and reducing exposure to external shocks.The study further notes that annual data may mask short-horizon dynamics and recommends future research that incorporates higher-frequency datasets and nonlinear modelling techniques.