Forecast of the Consumer Price Index in Mali: A Box-Jenkins Approach

Abstract

This article follows on from existing empirical work on harmonized consumer price indices (HCPIs) seeking to predict the dynamics of consumer price indices in Mali over the twelve months in 2026.The Box-Jenkins estimation methodology was used for short-term forecasting of the IHPC (Integrated Health Care Plan) in Mali. Data were obtained from the National Institute of Statistics (INSTAT) and cover the period from January 2018 to December 2025. The results of diagnostic tests allowed us to select the most appropriate model. Thus, the ARIMA (2, 1, 2) model, which has the lowest information criteria, was chosen for modeling. Similarly, the mean absolute deviation (MAPE) is relatively low (0.52%) compared to observed values, and Theil's U coefficients are less than 1. These results indicate that our model has excellent short-term forecast accuracy, but the overall forecast quality remains moderate.Therefore, further research should focus on comparing the predictive performance of models such as GARCH and SARIMA with that of an ARIMA model for seasonal decomposition, or even hybrid ARIMA-RNA (Artificial Neural Networks) approaches.

JEL ClassificationG21; G32; M12

Item typeEmpirical research