Statistical Evaluation of Seasonal Variation in Agricultural Commodity Prices in Nigerian Markets Using Spectral Analysis

Agricultural commodity price seasonality in Nigeria generates income volatility for farmers and food affordability challenges for consumers, and spectral analysis provides a rigorous frequency-domain statistical framework for identifying and quantifying periodic price cycles beyond simple seasonal decomposition. This study applies spectral analysis to evaluate seasonal variation in agricultural commodity prices in Nigerian markets using weekly price data for maize, tomatoes, yam, and groundnut from 2015 to 2023 from the National Agricultural Statistics System. Fourier power spectra, periodograms, and smoothed spectral density estimates were computed for each commodity price series. Significant periodicities were identified by Fisher's Kappa test and Bartlett's Kolmogorov-Smirnov test. Cross-spectral analysis examined lead-lag relationships between commodity prices across northern and southern market centres. All four commodity price series showed statistically significant annual periodicities (Fisher's Kappa significant at p < 0.05). Maize prices showed the strongest semi-annual harmonic, with a secondary six-month cycle reflecting the double harvest pattern in the Middle Belt. Tomato prices exhibited an irregular 14-week cycle corresponding to perishability-driven supply gluts. Cross-spectral analysis confirmed that northern market price cycles led southern market cycles by 2 to 4 weeks for maize and groundnut. The study provides a frequency-domain characterisation of Nigerian agricultural price seasonality and recommends spectral monitoring of commodity price cycles for early warning of food price stress in the NBS Agricultural Price System. Keywords: spectral analysis, commodity prices, agricultural markets, seasonality, Nigeria