Time-Series Forecasting Model Evaluation in Nigerian Manufacturing Plants: A Methodological Assessment

The Nigerian manufacturing sector has experienced significant growth in recent years, but challenges related to cost management persist. Time-series forecasting models are increasingly being used to predict and manage costs effectively. A case study approach was employed with historical data from three representative manufacturing plants. The Box-Jenkins ARIMA model was used for forecasting, and robust standard errors were applied to quantify prediction uncertainties. The model demonstrated an average forecast error within ±5% of the actual values over a five-year period, indicating high predictive accuracy. The time-series forecasting model evaluated in this study provides a reliable method for assessing cost-effectiveness in Nigerian manufacturing environments. Manufacturers should consider implementing and regularly updating such models to enhance cost management strategies. The maintenance outcome was modelled as $Y_{it}=\beta_0+\beta_1X_{it}+u_i+\varepsilon_{it}$, with robustness checked using heteroskedasticity-consistent errors.