ARIMA forecasting of prime lending rates from Nigeria’s financial industry

Purpose: This study investigates the behaviour and predictability of Nigeria’s prime lending rates from 1990 to 2026. The research identifies structural shifts, volatility patterns, and the effectiveness of autoregressive components in forecasting interest rate movements within the Nigerian financial ecosystem. Methodology: Adopting an ex post facto research design, the study utilizes monthly time-series data from the Central Bank of Nigeria. In this univariate framework, the prime lending rate serves as the dependent variable, while its own lagged values (AR) and stochastic shocks (MA) function as the independent variables. The analysis involves unit root testing, heteroscedasticity evaluations, and lag identification via ACF and PACF to fit an optimal ARIMA model. Results & conclusion: Findings reveal the PLR is integrated into order one, I(1), with a significant structural shift in January 1990. ARIMA (2,1,2) emerged as the most robust model for capturing mean fluctuations. However, residual diagnostics indicate significant volatility clustering and non-normality, suggesting that while the model effectively predicts price direction, it does not fully account for variance shocks over time. Implication of findings: Precise forecasting is essential for managing interest rate risk and credit pricing. The research recommends that the Central Bank of Nigeria and financial institutions integrate GARCH-type models with ARIMA frameworks to account for volatility clustering. Such evidence-based modelling is critical for developing resilient monetary policies and mitigating systemic financial instability.