A COMPARATIVE ANALYSIS OF COINTEGRATION TECHNIQUES: EVALUATING THE PERFORMANCE OF FMOLS, ARDL, AND VECM IN ESTIMATING LONG-RUN ECONOMIC RELATIONSHIPS

Abstract

This study conducts a comparative analysis of three prominent cointegration techniques Fully Modified Ordinary Least Squares (FMOLS), Autoregressive Distributed Lag (ARDL), and Vector Error Correction Model (VECM) to evaluate their performance in estimating long-run economic relationships in Nigeria from 1990 to 2023. Using macroeconomic variables (GDP, exchange rate, CO₂ emissions, industrialization, inflation, interest rate, and trade openness), the study confirms non-stationarity through Augmented Dickey–Fuller tests. To stabilize variance and reduce the degree of non-stationarity, all variables were transformed into natural logarithms prior to model estimation. FMOLS reveals a significant long-run negative impact of CO₂ emissions on GDP, while ARDL captures dynamic short-run effects such as exchange rate volatility and inflation adjustments. VECM validates a stable cointegrating relationship, with industrial output and CO₂ emissions driving long-run GDP movements, whereas exchange rates and inflation exert adverse effects. The error correction mechanism indicates rapid convergence toward equilibrium (14.2%–18.2% adjustment speed). Although FMOLS exhibits a lower explanatory power (adjusted R² = 0.451), ARDL and VECM demonstrate robustness in capturing short-run asymmetries and multivariate adjustments, respectively. Overall, FMOLS excels in estimating long-run elasticities, ARDL provides flexible lag dynamics, and VECM supports causality and long-run adjustment analysis. The findings highlight the importance of sustainable industrialization, inflation control, and cleaner energy adoption to mitigate the growth-constraining effects of CO₂ emissions.