From debt to trade: How macroeconomic forces shape ecological footprint in Vietnam–Insights from wavelet-based artificial neural network

Against the backdrop of increasing ecological pressure in emerging economies, this study examines how macroeconomic conditions are associated with ecological footprint (EF) in Vietnam. Using annual data from 2000 to 2024, the study employs Random Forest, Wavelet Artificial Neural Network (ANN) regression, and Wavelet-ANN Granger-type predictability analysis to capture nonlinear, horizon-dependent, and directional predictive relationships between EF and five macroeconomic factors: public debt (DEBT), inflation (INF), financial development (FD), trade openness (TRA), and labor force (LF). The findings reveal that the interplay between macroeconomic variables and EF is heterogeneous and vary across horizons. FD and LF are associated with lower EF over medium- and long-run horizons, while TRA and DEBT exhibit mixed and horizon-specific patterns. INF plays a comparatively weak and unstable role. The Wavelet-ANN Granger predictability results further illustrate that statistically significant predictive relationships are limited and horizon-specific. Specifically, TRA and LF emerge as the most persistent predictors of EF, DEBT becomes relevant over medium and long horizons, while FD provides predictive information in the medium horizon. Overall, the results suggest that ecological sustainability in Vietnam is more closely linked to structural macroeconomic conditions than to short-term fluctuations, underscoring the need to integrate environmental objectives into fiscal management, trade policy, financial-sector reform, and labor-market upgrading.