A Multidimensional Framework for Quantifying Brazil–China Commodity Trade Dependence Using the Commodity-Specific Sustainability Index

We propose the Commodity-Specific Sustainability Index (CSSI), a multidimensional system for quantifying Brazil–China commodity trade dependence that integrates environmental, economic, and social sustainability metrics with conventional trade dynamics. Traditional trade metrics often overlook sustainability risks due to their focus on volume or monetary value. The CSSI combines three dimensions of sustainability risk (environmental impact, economic resilience, and social well-being) into a single assessment framework for major commodities, including soybeans and iron ore. The framework uses a dynamic weighting mechanism that adjusts sub-indices depending on policy priorities and stakeholder inputs, and a Transformer-based time series model captures relationships between CSSI trends with bilateral trade flows along with external shocks, enabling the predictive analysis of sustainability-driven trade adjustments. Furthermore, the CSSI replaces conventional trade volumes with sustainability-adjusted counterparts that are then incorporated into standard trade frameworks such as gravity equations. Our analysis of soybeans and iron ore from 2015 to 2022 shows that conventional dependence metrics overestimate trade dependence by 12–19% (95% CI: 10.8–21.2%, p < 0.001) for commodities with a high environmental footprint. The predictive model, built entirely based on publicly accessible data sources, produces a mean absolute error of 5.5% (±0.8%) in forecasting quarterly trade flows, outperforming ARIMA (6.8% ± 0.5%) and LSTM (6.1% ± 0.6%). The CSSI’s novelty is its holistic approach to sustainability–trade connections, providing policy makers and researchers with a tool to assess long-term commodity resilience, beyond traditional economic metrics.