Unpacking Financial Returns in Global Transportation and Logistics: An Integrated Econometric, Time Series, and Machine Learning Analysis of ESG and Market Dynamics

This study analyzes the financial returns of globally operating transportation and logistics firms, examining the impact of Environmental, Social, and Governance (ESG) factors and broader market dynamics. Utilizing a dataset of 14 firms over a decade (2012–2021), the research first employs panel data regressions to quantify the influence of ESG scores and established asset pricing factors on excess returns. Second, market-level time series analysis, specifically Vector Autoregression (VAR) and GARCH modeling, explores the dynamic interrelationships between market returns and aggregate ESG performance. Finally, advanced machine learning classifiers (Random Forest, XGBoost) predict future excess return direction, identifying key predictive features through importance and SHAP analysis. Findings reveal that traditional market factors significantly influence firm returns, while ESG performance demonstrates a nuanced, often negative association. Machine learning models show high predictive power, highlighting critical financial and environmental determinants. These results are useful for investors seeking to optimize risk-adjusted returns, industry stakeholders making strategic investments in sustainability and technology, and policymakers developing resilient and green logistics frameworks.