Article

A Hybrid Machine-Learning and Econometric Panel VAR Approach to Assess FinTech Adoption and Digital Economy Growth in Emerging Markets

---

Abstract

The rapid rise of Financial Technology (FinTech) has significantly reshaped the global economic landscape, especially in emerging markets where digital platforms bridge gaps in financial access, innovation, and inclusion. However, empirical research quantifying the causal and temporal impact of FinTech adoption on digital economy growth remains limited due to methodological fragmentation and data complexity. This study proposes a hybrid analytical framework integrating Machine Learning (ML) and econometric panel modeling (Panel Vector Auto Regression, or PVAR) to examine how FinTech ecosystems drive digital economic transformation. Using a balanced panel of 10 emerging economies from 2015 to 2024, the ML layer—implemented through Gradient Boosting Regression (GBR), Random Forest (RF), and XGBoost—predicts the share of Digital GDP (DGDP) and evaluates feature contributions using SHAP (Shapley Additive Explanations). The econometric layer applies PVAR to analyze dynamic causality, Impulse Response Functions (IRF), and Forecast Error Variance Decomposition (FEVD) between FinTech Index (FTI), Financial Inclusion (FI), and DGDP. The results reveal a strong and statistically significant causality from FTI → DGDP (p = 0.001), with a peak IRF response of 0.82% after three periods. SHAP analysis confirms FTI (0.237) and Policy Innovation (0.194) as the dominant drivers, while inequality (GINI) and Inflation (INF) exert negative effects. This hybrid model effectively bridges the gap between predictive precision and causal interpretability, offering a replicable approach for data-driven policymaking. The findings emphasize that FinTech-driven digital growth depends not only on technological adoption but also on institutional readiness, human capital, and regulatory innovation, positioning the framework as a strategic tool for sustainable digital transformation in emerging economies.

Keywords: FinTech Adoption, Digital Economy, Financial Inclusion, Machine Learning, Panel VAR, SHAP Analysis

How to Cite:

Izumi, C. & Putawa, R. A., (2025) “A Hybrid Machine-Learning and Econometric Panel VAR Approach to Assess FinTech Adoption and Digital Economy Growth in Emerging Markets”, FinTech Innovation Journal 1(1), 1-16. doi: https://doi.org/10.63913/ftij.v1i1.79