Akademik Veri Yönetim Sistemi
Sustainable Finance, Springer Nature, ss.59-81, 2026
This study investigates the non-linear determinants of financial development across 157 countries with data (2000–2021) using the IMF’s Financial Development Index (FDI). While existing literature emphasizes linear relationships between financial institutions (FII) and markets (FMI), this research addresses a critical gap by employing machine learning—specifically decision tree modeling—to uncover hidden thresholds and interactions. Leveraging IMF data, the analysis reveals that financial market development (FMI) is the primary driver of FDI, with a root node split at FMI ≤ 0.357 distinguishing low/moderate development (FDI = 0.348) from higher development (FDI = 0.638). Institutional efficiency (FII) mediates outcomes, particularly in advanced economies, where FII > 0.513 elevates FDI to 0.793–0.881. The findings highlight stark disparities: advanced economies (FDI = 0.62, CoV = 12.9%) exhibit stability through institutional depth, while emerging (FDI = 0.33, CoV = 23.2%) and low-income economies (FDI = 0.15, CoV = 34.7%) face volatility due to over-reliance on FMI and weak institutions. The implied R2 values (0.98) suggest high explanatory power in terminal nodes, meaning the model captures most variance within those subsets. The study contributes a policy-focused framework, emphasizing institutional reforms for emerging markets and foundational access for low-income countries. Future research should integrate regulatory quality, fintech adoption, and macroeconomic stability variables to refine thresholds and address dataset biases. This work advances the scholarly discourse by bridging machine learning with financial economics, offering a roadmap for targeted interventions to enhance global financial resilience.