Data-Driven Analysis of Bank Ownership, Credit Allocation, and the Finance-Growth Nexus in the Western Balkans and CEE

Abstract

This paper explores the finance-growth nexus in ten economies of the Western Balkans and Central and Eastern Europe (hereafter referred to as the Extended Balkans) over 2001–2024. Moving beyond aggregate measures of financial depth, the study investigates how the structure of financial intermediation shapes the relationship between finance and economic growth. Particular attention is given to three dimensions: the nonlinearity of the financial depth–growth relationship, the role of foreign bank ownership, and the differential growth effects of household and corporate credit. The study adopts a data-driven empirical framework combining two-way (country and year) fixed-effects estimators with two-step system and difference Generalized Method of Moments (GMM) estimators. The results indicate that the aggregate private-credit-to-GDP ratio does not exhibit a robust positive linear effect on economic growth. Instead, an inverted U-shaped relationship emerges in the dynamic specification (system GMM: −0.38, p = 0.016 for credit and +0.004, p = 0.006 for its squared term), indicating a financial deepening threshold. Hansen panel threshold regression identifies a turning point at approximately 55% of GDP (95% CI: 53–59%), below which an additional percentage point of credit is associated with approximately 0.04 percentage points of higher growth, while beyond this threshold the marginal contribution becomes negligible. The analysis further provides evidence of a foreign-bank-ownership channel. Foreign ownership exhibits a negative level effect on growth (−1.19, p = 0.029), whereas its interaction with financial depth is positive (+2.43, p = 0.071), suggesting that the negative association weakens as financial depth increases. Evidence from the six-country IMF Global Debt Database (IMF-GDD) subsample reveals a pronounced credit-composition asymmetry, with household credit negatively associated with growth (−0.77, p < 0.001) and corporate credit positively associated with growth (+0.30, p < 0.001). The negative household-credit effect is confirmed in the post-crisis fixed-effects specification (−0.12, p = 0.043). The findings remain robust across leave-one-country-out, EU versus non-EU, and per-capita growth specifications, highlighting implications for macroprudential regulation, SME financing, and economic convergence within the EU accession framework.