ECONOMETRIC ANALYSIS OF THE FACTORS AFFECTING INTERBANK COMPETITION IN THE DIGITAL BANKING SERVICES MARKET

Abstract

This article presents a comprehensive econometric analysis of the factors affecting interbank
competition in Uzbekistan’s digital banking services market. Based on official quarterly data from the Central Bank of
Uzbekistan for the period 2018 Q1–2024 Q4 (28 quarters) and monthly data from January 2024 to January 2026 (25
months), the study applies a ten-stage econometric framework, including the Augmented Dickey–Fuller (ADF) and
Kwiatkowski–Phillips–Schmidt–Shin (KPSS) unit root tests, the Engle–Granger cointegration test, an Autoregressive
Distributed Lag (ARDL) model, multivariate OLS regression with Newey–West HAC standard errors, robustness
checks across six alternative specifications, the Chow structural break test, and bidirectional Granger causality
testing. The empirical findings reveal a statistically significant long-run negative relationship between remote banking
users and the net interest margin (NIM): the OLS coefficient is β = −1.144 (p < 0.001), with R² = 0.818 and F =
48.43, while the ARDL long-run coefficient of β = −1.247 confirms the robustness of the result. The Engle–Granger
test confirms cointegration between NIM and log(remote users) at the 1 percent significance level (test statistic =
−4.124, p = 0.008). The Herfindahl–Hirschman Index (HHI) declined from 1,500 to 924, representing a 38.4 percent
reduction over the analysis period, with a trend coefficient of −20.06 units per quarter (p < 0.001). Granger causality
testing confirms an asymmetric relationship: digital service users Granger-cause NIM (F = 8.67, p = 0.008), but not
vice versa. State-owned banks exhibit significantly higher NPL levels than private banks (Cohen’s d = 0.664; t = 3.32,
p = 0.003)