Assessment of the Impact of Digitization on Inflation Processes and Financial Stability in Azerbaijan

Abstract This study examines whether payment digitization is associated with inflation and selected financial vulnerabilities in Azerbaijan using 96 monthly observations from January 2018 to December 2025. The cashless share of domestic card transaction value is the main payment indicator; transaction counts and point of sale infrastructure provide alternatives. Persistence tests, exploratory ARDL bounds screening and dynamic change regressions are combined with seasonal and dated controls, heteroskedasticity and autocorrelation consistent inference and sensitivity analysis. The cashless value share increased from 17.375% to 67.754%, while overdue lending and deposit dollarization declined. These trends do not identify a causal relationship. In the principal inflation equation, the lagged change in the cashless share has a coefficient of 0.105 (p = 0.085; 95% confidence interval − 0.015 to 0.226). Payment lags are jointly significant in the untrended log-risk equation (p = 0.046), but their signed sum is imprecise (p = 0.092); adding a trend reverses its sign. The principal dollarization estimate is also imprecise (p = 0.832). Results vary across proxies, covariance bandwidths and samples. Inference is conditional on exploratory model selection, and the bounds screens do not justify an equilibrium interpretation. The findings document payment transformation while leaving its independent macrofinancial benefits unresolved, supporting statistical harmonization and targeted policy evaluation.