TIME SERIES ANALYSIS AND STRUCTURAL CHANGE IN ARMENIAN PUBLIC DEBT (2016-2025)

Abstract

Using monthly data and econometric methods, we analyze stationarity properties, optimal forecasting models, and regime changes in debt dynamics. Results show that all debt series are integrated of order one and require differencing for stationarity. ARIMA models indicate that total debt needs second differencing due to post-2022 acceleration, while external debt follows a random walk with drift. Structural break tests identify five significant breaks corresponding to April 2018, August 2019, January 2021, May 2022, and May 2024. These breaks delineate three fiscal regimes: Pre-Crisis (2016-2019) with stable external dominance, Crisis (2020-2021) with shock-driven accumulation, and Post-Crisis (2022-2025) with rapid domestic debt expansion. The compositional transformation from 79% external debt in 2016 to 52% domestic debt by 2024 represents a substantial shift in Armenia’s debt financing approach.