Explosive housing price dynamics in Türkiye: a city and district-level analysis before and during the COVID-19 pandemic

Abstract This study investigates house price dynamics in Türkiye using monthly data for 55 cities and 87 districts over the period 2010–2022. Employing a three-stage empirical framework, we first detect exuberant episodes using Generalized Sup Augmented Dickey-Fuller (GSADF) and backward Sup ADF (BSADF) right-tailed unit root tests at both the city and district levels. We then analyze co-explosivity among the five largest cities (Istanbul, Ankara, Izmir, Bursa, and Antalya) using logistic regression models. Finally, we explore the drivers of bubble formation across 55 cities using panel logistic regression models. The results indicate that most cities and districts experienced synchronized explosive house price increases across three periods: 2013–2015, 2017–2018, and 2020–2022. The 2021–2022 period saw the most widespread bubbles, particularly in the western and Marmara regions than in the eastern part of the country, indicating a clear regional asymmetry in housing market dynamics. The evidence suggests that housing price exuberance was more synchronized in the largest urban markets. The panel logistic regression analysis revealed that new house sales were associated with bubble formation. In contrast, mortgage-financed sales and higher mortgage rates reduced the likelihood of bubbles, highlighting the stabilizing role of credit costs. Macroeconomic shocks from the COVID-19 pandemic and low interest rates were key drivers of bubble dynamics. These findings suggest that housing market stability requires locally targeted macroprudential and credit policies to prevent local and global shocks.