Time-varying and Nonlinear Nature of Causality Between Cryptocurrency, Real Estate, and Stock Markets in South Korea

This study examines the time-varying and nonlinear causal relationships among cryptoassets, real estate, and stock markets in Korea, with particular focus on the Gangnam 3 district. The cryptocurrency indicator was expanded to include the market capitalization of Bitcoin, Ethereum, and altcoins (TOTAL2). Using time-varying parametric vector autoregression, Markov-switching, and nonlinear Granger causality econometric models, we investigate how asset market linkages evolve through phase transitions and structural shifts. The results show, that the causal relationship between virtual assets and Gangnam 3 real estate, overlooked in linear models, becomes statistically significant in nonlinear models under conditions of high volatility or sharp price changes, with direction and strength shifting over time. Additionally, asset market linkages adjusted in response to external shocks (e.g., interest rate changes, economic cycles, and global financial events). Furthermore, the Gangnam 3 real estate market displayed stronger interactions with crypto and stock markets than other regions of Seoul and the country, and responded differently to external shocks.