Regime-sensitive dynamics in global ESG markets: Evidence from a Markov-Switching TVP-VAR model
摘要
This study investigates the global return connectedness of ESG markets across low- and high‐volatility regimes. We employ a two-stage methodology: (i) a Markov-switching framework to identify transition probabilities between volatility states and (ii) a regime-specific time-varying parameter VAR model to measure dynamic connectedness. The dataset consists of daily observations from April 1, 2016 to January 31, 2023, covering ten developed and eleven emerging ESG indices. The results reveal three key insights. First, ESG markets exhibit substantially stronger return connectedness during high-volatility phases relative to stable periods, highlighting the susceptibility of ESG asset integration under stress. Second, ESG connectedness becomes significantly more pronounced after COVID-19 in high-volatility regimes, whereas lower-volatility periods show greater integration pre-pandemic. Third, the network structure demonstrates heterogeneous spillovers between developed and emerging regions; yet, Canada and the United States consistently act as dominant net transmitters of shocks across both regimes. These outcomes hold meaningful implications for risk management, diversification, and regulatory strategies in sustainable finance.