Campos-Martins, S., & Amado, C. (2025). Modelling dynamic interdependence in nonstationary variances with an application to carbon markets. Journal of Economic Dynamics and Control, 173. https://doi.org/10.1016/j.jedc.2025.105062This paper introduces a new multivariate conditional correlation GARCH model, the Multiplicative Time-Varying Extended Conditional Correlation GARCH (MTV-ECC-GARCH), designed to capture dynamic interdependence among assets or markets under nonstationary variance. The model extends traditional CC-GARCH frameworks by incorporating two key features: a nonstationary long-term component that captures structural shifts in unconditional volatility, and a short-term dynamic component allowing cross-market volatility interactions. Ignoring nonstationarity, the study notes, can lead to spurious volatility transmission. Parameter estimation is conducted using a maximization by parts algorithm, which simplifies the computation by estimating each variance equation separately. A Lagrange Multiplier (LM) test is proposed to detect volatility interactions under nonstationary conditions. Applying the model to carbon futures (CEF) and a media-based climate concern index (CCM), results show significant dynamic interdependence—particularly from climate-related media concerns to carbon market volatility—when nonstationarity is properly modeled, highlighting the model’s robustness and practical relevance for financial volatility analysis.