Quantile Connectedness and Extreme Risk Spillovers in the Capital Markets of Caribbean Rim Countries from a QVAR Approach

Abstract

This study examines systemic connectedness and risk spillovers in the capital markets of Caribbean Rim countries using a Quantile Vector Autoregression (QVAR) framework and weekly equity returns from five economies (the U.S., Mexico, Jamaica, Colombia, and Venezuela) over 2019–2025. Unlike mean-based models, QVAR effectively captures risk transmission across the entire conditional distribution and identifies tail dynamics during extreme events. The empirical analysis yields three key findings. First, under normal market conditions (τ = 0.50), risk transmission exhibits a “core-periphery” hierarchy with a modest Total Connectedness Index (TCI of 24.92), where the U.S. acts as the dominant net transmitter while Jamaica and Venezuela serve as passive net receivers. In extreme regimes, however, connectedness surges to roughly 70, triggering a structural role reversal: Mexico transforms into the primary risk transmitter during market crashes (τ = 0.05), while Venezuela becomes a severe risk-absorption trap. Second, quantile-specific spillovers exhibit a U-shaped descriptive pattern, where the estimated extreme upside connectedness (τ = 0.95 TCI of 70.57) is descriptively slightly higher than the extreme downside connectedness (τ = 0.05, TCI of 69.47). This descriptive pattern may be statistically associated with the tendency of illiquid peripheral markets lacking pricing autonomy to be swept into synchronized co-movements during speculative capital inflows. Finally, the median dynamic TCI fluctuates between 13.52 and 34.38, indicating that the system’s baseline connectedness experienced a significant structural decline in mid-2023 and has since remained at a consistently low level; while 2023 Fed easing expectations coincided with lowered connectedness to annual troughs, the 2025 U.S. “reciprocal tariff” event was followed by a multi-month rebound, which might descriptively reflect transmission via the macro-expectations channel. By characterizing the descriptive patterns of tail connectedness and topological reconfiguration, this study provides vital empirical evidence for regional financial surveillance.