Volatility Spillovers in Macao's Gaming Economy before and after COVID-19

Risk analyses of Macao's gaming-tourism economy have long rested on two implicit assumptions: that the co-movement between overnight visitor arrivals and gaming revenue is positive and stable, and that adverse shocks propagate no more severely than favourable ones. Using monthly data from February 2008 to July 2025 and a framework combining GJR-GARCH, Dynamic Conditional Correlation, and the Diebold-Yilmaz spillover decomposition, this paper tests both assumptions and finds that neither survives empirical scrutiny. The dynamic correlation between overnight arrivals and gaming revenue was positive and stable before 2020, effectively collapsed during the COVID-19 border closure, and has remained persistently negative since 2022, a reversal that has not self-corrected in three years of post-pandemic recovery. Separately, the spillover of bad volatility shocks between the two series is 30.6 percent, against 8.2 percent for good volatility shocks, a gap of 22.4 percentage points. Together, these results describe a gaming-tourism system in which the standard assumptions of stability and symmetry both fail simultaneously, and in which any risk model calibrated on pre-2020 data or symmetric shock mechanics will understate downside exposure by a structurally significant margin.