Dynamic network analysis of stock indices in the Americas

Purpose This study examines dynamic interdependence among stock indices in the Americas and the effects of global crises on these interdependencies that underline contagion patterns across developed and emerging markets. Design/methodology/approach A Dynamic Conditional Correlation (DCC-EGARCH) model is applied to measure time-varying correlations among stock indices from countries in the Americas. Network measures—Normalized Tree Length (NTL), Average Path Length (APL) and Mean Occupation Layer (MOL)—are used to track shifts in market connectivity. At the same time, the Bai and Perron test is used to identify structural changes linked to global events such as COVID-19. Findings Results show that crises increase market cohesion across the Americas, indicated by lower NTL and MOL values. This enhanced connectivity suggests that stock markets respond more synchronously during high-volatility periods, with the United States often central to the contagion network. Originality/value This study contributes to financial contagion research by applying dynamic network analysis to the Americas, offering tools to identify systemic risks. The findings provide insights for policymakers and investors into managing contagion risks in interconnected markets.