No contagion, only volatility: U.S. equity correlations during COVID-19

Purpose-During the COVID-19 crisis, correlations between U.S. equity returns and those of its three primary trading partners—Canada, China, and Mexico—rose sharply. In particular, the average correlation climbed from 0.56 in 2019 to 0.83 in 2020, the peak year. This study investigates whether this nearly 48% surge signals a contagion effect stemming from COVID-19. Methodology-Price data of ADRs for Canada, China, and Mexico, traded on the New York Stock Exchange were collected and returns on equally weighted portfolios for each country were computed. Using the returns on the country portfolios of ADRs and the US equity stock index S&P 500, cross-country correlations between the U.S. and each of its major trading partner countries were computed. These estimates were revised by applying the volatility adjustment procedure recommended by Forbes and Rigobon (2002). The revised estimates of correlations were tested whether they differed from the stable period values. Findings-During the pandemic, unadjusted Correlations between U.S. equities and each of its major trading partners increased. These estimates were then adjusted for the increased volatility. The revised correlations were not found to be significantly different from their pre-pandemic values. Conclusion-Estimates of correlations between U.S. equity and its major trading partner countries increased dramatically during the pandemic, implying possible contagion. This conclusion would be premature and incorrect as volatility changes are ignored in the estimation process. When corrected for it, the revised estimates of correlations do not support the presence of contagion effect. Keywords: COVID-19, pandemic, correlations, contagion, ADR