Cross-Sectional Aggregation and Persistence in Conditional Variance

This paper explores the interactions between cross-sectional aggregation and persistence of volatility shocks. We derive the ARMA-GARCH representation that linear aggregates of ARMA processes with GARCH errors admit, and establish conditions under which persistence in volatility of the aggregate series is higher than persistence in the volatility of the individual series. The practical implications of the results are illustrated empirically in the context of an option pricing exercise.