Are Jumps Contagious? An Empirical Investigation of Jumps Transmission Mechanisms in the Nasdaq Sector Indexes

This article relies on the comparison of the realized variance and the realized bipower variation to provide a nonparametric extraction of jumps in Nasdaq sector indexes. It proceeds with their empirical analysis along two directions: intensity and size. Whereas the jump intensity appears to be sector-specific, the average jump contribution to the total variance is approximately constant across sectors. Moreover, unlike volatility, jumps do not cluster through time. A multivariate analysis reveals that jump arrivals are significantly linked even though the contemporaneous occurrence of discontinuities does not increase the average jump size. Finally, despite their interdependence, jumps do not seem to be very contagious and few spillover effects can be observed.