Jump Contagion among Stock Market Indices: Evidence from Option Markets*

We analyze the contagious propagation of jumps among international stock market indices, using a rich panel of high-frequency stock and options data ( 692,892 option contracts) over the period 2006–2015. We propose a bivariate option pricing model designed to allow for time and space amplification of jumps in option markets. We develop a semi-parametric estimation procedure, which employs a continuum of moment conditions in GMM with implied states and non-parametric high-frequency spot volatility estimation. A partial-information approach is introduced to reduce the computational complexity arising in the multivariate setting. We find statistical evidence of jump contagion both within and between stock market indices. Our results reveal that jump contagion from the US to the UK is more pronounced than vice versa, whereas the jump contagion effects between the US and Germany stand on equal footing. We illustrate the statistical and economic importance of capturing jump contagion for risk management, option pricing, and scenario analysis. We show that accounting for jump contagion, employing scenarios based on the Global Financial Crisis, leads to an increase of capital requirements in the UK from 6.3% to 8.4% for each unit invested.