Testing for a Common Volatility Process and Information Spillovers in Bivariate Financial Time Series Models

Abstract

The paper considers the problem as to whether financial returns have a common volatility process in the framework of stochastic volatility models that were suggested by Harvey et al. (1994). We propose a stochastic volatility version of the ARCH test proposed by Engle and Susmel (1993), who investigated whether international equity markets have a common volatility process. The paper also checks the hypothesis of frictionless cross-market hedging, which implies perfectly correlated volatility changes, as suggested by Fleming et al. (1998). The paper uses the technique of Chesher (1984) in differentiating an integral that contains a degenerate density function in deriving the Lagrange Multiplier test statistic.