News impact curve for stochastic volatility models

This paper proposes a couple of new methods to compute the news impact curve for stochastic volatility (SV) models. The new methods incorporate the joint movement of return and volatility, which has been ignored by the extant literature. The first method employs the Bayesian Markov chain Monte Carlo scheme and the other one employs the rejection sampling. The both methods are simple, versatile, and applicable to various SV models. Contrary to the monotonic news impact functions in the extant literature, the both methods give the U-shaped news impact curves comparable to the GARCH models. They also capture the volatility asymmetry for the asymmetric SV models.