Estimation of continuous-time stochastic volatility models with jumps using high-frequency data

This paper proposes a method of inference for general stochastic volatility models containing price jumps. The estimation is based on treating realized multipower variation statistics calculated from high-frequency data as their unobservable (fill-in) asymptotic limits. The paper provides easy-to-check conditions under which the error in estimation resulting from this approximation is op(1) and additional ones under which it is op(1/T), where T is the number of days in the sample. Extensive Monte Carlo analysis shows that the proposed estimation method works well in finite samples, provided asymptotic approximations are used. The estimation technique is applied to the estimation of two semiparametric models.