In-sample Asymptotics and Across-sample Efficiency Gains for Volatility Measurement

Eric Ghysels ( UNC Chapel Hill), Per Mykland (University of Chicago), Eric Renault* ( UNC Chapel Hill)


We revisit the widely used in-sample asymptotic analysis developed by Jacod and Barndorff-Nielsen and Shephard which is extensively used in the realized volatility literature. We show that there are gains to be made in estimating current realized volatility from considering realizations in prior periods. Contrary to prior attempts in the literature, such as those of Andreou and Ghysels and Meddahi, we develop a new approach which is conditional on the volatility path.