Out of Sample Forecasts of Quadratic Variation
36 Pages Posted: 1 Jun 2006 Last revised: 12 Nov 2008
Date Written: November 10, 2008
We compare the forecasts of Quadratic Variation given by the Realized Volatility (RV) and the Two Scales Realized Volatility (TSRV) computed from high frequency data in the presence of market microstructure noise, under several different dynamics for the volatility process and assumptions on the noise. We show that TSRV largely outperforms RV, whether looking at bias, variance, RMSE or out-of-sample forecasting ability. An empirical application to all DJIA stocks confirms the simulation results.
Keywords: Market microstructure noise, high frequency data, measurement error, realized volatility, two scales realized volatility, out of sample forecasts.
JEL Classification: C14, C22, C53
Suggested Citation: Suggested Citation