The bias corrected weighted bipower variation on high-frequency financial data

It is important to construct precise estimator of financial volatility theoretically and empirically. Calendar effects are dominant character of high-frequency data and realized volatility is biased for microstructure noise in high frequency financial data. This paper make use of realized bipower variation which is robust to get rid of calendar effects by the method of being weighted and then to correct its bias, at last we have an unbiased and efficient financial volatility estimator named bias corrected weighted realized bipower variation.