Using intraday statistics for the estimation of the return variance

This paper proposes new estimators for the daily return variance which are based on common intraday statistics (opening, high, low, and closing prices). These estimators utilize information contained in products of absolute values of uncorrelated intraday statistics. An empirical study of nine components of the Dow Jones Industrial Average from 1962-01-02 to 2013-03-13 shows that they outperform existing estimators over all stocks and time periods. -- range-based estimation ; Brownian motion ; folded normal distribution