Empirical tests of the bias and efficiency of the extreme-value variance estimator for common stocks