Estimating the volatility of stock prices: a comparison of methods that use high and low prices

The volatility of stock prices has played an important role in the financial literature. Different methods of estimating the volatility are suggested and applied to British financial assets. Since we cannot observe the real volatility, we investigate the efficiency of the methods through simulation. The question of which estimator to use rather depends on the distributional assumption of returns. If it is log-normal, methods based on high/low prices are preferred. Furthermore, if there is drift in the data, then one may wish to use a procedure devised by Rogers and Satchell. If the drift is varying with time, the Rogers and Satchell's method is clearly superior.