Some comments on Hurst exponent and the long memory processes on capital markets