We present a simple and straightforward method for computing Hurst exponent of fractal, self-similar time-series by direct use of the defining relationship.
It is based on a set of series derived from the original time series constructed considering log-returns for increasing delays. The method can be applied to selfsimilar time series of any kind, irrespective of their origin.
In the present study we only consider financial time series and the computed Hurst exponent is used for estimation of market volatility. As case study, the exchange rates of ROL, CZK, GBP and SEK versus US Dollar and Euro are considered.