Modeling and simulation of currency exchange rates using MPRE

In this work, assuming as a model the Multifractional Processes with Random Exponent (MPRE), we propose a simulation algorithm able to replicate financial time series, specifically pertaining to the FX market. We show how, properly choosing the functional parameter of the MPRE, the simulated series fit with significant accuracy the actual ones. It is worthwhile to underline that the sole knowledge of the functional parameter ensures by itself that the surrogates succeed in replicating the empirical data. The results can be used in scenario analysis as well as in forecasting.