dc.description.abstract | A striking feature of the prices of financial assets is that their statistical
properties are to some degree universal across difierent assets,
regions and epochs. There is a vast amount of literature on modelling
these so-called stylized facts of financial data, but relatively recently
multifractal processes have been proposed as a new formalism for
financial modelling. Their main power lies in the fact that they capture
many of the main statistical properties of financial time series in an
effective way. The goal of this thesis is to present two multifractal models,
the Multifractal Model of Asset Returns and the Markov-Switching
Multifractal, and to study them in a more complete and rigorous way
than in the literature. | |