This thesis presents a mathematical formulation of informational inhomogeneity in financial
markets, with emphasis on its impact on asset volatility, the notion of information
extraction, and the role of information providers. We begin with a brief review of the BHM
framework, which models the market filtration by an information process consisting of a
signal and a noise term, such that the signal-to-noise ratio is determined by the information
flow rate. Motivated by the observations that valuable information is rarely circulated
homogeneously across financial markets, and that the information flow rate is typically
random, we introduce, in the first part of the thesis, an extension of the BHM approach
that leads to the simplest class of stochastic volatility models. In this extended framework
we derive closed form expressions: for (a) asset price processes; (b) pricing formulae for
options; and (c) option deltas. We show that the model can be calibrated to fit volatility
surfaces reasonably well, and that it can be used effectively to model information manipulation.
In the second part we introduce a framework for the valuation of information. In
particular, a new formulation of the utility-indifference argument is introduced and used as
a basis for pricing. We regard information as a quantity that converts a prior distributions
into a posterior distributions. The amount of information can thus be quantified by relative
entropy. The key to our theory is to equate the maximised a posterior utility with the a
posterior expectation of the utility of the a priori optimal strategy. This formulation leads
to one price for a given quantity of upside, and another for a given quantity of downside
information. Various intuitive, as well as counterintuitive implications (for example, price
of information is not necessarily an increasing function of the volume of information) of
our theory are discussed in detail.