Title: A Representation Theorem for Expectations of Functionals of Fractional Brownian Motion

Abstract: We show that we can represent the expectation of a functional of fractional Brownian motion as the result of applying an exponential operator to the functional, evaluated at a later time T, and along a particular path. This is a generalization of the work of Schellhorn and Jin [8], who proved this representation for the expectation of a functional of Brownian motion. Our representation is however less general than the one in Schellhorn and Jin [8], since the latter holds also for conditional expectations. The exponential operator can be calculated explicitly in a series expansion, which resembles the Dyson series of quantum mechanics. We provide three examples where our representation gives useful series expansions of expectations in several models. The first example is the fractional Merton model of interest rates, and the series expansion for the bond price can be simplified into a regular exponential. Whereas the latter is a well-known result, our next two examples seem to be original results. We show how to precisely calculate the first terms of the series of a bond price in a fractional Cox-Ingersoll-Ross model with time-dependent volatility. We also provide a new power series representation (divergent but sufficient to generate moments) of the characteristic function of exponential of fractional Brownian motion, and this representation can be naturally extended to all lognormal random variables.