On the minimal entropy martingale measure for Lévy processes

In the present paper, a new and simple approach is provided for proving rigorously that for general Lévy financial markets the minimal entropy martingale measure and the Esscher martingale measure coincide. The method consists in approximating the probability measure P by a sequence of Lévy preserving probability measures Pn with exponential moments of all order. As a by-product, it turns out that the problem of finding the minimal entropy martingale measure for the Lévy market is equivalent to the corresponding problem but for a certain one-step financial market. The existence of the Esscher martingale measure (and hence the minimal entropy martingale measure) will be characterized by using moment generating functions of the Lévy process.