We investigate the continuity of expected exponential utility maximization
with respect to perturbation of the Sharpe ratio of markets. By focusing only
on continuity, we impose weaker regularity conditions than those found in the
literature. Specifically, for markets of the form $S = M + \int \lambda d<M>$,
we require a uniform bound on the norm of $\lambda \cdot M$ in a suitable $bmo$
space.
In this paper we extend the stability results of [4]}. Our utility
maximization problem is defined as an essential supremum of conditional
expectations of the terminal values of wealth processes, conditioned on the
filtration at the stopping time $\tau$. The stability result, in particular,
implies that in the framework of [4], the optimal wealth at any given stopping
time is stable with respect to changes in the Sharpe ratio and initial wealth.
To establish our results, we extend the classical results of convex analysis to
maps from $L^0$ to $L^0$.