For utility functions $u$ finite valued on $\mathbb{R}$, we prove a duality
formula for utility maximization with random endowment in general
semimartingale incomplete markets. The main novelty of the paper is that
possibly non locally bounded semimartingale price processes are allowed.
Following Biagini and Frittelli \cite{BiaFri06}, the analysis is based on the
duality between the Orlicz spaces $(L^{\widehat{u}}, (L^{\widehat{u}})^*)$
naturally associated to the utility function. This formulation enables several
key properties of the indifference price $\pi(B)$ of a claim $B$ satisfying
conditions weaker than those assumed in literature. In particular, the
indifference price functional $\pi$ turns out to be, apart from a sign, a
convex risk measure on the Orlicz space $L^{\widehat{u}}$.