On the saddle point of a zero-sum stopper vs. singular-controller game

We construct a saddle point in a class of zero-sum games between a stopper and a singular-controller. The underlying dynamics is a one-dimensional, time-homogeneous, singularly controlled diffusion taking values either on $\mathbb{R}$ or on $[0,\infty)$. The games are set on a finite-time horizon, thus leading to analytical problems in the form of parabolic variational inequalities with gradient and obstacle constraints. The saddle point is characterised in terms of two moving boundaries: an optimal stopping boundary and an optimal control boundary. These boundaries allow us to construct an optimal stopping time for the stopper and an optimal control for the singular-controller. Our method relies on a new link between the value function of the game and the value function of an auxiliary optimal stopping problem with absorption. We show that the smooth-fit condition at the stopper's optimal boundary (in the game), translates into an absorption condition in the auxiliary problem. This is somewhat in contrast with results obtained in problems of singular control with absorption and it highlights the key role of smooth-fit in this context.