Stackelberg stopping games

TL;DR

This paper explores a Stackelberg variant of the Dynkin game, analyzing leader-follower dynamics, strategy concepts, and introducing entropy regularization to ensure equilibrium existence.

Contribution

It introduces a Stackelberg stopping game framework, characterizes leader strategies, and proposes entropy regularization to address equilibrium existence issues.

Findings

Precommitment strategies may not attain the supremum.

Randomized equilibrium strategies may not exist.

Entropy regularization ensures the existence of a regular randomized equilibrium.

Abstract

We study a Stackelberg variant of the classical Dynkin game in discrete time, where the two players are no longer on equal footing. Player 1 (the leader) announces her stopping strategy first, and Player 2 (the follower) responds optimally. This Stackelberg stopping game can be viewed as an optimal control problem for the leader. Our primary focus is on the time-inconsistency that arises from the leader-follower game structure. We begin by using a finite-horizon example to clarify key concepts, including precommitment and equilibrium strategies in the Stackelberg setting, as well as the Nash equilibrium in the standard Dynkin game. We then turn to the infinite-horizon case and study randomized precommitment and equilibrium strategies. We provide a characterization for the leader's value induced by precommitment strategies and show that it may fail to attain the supremum. Moreover, we construct a counterexample to demonstrate that a randomized equilibrium strategy may not exist. Then we introduce an entropy-regularized Stackelberg stopping game, in which the follower's optimization is regularized with an entropy term. This modification yields a continuous best response and ensures the existence of a regular randomized equilibrium strategy, which can be viewed as an approximation of the exact equilibrium.