Time-consistent stopping under decreasing impatience (Q1691445)

The paper is devoted to a dynamic theory for stopping problems in continuous time. Three main features are: first, the authors provide a precise definition of a sophisticated stopping policy (or equilibrium stopping policy) in continuous time. Specifically, they introduce the operator, which describes the game-theoretic reasoning of a sophisticated agent. Sophisticated policies are formulated as fixed points of this operator, which connects to the concept of subgame-perfect Nash equilibrium. Second, they introduce a new, iterative approach for finding equilibrium strategies. For any initial stopping policy \(\tau\), they apply the operator described above to \(\tau\) repetitively until it converges to an equilibrium stopping policy. Under appropriate conditions, this fixed-point iteration indeed converges, which is the main result of this paper. Third, when an agent starts to do game-theoretic reasoning and to look for equilibrium strategies, he is not satisfied with an arbitrary equilibrium. Instead, he works on improving her initial strategy to turn it into an equilibrium. Proposed iterative approach specifically represents this improving process.