Risk-Averse Stochastic Programming: Time Consistency and Optimal Stopping

Decision making under uncertainty includes reassessing and reevaluating risk after initial decisions. To this end, it is essential to consider a governing value process track its evolution over time. The paper, “Risk-Averse Stochastic Programming: Time Consistency Optimal Stopping,” by Pichler, Liu, Shapiro, develops consistent framework consolidating optimal stopping. paper the notion of time consistency for stochastic multistage optimization problem. Supermartingales envelopes characterizing decisions are given explicitly. With that, dynamic equations derived, which gradually reveal policy. Taking into account requires updating policies, as an explicit example on American options demonstrates.