Conditional convex orders and measurable martingale couplings

DOI10.3150/16-BEJ827MaRDI QIDQ2405124zbMATH OpenWikidataFDO

Authors Lasse Leskelä, Matti Vihola

Publication date 21 September 2017

Published in Bernoulli (Search for Journal in Brave)

Full work available at URL https://arxiv.org/abs/1404.0999, https://projecteuclid.org/euclid.bj/1494316832

zbMATH Keywords

conditional expectationconvex stochastic orderingcouplingsincreasing convex stochastic orderingprobability kernels

Mathematics Subject Classification ID

Inequalities; stochastic orderings (60E15) Martingales with discrete parameter (60G42)

Abstract: Strassen's classical martingale coupling theorem states that two real-valued random variables are ordered in the convex (resp. increasing convex) stochastic order if and only if they admit a martingale (resp. submartingale) coupling. By analyzing topological properties of spaces of probability measures equipped with a Wasserstein metric and applying a measurable selection theorem, we prove a conditional version of this result for real-valued random variables conditioned on a random element taking values in a general measurable space. We also provide an analogue of the conditional martingale coupling theorem in the language of probability kernels and illustrate how this result can be applied in the analysis of pseudo-marginal Markov chain Monte Carlo algorithms. We also illustrate how our results imply the existence of a measurable minimiser in the context of martingale optimal transport.

Review text

This paper provides a characterisation of the existence of a coupling between two probability kernels over a measurable space \(S\). It turns out that such a coupling exists if and only if the probability distributions induced at each point of \(S\) follow a certain convex stochastic ordering. The authors also consider the convex ordering of two random variables conditioned on a third one. They give a characterisation of this in terms of the existence of a special coupling between the three random variables.

Recommendations

• A new family of one dimensional martingale couplings
• Canonical supermartingale couplings
• Shadow couplings
• On the existence of ordered couplings of random sets -- with applications
• A coupling proof of convex ordering for compound distributions

Cited in

(6)

• Variational formulas for asymptotic variance of general discrete-time Markov chains
• A newsvendor problem with markup pricing in the presence of within-period price fluctuations
• The use of a single pseudo-sample in approximate Bayesian computation
• On Strassen's theorem for support functions
• Structural properties of multi-period martingale optimal transport problems and applications
• Quantizations of probability measures and preservation of the convex order

This page was built for publication: Conditional convex orders and measurable martingale couplings

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2405124)