Online pricing strategy adapts for better trade outcomes in multilateral markets
Support-Dependent Regret Rates for Profit Maximization in Multilateral Trade
Computer Science and Game Theory
Summary
Imagine an online marketplace where a middleman sets prices for buyers and offers payments to sellers, but a trade only happens if everyone agrees. The paper studies how to set these prices and payments when the values buyers and sellers have are unknown and possibly linked to each other. The authors created an algorithm that learns over time to make better deals by focusing on likely trade situations and ignoring less important ones. This helps improve profits by carefully adapting to the complexity and uncertainty of the market.
What this means in practice
- •For pricing strategy teams: Design pricing algorithms for complex marketplaces where multiple buyers and sellers interact and values are uncertain and correlated.
- •For e-commerce platform operators: Improve automated negotiation systems where trade requires agreement from all parties by focusing learning efforts on likely successful trades.
A theory result. No direct application yet.
Authors
Francesco Bacchiocchi, Matteo Castiglioni, Anna Lunghi, Alberto Marchesi
Abstract
We study online profit maximization in multilateral trade, a setting that generalizes both dynamic pricing and bilateral trade. At each round, an intermediary posts prices to buyers and offers payments to sellers; trade occurs only if every agent accepts. We consider valuations drawn from an unknown joint distribution, allowing arbitrary correlation across agents, under the assumption that each agent's marginal valuation distribution has unknown support of size at most $K$. For constant dimension $d$, we design an algorithm with regret $\widetilde O(\sqrt{K^dT})$. This extends optimal support-dependent guarantees for one-dimensional dynamic pricing to the multidimensional setting of multilateral trade. At the core of our approach is a normalized zooming procedure that adapts to the local probability of trade, avoiding the excessive refinement of nearly flat regions that can arise with standard zooming.