Firms learn to price in markets with crowded resources and competition
Multi-Agent Reinforcement Learning in Markets with Congestion
Computer Science and Game Theory
Summary
This paper studies how companies set prices when competing for customers who also consider how busy or crowded a service is. The companies don’t know exactly how price and congestion affect customer choices, so they learn by trying different prices over time. The researchers show that when companies learn independently, they can end up behaving like they are cooperating, even though they are competitors. This behavior depends on how the learning happens and how the situation is represented, offering new understanding of competition in crowded markets.
What this means in practice
- •For market platform designers: Design pricing mechanisms for online services where customers consider both price and waiting times, improving competition modeling.
- •For telecommunications network operators: Predict and influence how competing providers might set prices when network congestion affects customer choices.
Authors
Qixuan Zai, Randall Berry
Abstract
This paper investigates multi-agent reinforcement learning (MARL) in settings where firms compete for customers using congestible resources. We consider Bertrand competition in which firms compete by announcing prices and customers choose among firms based on both price and congestion. The relationship between price, congestion and the quantity of customers willing to accept service is governed by an unknown inverse demand curve, which firms must learn through experience. Each firm is modeled as a self-interested learning agent that chooses its price to maximize profit. A growing literature has shown that independently learning MARL agents can develop tacitly collusive behavior. We examine how such behavior emerges in markets with congestible resources. Our results provide insight into how learning dynamics, state representation, and strategic interaction jointly shape competition, with implications for both economic learning and the design of learning-enabled markets.