Markets with Heterogeneous Agents: Dynamics and Survival of Bayesian vs. No-Regret Learners
This work addresses the dynamics of heterogeneous learning agents in markets, providing insights for economics and machine learning, though it is incremental in proposing a hybrid approach.
The paper tackles the problem of comparing Bayesian and no-regret learners in asset markets, finding that low regret is insufficient for survival against Bayesian learners, while Bayesian learning is fragile and no-regret learning is more robust. It proposes a balanced strategy that improves robustness and adaptability to distribution shifts, bridging economic survival concepts with regret minimization.
We analyze the performance of heterogeneous learning agents in asset markets with stochastic payoffs. Our main focus is on comparing Bayesian learners and no-regret learners who compete in markets and identifying the conditions under which each approach is more effective. Surprisingly, we find that low regret is not sufficient for survival: an agent can have regret as low as $O(\log T)$ but still vanish when competing against a Bayesian with a finite prior and any positive prior probability on the correct model. On the other hand, we show that Bayesian learning is fragile, while no-regret learning requires less knowledge of the environment and is therefore more robust. Motivated by the strengths and weaknesses of both approaches, we propose a balanced strategy for utilizing Bayesian updates that improves robustness and adaptability to distribution shifts, providing a step toward a best-of-both-worlds learning approach. The method is general, efficient, and easy to implement. Finally, we formally establish the relationship between the notions of survival and market dominance studied in economics and the framework of regret minimization, thus bridging these theories. More broadly, our work contributes to the understanding of dynamics with heterogeneous types of learning agents and their impact on markets.