Pith. sign in

REVIEW 7 cited by

Algorithmic Collusion Without Threats

Not yet reviewed by Pith; the record is open.

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 2409.03956 v2 pith:7VSNCPVD submitted 2024-09-06 cs.GT cs.LGecon.TH

Algorithmic Collusion Without Threats

classification cs.GT cs.LGecon.TH
keywords moverpricespricingthreatsalgorithmsecondstrategiessupra-competitive
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved
0 comments
read the original abstract

There has been substantial recent concern that pricing algorithms might learn to ``collude.'' Supra-competitive prices can emerge as a Nash equilibrium of repeated pricing games, in which sellers play strategies which threaten to punish their competitors who refuse to support high prices, and these strategies can be automatically learned. In fact, a standard economic intuition is that supra-competitive prices emerge from either the use of threats, or a failure of one party to optimize their payoff. Is this intuition correct? Would preventing threats in algorithmic decision-making prevent supra-competitive prices when sellers are optimizing for their own revenue? No. We show that supra-competitive prices can emerge even when both players are using algorithms which do not encode threats, and which optimize for their own revenue. We study sequential pricing games in which a first mover deploys an algorithm and then a second mover optimizes within the resulting environment. We show that if the first mover deploys any algorithm with a no-regret guarantee, and then the second mover even approximately optimizes within this now static environment, monopoly-like prices arise. The result holds for any no-regret learning algorithm deployed by the first mover and for any pricing policy of the second mover that obtains them profit at least as high as a random pricing would -- and hence the result applies even when the second mover is optimizing only within a space of non-responsive pricing distributions which are incapable of encoding threats. In fact, there exists a set of strategies, neither of which explicitly encode threats that form a Nash equilibrium of the simultaneous pricing game in algorithm space, and lead to near monopoly prices. This suggests that the definition of ``algorithmic collusion'' may need to be expanded, to include strategies without explicitly encoded threats.

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Forward citations

Cited by 7 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score.

  1. Should Demand Models Incorporate Competitor Prices? Oblivious Learning and Algorithmic Collusion

    cs.GT 2026-06 unverdicted novelty 7.0

    In stylized competitive markets with noisy demand and iterated least squares learning, oblivious demand models yield transient collusive patterns that dissipate under sufficient exploration, informed sellers strictly ...

  2. Mean-based algorithms: A lower bound and regret

    cs.LG 2026-06 unverdicted novelty 7.0

    Derives first lower bound on γ_t for mean-based algorithms in unknown-horizon bandit settings, proposes two new algorithms, and shows some are also no-regret.

  3. Hierarchies of No-regret Algorithms

    cs.GT 2026-04 unverdicted novelty 7.0

    No-swap-regret players frequently receive lower utilities than no-regret players in two-player games due to slower effective learning rates, though the reverse holds in some random 7-action games.

  4. Markets with Heterogeneous Agents: Dynamics and Survival of Bayesian vs. No-Regret Learners

    cs.GT 2025-02 unverdicted novelty 7.0

    Bayesian learners can drive out no-regret learners despite logarithmic regret in stochastic markets, but no-regret is more robust; hybrids are proposed to combine strengths.

  5. Domination-Avoiding Learning Agents Cannot Collude

    cs.GT 2026-05 unverdicted novelty 6.0

    Domination-Avoiding agents provably avoid collusion in repeated price-competition markets and avoid playing strategies eliminated by iterated elimination of dominated strategies in any game.

  6. Misspecified Estimate-then-Optimize Leads to Supra-Competitive Prices

    cs.GT 2026-05 unverdicted novelty 6.0

    Misspecified estimate-then-optimize pricing converges to supra-competitive prices when initial random explorations occur in similar ranges, reaching monopoly levels under symmetry.

  7. Misspecified Estimate-then-Optimize Leads to Supra-Competitive Prices

    cs.GT 2026-05 unverdicted novelty 6.0

    Misspecified explore-then-exploit pricing with monopoly-style demand estimation leads to supra-competitive prices when firms explore similar price ranges on the same side of the Nash price.