Algorithmic collusion with endogenous exploration
Bibliographic Data
| ID | 19048416 |
|---|---|
| Authors | Ivan Conjeaud (0009-0003-1544-574X, corresponding author) |
| Year | 2024 |
| Publication date | 2024-01-01 |
| Open Access | Yes |
| Type | PREPRINT |
| Venue | SSRN Electronic Journal (REPOSITORY) |
| Journal identifiers | ISSN: 1556-5068 • E-ISSN: 1556-5068 |
| Publisher | RELX Group (Netherlands) (PUBLISHER • NL) |
| DOI | 10.2139/ssrn.4962504 |
| OpenAlex | W4403828262 |
| Language | EN |
| References cited | 26 |
I study a two-stage model in which two players simultaneously choose an exploration parameter for their Q-learning algorithms, which then repeatedly play a one-shot game chosen from a class of social dilemmas including the prisoner's dilemma, first and second price auctions as well as Bertrand competition with horizontally differentiated products. The players collect the limit average payoffs obtained by their algorithms. I show that in equilibrium, both players receive payoffs that are strictly higher than the payoffs received in the unique strict Nash equilibrium of the one-shot game. The case of the repeated prisoner's dilemma is then studied using extensive numerical simulations. Their results allow to gain insight on (i) the mechanism causing algorithmic collusion and (ii) the strategic role of exploration levels in the game. These findings have important implications for algorithmic collusion.
Business · Collusion · Economics · Industrial organization · Mathematical economics · Microeconomics · Auction Theory and Applications · Computability, Logic, AI Algorithms · Computer Science
Human-level control through deep reinforcement learning
Least squares quantization in PCM
The Folk Theorem in Repeated Games with Discounting or with Incomplete Information
A Non-cooperative Equilibrium for Supergames
Algorithmic Pricing and Competition
Effective Choice in the Prisoner's Dilemma
The Emergence of Cooperation among Egoists
| Highly cited | No |
|---|