Reinforcement-learning pricing algorithms sometimes converge to supra-competitive prices even in markets where collusion is impossible by design or cannot be an equilibrium outcome. We analyze when such spurious collusion may arise, and when instead the algorithms learn genuinely collusive strategies, focusing on the role of the rate and mode of exploration.
Calvano, E., Calzolari, G., Denicolò, V., Pastorello, S. (2023). Algorithmic collusion: Genuine or spurious?. INTERNATIONAL JOURNAL OF INDUSTRIAL ORGANIZATION, 90, 1-5 [10.1016/j.ijindorg.2023.102973].
Algorithmic collusion: Genuine or spurious?
Calvano E.;Calzolari G.;Denicolò V.;Pastorello S.
2023
Abstract
Reinforcement-learning pricing algorithms sometimes converge to supra-competitive prices even in markets where collusion is impossible by design or cannot be an equilibrium outcome. We analyze when such spurious collusion may arise, and when instead the algorithms learn genuinely collusive strategies, focusing on the role of the rate and mode of exploration.File in questo prodotto:
| File | Dimensione | Formato | |
|---|---|---|---|
|
True_Or_Fake_IJIO_Revision+submitted.pdf
Open Access dal 26/06/2025
Tipo:
Postprint / Author's Accepted Manuscript (AAM) - versione accettata per la pubblicazione dopo la peer-review
Licenza:
Licenza per Accesso Aperto. Creative Commons Attribuzione - Non commerciale - Non opere derivate (CCBYNCND)
Dimensione
526.5 kB
Formato
Adobe PDF
|
526.5 kB | Adobe PDF | Visualizza/Apri |
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
