Algorithmic Collusion: Supra-Competitive Prices via Independent Algorithms

Motivated by their increasing prevalence, we study outcomes when competing sellers use machine learning algorithms to run real-time dynamic price experiments. These algorithms are often misspecified, ignoring the effect of factors outside their control, e.g. competitors' prices. We show that the long-run prices depend on the informational value (or signal to noise ratio) of price experiments: if low, the long-run prices are consistent with the static Nash equilibrium of the corresponding full information setting. However, if high, the long-run prices are supra-competitive---the full information joint-monopoly outcome is possible. We show this occurs via a novel channel: competitors' algorithms' prices end up running correlated experiments. Therefore, sellers' misspecified models overestimate own price sensitivity, resulting in higher prices. We discuss the implications on competition policy.

[1]  Mallesh M. Pai,et al.  Competing Models , 2019, ArXiv.

[2]  Ran Spiegler,et al.  The market for quacks , 2006 .

[3]  Peter Auer,et al.  Using Confidence Bounds for Exploitation-Exploration Trade-offs , 2003, J. Mach. Learn. Res..

[4]  T. Lai,et al.  Optimal learning and experimentation in bandit problems , 2000 .

[5]  Emilio Calvano,et al.  Artificial Intelligence, Algorithmic Pricing and Collusion , 2018, American Economic Review.

[6]  Francesco Bonchi,et al.  Algorithmic Bias: From Discrimination Discovery to Fairness-aware Data Mining , 2016, KDD.

[7]  Nabil I. Al-Najjar.  Decision Makers as Statisticians: Diversity, Ambiguity, and Learning , 2009 .

[8]  Emilio Calvano,et al.  Algorithmic Pricing What Implications for Competition Policy? , 2018, Review of Industrial Organization.

[9]  Omar Besbes,et al.  Dynamic Pricing Without Knowing the Demand Function: Risk Bounds and Near-Optimal Algorithms , 2009, Oper. Res..

[10]  Sanjog Misra,et al.  Scalable Price Targeting , 2017 .

[11]  Mallesh M. Pai,et al.  Coarse decision making and overfitting , 2014, J. Econ. Theory.

[12]  Daron Acemoglu,et al.  Fragility of Asymptotic Agreement Under Bayesian Learning , 2008 .

[13]  Annie Liang,et al.  Games of Incomplete Information Played by Statisticians , 2016, ArXiv.

[14]  R. Porter,et al.  NONCOOPERATIVE COLLUSION UNDER IMPERFECT PRICE INFORMATION , 1984 .

[15]  John R. Hauser,et al.  Website Morphing , 2009, Mark. Sci..

[16]  Anton J. Kleywegt,et al.  Learning and Pricing with Models That Do Not Explicitly Incorporate Competition , 2015, Oper. Res..

[17]  Catherine E. Tucker,et al.  Algorithmic Bias? An Empirical Study of Apparent Gender-Based Discrimination in the Display of STEM Career Ads , 2019, Manag. Sci..

[18]  Wang Chi Cheung,et al.  Dynamic Pricing and Demand Learning with Limited Price Experimentation , 2017 .

[19]  Peter Auer,et al.  Finite-time Analysis of the Multiarmed Bandit Problem , 2002, Machine Learning.

[20]  A. V. den Boer,et al.  Dynamic Pricing and Learning: Historical Origins, Current Research, and New Directions , 2013 .

[21]  J. Gittins.  Bandit processes and dynamic allocation indices , 1979 .

[22]  Peter Auer,et al.  UCB revisited: Improved regret bounds for the stochastic multi-armed bandit problem , 2010, Period. Math. Hung..

[23]  J. Harrington.  DEVELOPING COMPETITION LAW FOR COLLUSION BY AUTONOMOUS ARTIFICIAL AGENTS† , 2018, Journal of Competition Law & Economics.

[24]  Christo Wilson,et al.  An Empirical Analysis of Algorithmic Pricing on Amazon Marketplace , 2016, WWW.

[25]  R. Agrawal.  Sample mean based index policies by O(log n) regret for the multi-armed bandit problem , 1995, Advances in Applied Probability.

[26]  Eric M. Schwartz,et al.  Dynamic Online Pricing with Incomplete Information Using Multi-Armed Bandit Experiments , 2018, Mark. Sci..

[27]  B. Jullien,et al.  OPTIMAL LEARNING BY EXPERIMENTATION , 1991 .

[28]  R. Weber.  On the Gittins Index for Multiarmed Bandits , 1992 .