Successfully reported this slideshow.
We use your LinkedIn profile and activity data to personalize ads and to show you more relevant ads. You can change your ad preferences anytime.

Algorithms and collusion – Ariel EZRACHI – June 2017 OECD discussion

16,146 views

Published on

This presentation by Ariel EZRACHI (Director of the centre for competition law and policy, Oxford University, UK) was made during the discussion “Algorithms and collusion” held at the 127th meeting of the OECD Competition Committee on 23 June 2017. More papers and presentations on the topic can be found out at oe.cd/1-0.

  • Be the first to comment

  • Be the first to like this

Algorithms and collusion – Ariel EZRACHI – June 2017 OECD discussion

  1. 1. Ariel Ezrachi Slaughter and May Professor of Competition Law The University of Oxford, Centre for Competition Law and Policy Virtual Competition OECD Roundtable on Algorithms and Collusion 23 June 2017 Part I
  2. 2. Digitalised Collusion ‘Classic’ collusion  Airline Tariff Publishing  Topkins  Trod Hub and Spoke  Intentional and unintentional alignment of price.  Eturas and Others  Dynamic personalised pricing - example: Gas stations in the Rotterdam area using the same a2i Systems pricing software... 2
  3. 3. Tacit Collusion on Steroids Relevant for transparent concentrated markets with entry barriers, homogeneous products, no brand recognition or loyalty.  ‘Simple’ algorithm-stabilising-mechanism  Transparency | Stability | Speed | Wider participation. Consider increased use of algorithm and increased market concentration. 3
  4. 4. Illustration – Retail Petrol Markets  Chilean retail petrol industry.  German ‘real-time’ transparency measures.  Perth, Australia – ‘Fuel watch’. Improved transparency via online system + Automated real time pricing = Increased likelihood for tacit collusion and softening of competition.
  5. 5. Future technologies So far, analysis based on existing technology. In the future, technology may provide further enhancements:  Big data and big analytics  God View and Artificial Intelligence  Ability to adjust to changing market reality.  Better understanding of the signalling of others.  Detachment from human decision making.
  6. 6. Important caveats  When tacit collusion is the dominant strategy for humans, it would be rational to weave it into the ‘digital strategy’ - the algorithm.  Other outcome are possible when market power or maverick present, or when market conditions dictate competition.  Personalisation is a leading trend and could undermine tacit collusion. Note the emerging hybrid scenario which facilitates simultaneously both tacit collusion and behavioural discrimination
  7. 7. Enforcement Challenge I – Policy & legality  Rational reaction or market manipulation ? Challenge II – Detection  Stealth mode - In the future, could we identify what is the competitive price ? Challenge III - Enforcement  What are the counterfactuals ?  Audit algorithms ?  Shifting the burden to the companies ?  Strict liability ? 7
  8. 8. Ariel Ezrachi Slaughter and May Professor of Competition Law The University of Oxford, Centre for Competition Law and Policy Virtual Competition OECD Roundtable on Algorithms and Collusion 23 June 2017 Part II
  9. 9. Market led Solutions (regardless of illegality)  The market’s ability to correct itself.  Disruptive innovation.  Product differentiation (but note hybrid scenario).  Secrete Deals  Buying Alliances  Joint bids  Algorithmic consumers
  10. 10. Enforcement Challenge I – Policy & legality  Rational reaction or market manipulation ?  Strict liability ? Challenge II – Detection  Stealth mode - In the future, could we identify what is the competitive price ? Challenge III - Enforcement  What are the counterfactuals ?  Audit algorithms ?  Shifting the burden to the companies ? 10
  11. 11. Policy and illegality IF behaviour is condemned – Change the available signals. Clarify the boundary of legality – - Collection and analysis of data. - Identification and action on deviation. - Use of same data pool. - Use of same / similar algorithm. - Alignment of strategy. - Facilitating action. Strict liability ?
  12. 12. Detection and Enforcement Solutions  Audit algorithms (but be aware of limited effectiveness)  Algorithmic Collusion Incubator  Deceleration.  Transparency.  Disruptive algorithm.  Sponsored entry.  Secret deals  Sharing of data.  Market/sector investigations. … Beware of possible distortions.

×