Collusive Compensation Schemes Aided by Algorithms

Sophisticated collusive compensation schemes such as assigning future market shares or direct transfers are frequently observed in detected cartels. We show formally why these schemes are useful for dampening deviation incentives when colluding firms are temporary asymmetric. The relative attractiveness of each of these schemes is shaped by firms’ ability to predict future market conditions, possibly aided by algorithms. Prices and profits are inverse u-shaped in prediction ability. Assigning future market shares is optimal when prediction ability is intermediate, and otherwise direct transfers are optimal. Competition authority's limited resources should be utilized to respond to these changing market conditions.