Competition Commissioner Vestager sums up the situation of companies and the relationship between companies and competition law when she says: It may not be easy to understand how these price algorithms work, but companies can assume their responsibility for anti-trust collusion, i.e. cooperation , not by hiding behind pricing algorithms.<\/p>\n
##Pricing algorithms and their function
\nIt is not obvious at first glance why pricing algorithms could be questionable under antitrust law and even lead to the formation of a cartel by companies. The case is as follows: companies are increasingly focusing on observing the prices of competing products, especially on the Internet. They use these observations to adjust their own prices using a specific price algorithm. For this purpose, the price algorithm as software collects large amounts of data on the Internet, in particular on competitive prices, product availability or consumer behavior. On this basis, it proposes the best price to the company. The European Commission's e-commerce sector inquiry has revealed that 67% of companies that monitor their competitors' prices use software to do this and 80% of these use pricing software at the same time to increase or decrease their own prices accordingly. This is also known as *dynamic pricing*. So far so good.<\/p>\n
##The antitrust problem
\nAntitrust law prohibits all coordinated behavior between companies that could lead to restrictions of competition, in particular through price fixing. A concerted practice is understood to mean any form of coordination of behavior in the absence of an agreement or decision by a body. These price agreements do not have to be \u201cclassically\u201d agreed in some back room at association meetings. It is sufficient if there is a so-called tacit "contact" between companies about the respective competitive behavior. Something else, however, is the conscious parallel behavior. Conscious parallel behavior means that every company and retailer is allowed to observe the prices of its competitors without contacting them and to align their own competitive behavior accordingly.<\/p>\n
##Different types of pricing algorithms
\nWhen using price algorithms, the boundaries can be fluid. When using dynamic price management, the respective company no longer makes its own decisions. The own price adjustment takes place at least partially automatically under the analysis of large amounts of data. The price algorithm used in each case reacts to the other price algorithms used by the other companies. It is even possible that several companies use the same price algorithm within the framework of standard software. The contact can not only be made through personal contact, but also tacitly through software and price algorithms. The fact that in the case of price algorithms there is no "mastermind" in the respective company who is aware of this contact is irrelevant. The mutual reaction of the price algorithms to each other or the use of the same price algorithm can be sufficient for coordinated behavior. <\/p>\n
##Proof of "contact"
\nAt present, it is still widely read that it is extremely difficult to prove that and how price algorithms react to each other and to what extent the necessary "contact" between companies can be seen in this. However, the technical possibilities of the antitrust authorities to prove the use and effectiveness of the price algorithm in specific individual cases are also increasing. However, all companies that use such pricing algorithms must be aware that their competitor is using the same or similar software for price adjustment and this can lead to cartel formation in the legal sense. This applies in particular if it is known in an industry which company uses which price algorithms. In July 2017, the European Commission launched a public tender for a more detailed investigation into the use of pricing algorithms. The antitrust authorities will therefore examine the use of price algorithms more closely in the future. <\/p>\n