On June 25, 2018, the US Supreme Court rendered its final decision with respect to the claims that American Express ("AMEX") violated Article 1 of the Sherman Act via anti-steering provisions in its agreements with merchants. The decision contains significant insights as to how relevant markets should be defined for "transaction platforms" and how this relevant market definition determines the way in which an effect-based competition law analysis should be conducted.

Before moving on with the competition law related issues, background information regarding transaction platforms in general, AMEX's business model and the anti-steering provisions used by AMEX that constitute the subject of Supreme Court's assessments will be provided.

Transaction Platforms: Transaction platforms provide services to two different customer groups that depend on these platforms to intermediate between them and the value of these platforms to one group depends on how many members of the other group participate. Therefore, such platforms must take into consideration the effects of their pricing decisions on both consumer groups and design their pricing structure accordingly. The distinctive feature of transaction platforms is that such platforms must simultaneously serve to both customer groups as the service that they provide is the "transaction" itself and this service may not be provided to different customer groups separately. To exemplify, AMEX (and other players in the credit card market) has two customer groups which are merchants and cardholders. The main function of AMEX is to ensure that AMEX cardholders can realize transactions with merchants that accepts AMEX credit cards. Both customer groups derive more benefit from AMEX as the amount of transactions increase and the amount of transactions depend on the number of customers in each customer group. Moreover, the number of customers in one group depends on the number of customers in the other group (i.e. more cardholders would prefer AMEX if it is accepted by a greater number of merchants and vice-versa).

AMEX's business model: The success of AMEX's business model depends on the amount cardholders spend and AMEX provides better rewards and advantages in order to promote more spending. To finance these better rewards and advantages, AMEX charges higher (when compared to competitors) merchant fees. Therefore, some merchants may have an incentive to dissuade cardholders from using AMEX.

Anti-steering provisions: Anti-steering provisions refer to provisions in the agreements between AMEX and merchants (these constitute vertical agreements from a competition law perspective) which prohibit merchants from dissuading the cardholders from using AMEX and encouraging them to use other credit cards (e.g. Mastercard and Visa).

The complaint was that anti-steering requirements imposed by AMEX restricted competition and led to higher merchant fees. At the core of the compliant, lied the assumption that there were two separate markets in which AMEX operates - one for merchants and one for cardholders and that these two markets should be assessed separately. Although this assumption was accepted by the District Court, the Supreme Court rejected it and held that there is only one market in which AMEX operated and this was the market for "transactions".

Supreme Court's market definition created a fundamental change in the way in which the competitive effects of anti-steering requirements should be assessed. Whereas in the presence of two markets, the claimants could argue that increased merchant fees constitute proof of anti-competitive effects in and of themselves, this was no longer possible in the presence of a single market. The Supreme Court also provided detailed explanations concerning the pricing strategies of two-sided platforms clarifying that "the fact that two-sided platforms charge one side a price that is below or above cost reflects differences in the two sides' demand elasticity, not market power or anticompetitive pricing".

The Supreme Court pointed out that the claimants had the duty to prove the anti-competitive effects in the "market for transactions" by showing that the alleged anti-competitive conduct led to a decrease in output (i.e. number of transactions) and/or a restriction of competition between credit card companies. The Supreme Court held that the claimants failed to show any anti-competitive effects of the anti-steering provisions in the transactions market and stated that there was evidence on the contrary showing that output in the market has expanded and the overall quality improved during the time when anti-steering provisions were in force (the Court stressed that high merchant fees allowed AMEX to provide better services to the cardholders and to maintain its business model). Hence the Supreme Court reversed the judgment of the District Court and concluded that anti-steering provisions in AMEX's agreements with the merchants were not anti-competitive.

The reasoning of the Supreme Court may have some implications for the ongoing Sahibinden.com investigation of the Turkish Competition Authority ("TCA").

Sahibinden.com is a two-sided platform that serves both to the customers that desire to purchase or rent certain products (buyer group) and to those who want to sell or rent their own -or third parties'- products (seller group). Sahibinden.com adopts a pricing model where the services offered to the buyer group are completely free whereas the sellers pay a certain fee for listing their products on the platform (there are various different methods for calculating this fee). Unlike AMEX, Sahibinden.com does not charge a commission over each transaction. However, the value of the platform is exclusively dependent upon its ability to match the seller group and buyer group and the benefit each group receives from using the platform is directly proportionate with the number of the other customer group. Thus, Sahibinden.com must consider both customer groups when designing the most efficient pricing structure. Currently, Sahibinden.com prefers to subsidize the buyer group at the expense of the seller group. Yet, this does not mean that it is not adopting the most efficient pricing structure for both customer groups as this may well be maximizing the output by creating the optimum buyer-seller balance (i.e. the number of transactions).

Within the scope of its investigation, TCA is examining whether Sahibinden.com (which is allegedly in a dominant position in the markets for "online automotive listings" and "online real estate listings") is charging excessive listing fees and exploit its customers in the seller group.

Although Sahibinden.com investigation concerns an abuse of dominance claim and therefore differs from the AMEX case, the effect-based analysis to be conducted may indeed be quite similar in both cases as both are related with the effects of certain strategies adopted by two-sided platform operators. In light of the reasoning of the Supreme Court, the TCA needs to consider whether (i) it should solely focus on the services Sahibinden.com provides to the seller group or (ii) it should regard the services provided to both customer groups as a whole and conduct its assessment accordingly. The effect-based analysis to be conducted under each scenario would be materially different. If the former approach is adopted, the TCA would need to rely on a classical excessive pricing tests that examine the relationship between the costs of a service and its price and the comparison of the dominant undertaking's price with that of its competitors. Under the latter approach, which was developed by the Supreme Court, the TCA would need to assess whether the output (e.g. the amount of transactions) is reduced or the competition between platforms is restricted because of the alleged excessive pricing.

Up until now, the TCA did not assess the alleged anti-competitive effects of pricing strategies of a two-sided online platform and the Sahibinden.com investigation will be a first. It would be interesting to see if the TCA would follow the lead of the Supreme Court and develop a different test due to the characteristics of the market or it would simply apply the test used in the "classical industries" to two-sided platforms without a major modification.

The content of this article is intended to provide a general guide to the subject matter. Specialist advice should be sought about your specific circumstances.