Serbian Anti-Trust Commission fines two coffee producers and conditionally approves their merger

On 29 February 2024, the Serbian competition authority published its decision dated 29 December 2023 imposing fines on Croatian Atlantic Group and its two Serbian subsidiaries (“Atlantic“), and Serbian coffee producer Strauss Adriatic (“Strauss“), for concerting to exchange commercially sensitive information on wholesale prices of ground coffee. Atlantic Group and the subsidiaries were jointly and severally fined with approximately EUR 1.6 million (approx. 0.9% of the relevant turnover) while Strauss was fined with approximately EUR 400,000 (approx. 1% of the relevant turnover). This is the first decision of the Serbian Commission that establishes a single and continuous infringement in the form of concerted practice.

On the same day when it published the infringement decision, the Serbian authority published its decision rendered the day before, conditionally approving the merger between Atlantic Group’s Serbian subsidiary and Strauss (to be discussed in a separate blog post).

The infringement decision gives valuable insight into how the Commission creates its initial suspicion of infringement leading to a decision to launch an investigation, as well as into the methodology the Commission uses to draw conclusions from the facts established during the investigation. The decision generously quotes from the parties’ internal correspondence and public statements, amounting to a do’s and don’ts guide for market participants.

What led the Serbian competition watchdog to suspect the infringement?

The investigation was launched in September 2021 following a dawn raid performed on both parties. The Commission built up suspicion of infringement based on the insights obtained while analysing of the retail sector concerning certain food and beverage products, including ground coffee. As part of the sector inquiry, the Commission looked at the retail prices of ground coffee in Serbia in the period between 2015 and 2020 and compared their movement with the movement of commodity exchange price of raw coffee. According to the public statement of Strauss’ representative, the price of raw coffee participates in the wholesale price with 90%.

During the exercise, the Commission noticed that the retail prices of Atlantic’s and Strauss’ comparable coffee brands simultaneously went up at the beginning of 2017 and remained in correlation throughout the analysed period, even though the price of raw coffee continuously went down during the same period. The authority further took note of the overlapping public statements made by Atlantic’s and Strauss’ respective representatives in the period 2017 until July 2021, maintaining that the prices would not change. The tone changed in July 2021 when the representatives of both market players consecutively announced in public their intention to increase prices. The Commission compared Strauss’ publicly available financial statements for 2015 and 2016, which referred to limited opportunity for price increase in Serbia due to fierce competition and consumers’ preference for cheaper brands, with the 2017 financial statements which recorded significant growth of turnover as a result of price increase and not larger market share.

This, coupled with the fact that the wholesale market of ground coffee in Serbia is oligopolistic (according to the Commission, Atlantic and Strauss together hold 80% of the market), and as such conducive to collusion, led the Commission to suspect that Atlantic and Strauss coordinated their wholesale price of ground coffee.

Finding of concerted practice

The Commission established concerted practice created and maintained both through indirect means (signaling via media outlets and getting hold of each other’s wholesale price lists ahead of their implementation by the retailers) and through direct information exchange.  The Commission also found the parties’ coordination was supported by retail price maintenance arrangements, but it did not bring separate RPM charges.

It reiterated the conditions for the finding of the concerted practice developed under the EU law: (i) direct or indirect contact between firms the intention or effect of which is either to influence the conduct of the market or to disclose intended future behaviour to competitors; (ii) conduct on the market in line with the concert; and (iii) a causal relationship between the two.

Contacts and market conduct

Period of intense competition 2012-2014

Based on an inspection of the parties’ internal e-mails captured during the dawn raids, the Commission established that the parties perceived the period between 2012 and 2014 as a period of intense competition and even a “price war”, which they both considered undesirable.

The beginning of price coordination (2014-2015)

The Commission established based on inspected internal e-mail correspondence that the parties’ representatives met in September 2014 to discuss the price increase. The parties continued communication on imminent wholesale price increases at the beginning of 2015. At that time, they came into possession of each other’s wholesale price lists before those price lists were actually implemented, which allowed them to align. According to the Commission, from this moment until July 2021, the wholesale prices of the parties’ brand pairs were almost identical.

In addition, following the adoption of new wholesale price lists, each party notified identical recommended retail prices to the retailers and those recommendations were implemented. In an obiter, the Commission qualifies this as RPM aimed to support horizontal price coordination at the wholesale level.

Despite the foregoing evidence of collusion, the entire 2015 was in fact marked by fierce competition between the two sets of brands through various discounts organized at the retail level.

Exchange of information on capacity utilization (2016)

One of the e-mails dating from 2016 reveals that, as part of the discussions of a potential joint venture that was supposed to last between three and five years, the parties exchanged information on the utilization of their respective production facilities in Serbia. The Commission observed in the decision that the discussions of a potential transaction are not illegal per se but could be illegal if used as a framework for the exchange of commercially sensitive information. In this particular case, it was improper for the parties to exchange commercially sensitive information on utilization even before they commenced negotiations of the transaction and “before they were certain that the communication leads to a potential transaction”.

Price increase and subsequent period of stability (2017-2021)

Towards the end of 2016, both parties individually contemplated price increases of about 10% because the price of raw coffee had gone up. In December 2016, Strauss’ managing director publicly announced his company would increase prices. In January 2017, Strauss came into possession of Atlantic’s wholesale price list to be valid from 1 February, which provided for an average price increase of 11.5%. Around the same time, Atlantic sent an e-mail to Strauss with a table overview of retail prices of Strauss’ brands, whereupon Struss internally commented that those discounted prices should no longer apply. It ended up aligning its new price list to Atlantic’s price list it had obtained and increasing its prices by 11.2% on average, and not 10% as initially contemplated.

According to the Commission, in the ensuing period the parties remained aware of price stability and designed their relationships with retailers with a view to maintaining retail price parity of their comparable brands.

In July 2021, Strauss’s representative gave three consecutive statements in the press announcing the price increase. After the third statement was given, Atlantic’s representative announced a price increase. The Commission assigned particular weight to the fact that Atlantic publicly announced its price increase in the form of a paid simulated interview with its representative and qualified this reaction as an affirmative response to Strauss’ invitation to a price increase. Subsequently, parties came into possession of each other’s new price lists prior to their implementation.

Causality between coordination and conduct on the market

The Commission used correlation analysis to support its conclusion on collusion. It established a high positive correlation between the prices of raw coffee at commodity exchange and prices paid for imports of raw coffee (on the basis of 65% robusta and 35% arabica mix) and a strong correlation coefficient of almost 1 between import prices paid by Atlantic and import prices paid by Strauss.

On the other hand, at the wholesale level where strategic commercial decisions of the parties are taken, the prices did not correlate to the prices of raw coffee as the most important input cost. This led the Commission to conclude that the parties did not make strategic decisions on the wholesale prices on the basis of the prices of the most important raw material. Because the cost of raw coffee contributes to the wholesale price of ground coffee by 90%, and because the investigation was not directed at establishing an agreement to fix prices at a particular level, the Commission excluded the cost of other inputs from the analysis. In light of the fact that the wholesale prices were not in trend with the prices of raw coffee and that the parties, following the information exchange, adjusted upwards their initial internal price plans, the Commission rejected parties’ submissions that their wholesale prices were formulated on a cost basis.

A positive correlation was also observed between wholesale and regular retail prices in Serbia, meaning that price increases at the wholesale level spilled over to the retail level. Another observation made by the Commission is that prices of individual coffee brands at different retailers were almost identical during the relevant period, reflecting a lack of competition at the retail level.

The Commission rejected Strauss’ defense that, as a price follower, it was merely intelligently adjusting its prices to those of the price leader. The defense failed due to the evidence of the exchange of commercially sensitive information and the possession of, and alignment with, the competitor’s price list before the latter became public. The Commission noted that normal market conduct would not have permitted Strauss, a company operating within a strong hierarchical group system, to adjust its price list to the competitor’s within hours of receiving the latter.

The fact that Strauss, whose capacities were under-utilized, did not choose to increase its market share by increasing production at the time when the price of raw coffee went down, also led the Commission to conclude that Strauss chose price stability in concert with the competitor over competition.

Single continuous infringement

The Commission found that the infringement was single and continuous, lasting from 16 September 2014 until 14 September 2021 as the date of initiation of the proceedings. On that basis, it rejected parties’ objections that seizure of correspondence from 2014 and 2015 fell outside the five-year statute of limitations.

The authority confirmed the elements of single and continuous infringement formulated in the EU case law: (i) the existence of an overall plan pursuing a common objective; (ii) the intentional contribution of the undertaking to that plan, and (iii) awareness of the offending conduct of the other participants.


Atlantic/Strauss infringement decision is significant because:

  • it reiterates the willingness of the Serbian Commission to interpret the national competition law in line with EU-developed substantive law when assessing competition law infringements;
  • it highlights the significance of sectoral analysis as the initial source of indications of potential competition law infringements;
  • it emphasizes the importance of dawn raid for evidence gathering;
  • it confirms the authority’s capacity to conduct complex investigations of long-term conduct and produce a mature and comprehensive analysis.