It seems undeniable that within the EU interest is growing in purchasing alliances and particularly in the grocery retail sector. However, care must be taken to avoid drifting into a buyer cartel. The decisional practice of the EC and NCAs shows a clear distinction between buyer cartels and legitimate joint purchasing agreements. On the one hand, buyer cartels are viewed as restriction of competition by object and are per se prohibited. As a result, when such infringements are assessed, hefty fines are usually imposed. On the other hand, investigations into legitimate purchasing alliances, which are subject to an effects-based analysis, are rarely subject to fines, but rather, are often resolved by commitments. Only where joint purchasing agreements are used as a trojan horse for buyer cartels are they assessed more harshly.
By Liliane Gam1
Competition policy and enforcement action against collusion in Europe has largely focused on horizontal coordination among suppliers. However, in recent years there has also been a heightened interest in joint purchasing agreements or purchasing alliances by competition authorities, with increased competition enforcement, especially in the retail sector.
Unlike buyer cartels, which are regarded as having as their object the restriction of competition and are thus prohibited by competition law, joint purchasing agreements are usually unlikely to have as their object the restriction of competition since it is primarily a means of increasing bargaining power. Such agreements require an effects-based analysis.
This paper provides an introduction to the sharpened focus on purchasing alliances in European competition policy (Section II), followed by an overview of the legal framework for the analysis of purchasing alliance and their distinction from buyer cartels (Section III), as well as a summary of the recent enforcement practice of NCAs and the EC (Section IV). Finally, some key considerations to take account of when setting up a purchasing alliance are provided (Section V).
II. EUROPEAN POLICY ACTION CONCERNING PURCHASING ALLIANCES
There is a growing sense in European competition policy that more attention needs to be placed on joint purchasing agreements. Last decade, the European Commission (“EC”) and National Competition Authorities (“NCAs”) began to take a closer interest in joint purchasing agreements in the course of their retail market monitoring work. A 2010 report2 notes that while joint purchasing agreements are normally pro-competitive, when concluded among market players to achieve volumes and discounts similar to their bigger competitors, large transactional purchasing alliances may in certain circumstances raise concerns as to their effect on competition and ultimately, on consumers, especially where they “reduce the participants’ incentives to expand into each other’s domestic markets or may contribute to a standardisation of their purchasing policies, which could have a negative longer term impact on product variety and/or the ability of food suppliers to innovate.”3
A couple of years later, the Food Subgroup of the European Competition Network (“ECN”) published a report which pointed to a growing phenomenon of international purchasing, which needed to be monitored.4 This may be, among others, one of the reasons why the EC subsequently created a Food Task Force within DG COMP.5 The growing concerns towards purchasing alliances was also evident in a number of sector inquiries conducted by NCAs.
For instance, in 2014, the German Federal Cartel Office (“FCO”) concluded a sector inquiry into the food sector where it considered purchasing alliances alongside concerns about increasing concentration and complaints about buyer power. It noted that purchasing alliances involving large retailers should be carefully examined as they could be regarded as a pre-stage of a merger.6
At the same time in France, a significant number of purchasing alliances were concluded, which prompted a request from the Government for an opinion from the French Competition Authority (“FCA”) on the impact of such alliances on competition. In its opinion,7 the FCA expressed concerns that they may give rise to competition risks on both the upstream and downstream markets. The FCA recommended the introduction of a pre-closing filing obligation. Since then, any purchasing agreement in the retail sector must be communicated to the FCA if the turnover exceeds certain thresholds.8 In addition, the FCA may carry out a competitive assessment of the agreement’s implementation.9 If a buying agreement raises competition concerns, the parties must submit commitments.
This context of increased focus on purchasing alliances in the retail sector resulted in several investigations by NCAs. While there have also been a number of investigations in relation to traditional buyer cartels, the investigations into purchasing alliances show that the frontier between legitimate purchasing alliances and buying cartels may not be easy to draw. As Margrethe Vestager stated recently, it is important to “keep a careful eye on how such alliances work in practice, in order to make sure they are not a cover for cartel behaviour.”10
Finally, in EU competition law, purchasing alliances are assessed under the framework of the EC’s Horizontal Guidelines,11 which are currently subject to reform. The EC recently published a staff working document on this topic, indicating that further clarification was required regarding the distinction between buyer cartels and joint purchasing agreements. This seems to confirm that provisions laid down in the Horizontal Guidelines regarding joint purchasing agreements may no longer be regarded to be adapted to market developments. An overhaul of these provisions is thus to be expected in the upcoming months.
III. FRAMEWORK OF ANALYSIS FOR PURCHASING ALLIANCES
Article 101(1) of the Treaty on the Functioning of the European Union (“TFEU”) prohibits agreements or concerted practices between undertakings which have as their ‘object’ or ‘effect’ an appreciable prevention, restriction, or distortion of competition within the internal market and may affect trade between Member States, unless they fall within an excluded category or are otherwise exemptible in accordance with Article 101(3) of the TFEU.
As stated above, the Horizontal Guidelines provide a framework for the assessment under Article 101 TFEU of purchasing alliances. The latter can contravene competition law by (i) constituting an object restriction or (ii) having the effect of restricting competition. These Horizontal Guidelines are a (non-binding) soft law tool.
A. Restrictions of Competition by Object
A joint purchasing arrangement is unlikely to have the ‘object’ of restricting competition. Only where such a purchasing arrangement facilitates a cartel among its participants will there be an object restriction.12 An object restriction of this sort might take the form of output limitation (demand withholding), market sharing/allocation (e.g. through purchasing quotas), price fixing or ‘profit allocation’ with upstream suppliers.
Where there is no obvious anticompetitive object, it is necessary to consider whether the arrangement has an appreciable effect on competition.
B. Restriction of Competition by Effect
As to the legal assessment, the EC will analyze two markets regarding joint purchasing: the purchasing market (upstream), where the parties jointly purchase their supplies, and the selling market (downstream), where the parties sell individually their respective products.
Joint purchasing arrangements are unlikely to fall under Article 101(1) TFEU where the parties – due to lack of market power – would not be able to buy certain supplies without the agreement.15 This may be the case, for example, where a supplier will not deliver goods below a certain quantity threshold.
C. Safe Harbor
The Horizontal Guidelines further define a general safe harbor for agreements that have a market share of at most 15 percent on both the upstream16 (i.e. their purchases account for less than 15 percent of total purchases on that market) and the downstream markets. Those agreements are said to be unlikely to have restrictive effects on competition. One exception to this would be where there are parallel networks of similar joint purchasing arrangements that could dampen competition.
D. Analysis Beyond the Safe Harbor
A more detailed assessment needs to be conducted to assess the risks, taking into account factors such as: (i) the parties’ market shares in both the purchasing and selling markets; (ii) the level of market concentration (market shares of the parties’ competitors in both markets); (iii) whether the sellers have countervailing power; (iv) whether the parties have similar cost structures; or (v) whether there is information exchange – or measures limiting it – which could lead to a collusive outcome.
The Horizontal Guidelines note that from an upstream perspective, anticompetitive buying power is likely to arise if an agreement accounts for a sufficiently large proportion of the total volume of a purchasing market so that access to the market may be foreclosed to competing purchasers.18
From a downstream perspective, a high commonality of costs may facilitate the coordination of behavior. The Horizontal Guidelines also note that information exchange may give rise to concerns, but that if the information exchanged does not exceed the sharing of data necessary for purchasing then the agreement would be more likely to meet the criteria of Article 101(3) TFEU.19
- Individual exemptions
If a joint purchasing arrangement infringes Article 101(1) TFEU, it may nevertheless be exemptible under Article 101(3) TFEU provided that it meets the four cumulative conditions of that paragraph, namely (i) efficiency gains; (ii) indispensability; (iii) pass-on to consumers; (iv) no elimination of competition. Regarding purchasing agreements, these conditions are detailed in paragraphs 217-220 Horizontal Guidelines.
Joint purchasing arrangements can give rise to efficiency gains. In particular, they can lead to discounts from volume bundling; increased supply chain efficiency (e.g. reduced transport costs); economies of scale (e.g. centralized ordering); transaction cost savings (e.g. negotiation costs); and counteracting significant supplier power in the selling market,20 provided that the cost savings are passed on to the purchasers’ customers; and any restrictions on competition are essential to create the cost savings or economic benefits.
IV. RECENT ENFORCEMENT PRACTICE FOR PURCHASING ALLIANCES
In the past ten years, enforcement activity has been increased in relation to purchasing alliance and buyer cartels. While the EC’s practice has so far mostly focused on buyer cartels, at NCA level the focus has been on purchasing alliances. Below we first take a look at the rather classic buyer cartels at both EC and Member State level (IV.A) and then at investigations into purchasers alliances between retailers (IV.B).
A. Enforcement Action Involving Purchasing Cartels
- Billa/Julius Meinl (Czech Republic)
On October 13, 2003, the Czech Republic Competition Authority imposed fines totaling CZK 51 million on the retail chains Billa and Julius Meinl for a price-fixing cartel in the purchase of goods.21
In the context of their cooperation, parties exchanged information about their respective purchase prices and bonus and discount systems, compared them, and required suppliers to provide them with the same financial terms (under the threat of de-listing). The entire agreement was prohibited.
- Raw Tobacco (EC)
On October 20, 2005, the EC imposed fines totaling EUR 56 million on four tobacco processors for colluding on the prices paid to tobacco growers or intermediaries and on the allocation of suppliers.22 The EC also imposed fines on the Italian trade associations of processors and tobacco growers for engaging in collective price negotiations.
Parties breached Article 101(1) TFEU by exchanging sensitive information, coordinating bids for public auctions or prices to be paid to growers, and allocating suppliers (growers and intermediaries) and quantities among themselves.
- Car Battery Recycling (EC)
In this case, communications relating to purchase prices showed the parties’ intention to influence purchase prices. The aim was to undercut the amount they owed for used batteries. This collusion took the form of frequent discussions between undertakings. Moreover, the undertakings were well aware of the illegal character of their contacts as they used a coded language (e.g. weather conditions).
- Ethylene (EC)
The most recent decision issued by the EC in relation to a buyer cartel relates to the so-called ethylene cartel.25 On July 14, 2020, the EC imposed fines totaling EUR 260 million on three undertakings for taking part in a price-fixing cartel for the purchase of Ethylene.
Companies buying ethylene usually do so under supply agreements. However, given that the ethylene market is characterized by a highly volatile price, a formula is used to determine the price on a monthly basis. The monthly contract price results from individual negotiations between buyers and suppliers of ethylene. The investigation revealed that these undertakings coordinated their negotiation strategies with suppliers in order to influence the price to their advantage. The coordination involved the exchange of a great deal of sensitive information.26
- Steel (Germany)
On November 21, 2019, the FCO imposed fines totaling around EUR 100 million on car manufacturers for taking part in a price-fixing cartel for the purchase of steel products.27
These German car manufacturers, which met twice a year with steel manufacturers, forging companies and large systems suppliers under the umbrella of the “German association for steel and metal processing,” exchanged information on uniform surcharges for the purchase of long steel products. Such exchange could lead to price collusion.
B. Enforcement Action Involving Purchasing Alliances
- P&H/Makro (UK)
On April 27, 2010, the UK Office of Fair Trading (“OFT”) issued an opinion in relation to a proposed joint purchasing agreement between Makro and Palmer & Harvey.
The OFT concluded in general that, in the absence of parallel networks of similar agreements in the relevant market, joint purchasing agreements were unlikely to cause harm where the parties had no downstream market power. It thus considered that the arrangement between Makro and P&H was unlikely to reduce competition as the companies would not have the power to raise prices/reduce output in the downstream markets.
In its analysis, the OFT identified a concern that certain exchanges of information between the companies could potentially lead to a reduction in competition. However, the parties subsequently agreed to ensure the data they supplied to each other was general and aggregated, preventing either company from extrapolating specific or sensitive information. The OFT opinion did not consider the upstream market in any detail.
- Centrale Italiana (Italy)
On September 17, 2014, the Italian Competition Authority (“ICA”) closed an investigation against a joint purchasing organization (Centrale Italiana), and five retailers.28
The parties involved in the alliance had a market share of 23 percent on Italy’s upstream procurement market and around 40 percent on the downstream retail markets in some local areas.
The ICA considered that upstream, the alliance could reduce the possibility of suppliers to compete or decrease the variety and quality of products. At downstream level, the ICA considered that the arrangement could reduce the incentive of the parties to compete – as alliance members were aware of the turnover achieved by their competitors within the alliance, which could lead to the coordination of their sales policies.
The ICA closed its investigation given that the parties agreed and committed to put an end to their alliance or any form of commercial cooperation.
- Retail Trade Group (Germany)
On April 4, 2017, the FCO indicated that it had no objection to the creation of a joint purchasing joint venture (Retail Trade Group) between four (small) food retailers.29
While noting that the cooperation could have effects on the market, the considered effects on downstream retail markets would be limited given that the four major retailers dominated more than 85 percent of the market. As a result, the cooperation ensured the competitiveness, and ultimately the independence of the smaller retailers. The latter was necessary in order to provide alternatives and variety to consumers. Moreover, upstream, the effects were also limited given that the combined shares of the RTG members amounted to less than 15 percent – despite downstream market shares above in some territories – in all product categories and were thus under the safe harbor.
The FCO therefore decided not to investigate but indicated that a probe under competition law might be carried out later if the cooperation was extended.
- Casino/Intermarché (EC)
On November 4, 2019, the EC opened a formal antitrust investigation into a joint purchasing joint venture between Casino and Intermarché.30
This investigation differs from other EC precedents, which concerned full-on buyer cartels, whereas here, a joint purchasing arrangement (more particularly, a joint purchasing joint venture) is at the root of the investigation. While a separate entity usually adds some layer of protection, the EC is investigating whether this separate entity exceeded its initial framework and served as a vehicle to coordinate their behavior downstream in relation to the development of shop networks or pricing policy. The investigation is still ongoing.
- Casino/Auchan/Metro/Schiever (France)
On October 22, 2020, the FCA rendered binding the commitments submitted by retailers for their purchasing alliance to alleviate the competition concerns raised by such alliance.31
The alliance at issue was organized through a separate legal entity that organized tenders for a range of products. The FCA considered that there was a risk of foreclosure of suppliers, especially SMEs and very small businesses. Moreover, if a cut in margins to suppliers were to occur, it could reduce their ability to invest and innovate, and finally their incentive to stay in the market. The FCA took into account: (i) the suppliers’ limited market power (e.g. short-term contracts); (ii) the market’s low profitability and decreasing volumes over the years; and (iii) the fact that a significant share of the supply of retailers’ own-brand labels was produced by SMEs, which were more susceptible to sudden changes in marketing conditions.
To alleviate these concerns, the parties agreed to exclude certain categories of products from struggling sectors and limit the scope for other products to 15 percent of all purchases.
- Carrefour/Tesco (France)
On December 17, 2020, the FCA rendered binding the commitments submitted by retailers for their purchasing alliance to alleviate the competition concerns.32
The FCA identified a range of possible competition effects similar to those identified for the Casino/Auchan/Metro/Schiever alliance. The commitments are therefore very similar to the abovementioned case. The parties agreed to (i) exclude certain agricultural products; (ii) limit their cooperation for other product families to 15 percent of all purchases; and (iii) ensure SMEs the possibility of bidding for tenders launched by retailers.
- Carrefour/Provera (Belgium)
On April 28, 2021, the Belgian Competition Authority (“BCA”) rendered binding the commitments submitted by retailers for their purchasing alliance.33
As part of this alliance, Carrefour agreed to negotiate the purchase of product on behalf of both retailers.
The BCA considered that the safeguards put in place were not sufficient to prevent the exchange of sensitive information:
- Commercial negotiations within the alliance were conducted by the same employees as were also in charge of the negotiations of Carrefour’s separate purchases. The BCA took the view that a non-disclosure agreement, a separate e-mail address or committing not to use the information was not sufficient to prevent the exchange of information;
- Provera shared future purchases information with Carrefour;
- In order to negotiate the discounts to be granted by the suppliers in exchange of promotional services to be provided by Carrefour and Provera in their respective stores, the parties used a so-called “scorecard” system. While the parties negotiated the details for the compensation separately, the BCA considered that through the scorecard system, they obtained information that they would not have obtained otherwise, and which could contribute to the coordination of their commercial policies; and
- At management level, there was a risk of information exchange given that they were informed of the negotiations. Even though the cooperation provided that only the necessary information was to be shared, there was also a risk of coordination.
In order to address competition concerns, the parties agreed to create a separate legal entity for the conducting of their purchasing operations and to implement stricter control over the persons with access to information. Parties also committed to ensure that negotiations would henceforward be restricted to the bare financial aspect, allowing each party to define its business strategy in complete independence.
The decisional practice of the EC and NCAs shows a distinction between buyer cartels and legitimate joint purchasing agreements. On the one hand, buyer cartels are viewed as restriction of competition by object and are per se prohibited. As a result, hefty fines34 are usually imposed, even where the parties cooperate with the investigation. On the other hand, legitimate purchasing alliances, which are subject to an effects-based analysis, are rarely subject to fines. Either the cooperation is terminated (Centrale Italiana), or the parties to the cooperation make some changes to it in order to alleviate the competition concerns (e.g. Carrefour/Provera). The only exceptions where fines are imposed is where purchasing alliances are used as a vehicle for covert cartels (Billa/Julius Meinl).
V. KEY TAKEAWAYS WHEN SETTING UP PURCHASING ALLIANCES
It is important to ensure when setting up a purchasing alliance that there is no restriction of competition by object and that the effects on competition are limited to what is strictly necessary to achieve the efficiencies resulting from such cooperation. Great care must indeed be taken to ensure that a cooperation does not become a vehicle for the exchange of sensitive information or coordination of behavior.
In order to reduce the risks, parties to a purchasing alliance may consider the following:
- The parties should make sure that the arrangement falls within the 15 percent safe harbor. In this regard, it should be ensured that the 15 percent market share threshold is not exceeded even under a narrow definition of the market.
- The parties should be able to purchase products outside the arrangement. In addition, the purchasing arrangement should not prevent the parties from competing with each other in the downstream market.
- The parties should consider: (i) limiting the exchange of commercially sensitive information to what is necessary to the joint purchasing activities; (ii) implementing “clean teams” to handle the joint purchasing and information; (iii) carefully managing documents, in particular those describing the purposes of the transaction. However, the safest way might be to organize the joint purchasing arrangement through a separate legal entity, with information barriers in place.
It remains to be seen what changes will be made to the Horizontal Guidelines as these will ultimately determine the framework of assessment.
1 The author is a Managing Associate at Linklaters LLP in Brussels. The views expressed here are the author’s alone and should not be attributed to Linklaters LLP. The author wishes to thank Jonas Koponen (Partner, Linklaters LLP, Brussels) and Simon Graff (Legal Intern, Linklaters LLP, Brussels) for their valuable contribution.
2 Commission SWD on Retail Services in the Internal Market – Accompanying document to the Report on Retail Market Monitoring: “Towards more efficient and fairer retail services in the Internal Market for 2020,” July 5, 2010, SEC(2010)807.
3 Ibid. p. 47.
4 Report on competition law enforcement and market monitoring activities by European competition authorities in the food sector, May 2012.
5 Speech of Joaquín Almunia VP of the EC responsible for Competition Policy, “Competition policy as a pan-European effort,” Nicosia, October 2, 2012.
6 Bundeskartellamt, Summary of the Final Report of the Sector Inquiry into the food retail sector, September 24, 2014, p. 16.
7 Avis 15-A-06 du 31 mars 2015, relatif au rapprochement des centrales d’achat et de référencement dans le secteur de la grande distribution.
8 Loi Macron: “LOI n° 2015-990 du 6 août 2015.” See also Article L.462-10 of the French Commercial Code.
9 Loi Egalim: “LOI n° 2018-938 du 30 octobre 2018.” See also Article L.462-10 of the French Commercial Code.
10 EuroCommerce Policy Talk Webinar, Keynote speech, March 12, 2021.
11 Guidelines on the applicability of Article 101 of the Treaty on the Functioning of the European Union to horizontal co-operation agreements (2011/C 11/01) (“Horizontal Guidelines”).
12 Horizontal Guidelines, para. 205.
13 For example, Judgment of September 24, 2019, HSBC Holdings and Others v. Commission, T-105/17, EU:T:2019:675, para. 64.
14 Horizontal Guidelines, para. 206.
15 By analogy with Horizontal Guidelines, para. 237.
16 That is the market from the supplier’s perspective (i.e. if there were a 5-10 percent increase in price, what goods or services could the supplier switch to).
17 Horizontal Guidelines, para. 208.
18 Horizontal Guidelines, para. 210.
19 Horizontal Guidelines, para. 216.
20 Judgment of December 15, 1994, Gottrup-Klim and Others Grovvareforeninger v. Dansk Landbrugs Grovvareselskab, C-250/92, EU:C:1994:413, para. 32.
21 Press release, Cartel Agreement Billa/Julius Meinl, September 21, 2006.
22 Commission decision of October 20, 2005 (COMP/C.38281.B2) – Raw Tobacco IT.
23 One of the parties successfully appealed on the basis of evidence of a time-limited involvement. See Judgment of November 7, 2019, Campine and Campine Recycling v. Commission, T-240/17, EU:T:2019:778.
24 Commission decision of February 8, 2017 (AT.40018 – Car battery recycling).
25 Commission decision of July 14, 2020 (AT.40410 – ETHYLENE).
26 Ibid. para. 42.
27 Bundeskartellamt 2019 Annual Report, p. 21.
28 Italian NCA, I768, Centrale d’acquisto per la Grande Distribuzione Organizzata.
29 Press release, “Bundeskartellamt currently has no objection to food retail joint venture ‘Retail Trade Group.’”
30 AT. 40466, Alliance Casino & Intermarché.
31 Decision of the French NCA, October 22, 2020, No 20-D-13.
32 Decision of the French NCA, December 17, 2020, No 20-D-22.
33 Decision of the Belgian NCA, April 28, 2021, Case CONC-I/O-19/0013.
34 See Raw tobacco; Car battery recycling; Ethylene; Steel.