On 18 May and 22 June 2022, the EU General Court upheld two European Commission prohibition decisions. Both judgments endorsed the European Commission’s assessment of the mergers at issue, recognising a broad margin of discretion in the European Commission’s evaluation of the concentrations’ effects and the remedies.
Key takeaways
- The EU General Court (“GC”) ruled that the European Commission (“EC”) must only prove that the transaction will likely result (e., balance of probabilities standard) in a significant impediment of effective competition (“SIEC”), rather than showing that there is a strong probability of a SIEC;
- The SIEC can occur only on a part of the relevant market;
- The fact that two competitors are close may contribute to the finding of a SIEC (they do not have to be each other’s closest competitors);
- The EC can rely on a body of evidence to prove a SIEC, including customer responses to market test (which bear particular credibility), and does not have to do complex economic studies (such as the SSNIP test when defining markets);
- The EC enjoys broad discretion in assessing remedies proposed by the parties, and the GC will only verify if the EC committed a manifest error of assessment;
- Commitments need not remove the entire overlap between the merging parties but must be sufficient to remove the SIEC and must be viable. To achieve viability, merging parties may have to include in the divestiture package products beyond the overlapping products;
- Upfront buyer remedies are equivalent to fix-it-first remedies only if they include specific requirements which are sufficiently clear and provide enough certainty to the EC that a suitable buyer will operate the divested business in a viable manner.
Background
Wieland/ARP/Schwermetall – creation of a dominant position
On 13 June 2018, Wieland-Werke AG (“Wieland”) notified the EC of its proposed acquisition of Aurubis Rolled Products (“ARP”) and Aurubis’s 50% stake in Schwermetall (“Schwermetall”), a 50/50 joint venture between Wieland and ARP. Schwermetall manufactures pre-rolled strips which it sells to its parent companies and other downstream manufacturers of rolled products.
In Phase I, the EC found that the transaction raised competition concerns, as it would have allowed Wieland to eliminate the competitive pressure exercised by ARP and become a dominant player in the markets for rolled products in the European Economic Area (“EEA”) (Wieland and ARP were the first- and second-largest suppliers on that market). The EC also found that Wieland’s acquisition of sole control over Schwermetall would result in vertical effects (that would further reinforce alleged horizontal effects), namely through access to sensitive information about Schwermetall’s customers (Wieland’s competitors on the downstream market) and likely price increases in the downstream market.
The EC considered that the commitments proposed by Wieland in Phase I and Phase II were insufficient to resolve these concerns and blocked the Wieland/ARP/Schwermetall transaction on 5 February 2019 (the “Wieland decision”). Wieland appealed the EC decision before the GC.
The Wieland decision was taken a few days after the EC unconditionally cleared a parallel transaction between KME AG and MKM Mansfelder Kupfer und Messing GmbH (“MKM”), the third- and fourth-largest suppliers of rolled products in the EEA. That transaction was notified only 11 days before the proposed Wieland/ARP/Schwermetall merger. The EC applied the priority rule, under which two separate transactions in the same industry are treated on a “first come, first served” basis: the transaction notified first is analysed as if the second transaction will not occur, but the second transaction is analysed as if the first transaction has already closed.
On 18 May 2022 (more than three years after the EC decision was adopted), the GC handed down its judgment, rejecting Wieland’s appeal in its entirety (“Wieland judgment”).
Thyssenkrupp/Tata Steel – “gap case”
On 25 September 2018, Thyssenkrupp and Tata Steel notified the EC of their plan to create a joint venture that would combine their flat carbon steel and electrical steel activities. The EC considered that the transaction, which involved the second and third-largest producers of flat carbon steel in the EEA, raised competition concerns as, although it would not create a dominant company, it would create a market leader and remove an important competitor in some already highly concentrated markets (so-called gap case). [1]
Following an in-depth investigation, the EC prohibited the concentration on 11 June 2019 (“Thyssenkrupp decision”). According to the EC, the transaction would have reduced competition and choice in suppliers, as well as resulted in increased prices for various types of steel. The competitive pressure exerted from third‑country imports and the remedies proposed by the parties were considered insufficient to ensure effective competition post-transaction.
On 22 August 2019, Thyssenkrupp appealed the EC’s decision before the GC. The appeal was rejected by the GC almost three years later, on 22 June 2022 (“Thyssenkrupp judgment”). [2]
While the two cases involved different issues and different theories of harm, they gave the GC the opportunity to discuss fundamental issues in merger control such as the required standard of proof, the assessment of evidence and the assessment of remedies offered by the parties.
Standard of Proof
Likelihood of SIEC: strong probability vs balance of probabilities (more likely than not)
The Significant Impediment of Effective Competition (SIEC) test set forth in the Merger Regulation was introduced to allow the EC to prohibit anti-competitive mergers that could not be dealt with under the “dominance test”, i.e., gap cases. [3] However, almost 20 years after its introduction, the EC’s application of the SIEC test still gives rise to many questions, amongst which is the standard of proof to which the EC is held.
In CK Telecoms, a gap case on which the GC ruled in 2020, the EC had blocked the proposed acquisition of Telefónica UK by Hutchison, considering that the merger would have created a new market leader while removing an important competitor in the UK mobile market. On appeal, however, the GC annulled the EC decision stating inter alia that “the Commission is required to produce sufficient evidence to demonstrate with a strong probability the existence of significant impediments following the concentration”. [4] In particular, the GC found that the necessary standard of proof in a case involving such theories of harm is stricter than the “balance of probabilities” (“more likely than not”) standard.
In the Thyssenkrupp judgment, by contrast, the GC found that the EC must “show with a sufficient degree of probability, […] that the transaction significantly impedes effective competition in the internal market”. [5] In that case, it backed the EC conclusion that the parties’ main competitor (and market leader) would not have had the incentives to intensify competition (e.g., by increasing production) against the merging parties after the transaction and thus compensate for the loss of competition. It concluded that the EC had sufficiently proven the likelihood of a SIEC.
Thus, in two “gap” cases, the GC appeared to have required from the EC two different standards of proof, one of “sufficient degree of probability” and the other of “strong probability”. While it is not entirely clear what “sufficient degree of probability” actually means, e.g., whether this means “balance of probabilities” or something else, and what the difference to the equally unclear “strong probability” standard is, one may assume that the GC meant to deviate from the standard applied in CK Telecoms (“strong probability”) and ease the burden on the EC to prove a SIEC.
The final word on this will rest with the European Court of Justice (“ECJ”). Indeed, that precise point was hotly debated at the hearing before the ECJ in the appeal brought by the EC in the CK Telecoms case on 14 June 2022. The next step of this saga will be the Opinion of Advocate General Kokott, scheduled for 20 October 2022.
SIEC on the whole relevant market or part of it?
In Wieland, the EC defined an overall market for rolled products but found anti-competitive effects on competition only on certain segments of this market (notably high-end or specialised and low-end or commoditised rolled products). For the EC, the different segments required differentiated analyses in light of product performance, customers’ specifications and intensity of competition. [6] Because of the oligopolistic structure and the very high level of concentration in the high-end segment, the EC concluded that the weakening of competition in this segment would affect the competition in the relevant market as a whole. [7] Wieland argued that the prohibition decision was inconsistent as the EC could not limit itself to finding anti-competitive effects only on part of the market.
The GC rejected Wieland’s argument and confirmed that the existence of an overall market does not exclude the possibility of identifying different competitive dynamics in different market segments. For the GC, this is in line with the objective of Article 2 of the Notice on the definition of the relevant market, i.e., market definition is not an end in itself, but just a tool to define the “arena” in which competition takes place and in respect of which market power and a transaction’s effects need to be analysed. Thus, a distinction must be made between the definition of the relevant market, which is “the market on which the notified concentration occurs and which might therefore be affected by it”, and the assessment of the effects of competition on that market. A concentration can impede competition differently in different parts of the relevant market (given that different segments may feature different dynamics, and competitors may act differently in these segments), without calling into question the correctness or appropriateness of the definition of that (differentiated product) market.
Analysing the impact of transactions on various sub-segments of the relevant market has long been the EC’s practice. However, as a result of this ruling by the GC, one may expect that less weight will be given to market definition in itself and, conversely, more weight given to the analysis of the effects of the concentration on individual market segments. An open question, however, is whether anti-competitive effects in a segment of the market only are always “significant” enough to constitute a SIEC.
Closeness of competition: “closest”, “particularly close” or just “close”?
In both judgments, the GC reviewed the question of closeness of competition, i.e., the level of substitutability of the merged parties’ competing products. The higher the degree of substitutability and closeness of competition, the more likely the proposed transaction could eliminate competitive constraints that the merging parties exert on each other. [8]
In the Wieland decision, the EC found that the proposed concentration would eliminate the competitive constraints that ARP exercised on Wieland because ARP was an important and close competitor to Wieland. [9] The EC considered in particular that it only had to establish that the parties were “close competitors”, rather than the “closest competitors”. The GC confirmed that the concept of “close competitor” can be “a central factor in the analysis as to whether the reduction in competitive constraints arising from a concentration could lead to significant price increases on the relevant market”. [10] It concluded that the EC assessment that the parties to the concentration at issue were close competitors was “legally sufficient”. [11]
Here again, one may wonder if the Court departed from CK Telecoms which seemingly demanded a “particularly close competitor”. [12]
Still, there is an important difference between these two cases and the CK Telecoms case, which could explain the divergent approach. In CK Telecoms the EC analysed a market characterised by a high degree of product homogeneity (mobile communications services) where all competitors (mobile operators) are close. Indeed, in CK Telecoms, the EC found that all four UK operators were close, and its analysis focused on the degree of closeness between the merging parties. [13]
By contrast, in Wieland, the market was “highly differentiated and characterised by the existence of different segments driven by specific competitive dynamics”, and the analysis of closeness focussed predominantly on closeness of substitutability between the merging companies’ products. [14] Equally, in the Thyssenkrupp judgment, the GC fully endorsed the EC’s analysis that the parties were close (and not necessarily closest) competitors, as they had (i) similar technical capabilities; (ii) similar focus on specific products or segments; (iii) large sales at common customers; and (iv) geographic proximity. This case provides a useful illustration of qualitative factors which can be relied upon by the EC to show closeness of competition.
Mixed theories of harm
In the Wieland decision, the EC considered that Wieland’s vertical integration post-transaction of Schwermetall (a 50-50 JV with Aurubis over which Wieland acquired sole control as a result of the transaction) would involve the disappearance of Aurubis’ power of veto and Schwermetall’s commercial autonomy. [15] As a result, according to the EC, the transaction would have given the merged entity the opportunity to increase prices in the downstream market where Wieland and ARP competed, and put competitors at a competitive disadvantage (by raising costs of inputs).
Wieland took issue with this “mixed” approach (taking into account horizontal and vertical effects) and argued that the EC should have assessed the vertical effects of the concentration solely in accordance with the Non-Horizontal Merger Guidelines (rather than para. 36 of the Horizontal Merger Guidelines). [16] The GC rejected the argument and determined that the EC was entitled to find that these vertical effects reinforced the horizontal effects, in line with paragraphs 36 of the Horizontal Merger Guidelines and 18 and 78 of the Non-horizontal Merger Guidelines and that therefore it was not required to apply the typical ability-incentives-effects analytical framework that is applied to purely vertical mergers. [17]
In the Thyssenkrupp judgment, the parties argued that the EC used two separate and contradictory theories of harm in relation to certain markets, i.e., the reinforcement of a dominant position and the elimination of significant competitive constraint on oligopolistic markets (“gap case”). The GC’s reasoned that these theories of harm are not mutually exclusive, and that, in any event, the EC found, primarily, that the merger would create a dominant position and, in the alternative, that it would eliminate an important competitive force. [18]
Standard of Evidence
Evidence: consistent, reliable and relevant: not necessarily economic
Both judgments recalled that the EC must conduct its analysis on a body of evidence that should be analysed as a whole, but they also made it clear that the EC enjoys some discretion in the type of evidence it relies upon in any individual case.
For example, confirming its Topps judgment, the GC held that there is no need to conduct a SSNIP test or any other econometric test to assess the substitutability of the different products. [19] Evidence from the market test can be sufficient. While there is no hierarchy of evidence, the EC must base its assessment on a set of “reliable, substantiated, consistent, relevant and sufficient items of evidence” that will be carefully reviewed by the Court. [20] The GC also confirmed that the EC does not need to take into account the economic evidence submitted by the parties where it is “not sufficiently convincing or relevant”. [21] In the same vein, the Wieland judgment supports the EC’s approach to rely largely on the internal documents of the parties, without carrying out extended economic analyses (for example a switching analysis) to prove that a SIEC will likely arise. [22] It is interesting to contrast this position with antitrust cases, especially cases of abuse of a dominant position, in which the GC appears to require more economic rigour from the EC. [23] The difference may be explained by the so-called “need for speed” in merger cases (i.e., the fact that the EC must decide within a short timeframe), underscored several times by the GC in both the Thyssen and Wieland cases, and that Article 102 cases typically relate to past conduct for which effects can be determined, while merger control seeks to predict the impact of a change of market structures.
Internal documents and third party responses
The GC’s judgments confirmed that internal documents have high probative value, especially where they were drawn up by persons directly involved in the parties’ business. [24] Similarly, third parties’ statements are particularly credible, especially where they are supported by other items of evidence. [25] However, the GC noted that in an oligopolistic market, customers’ responses are more credible than those of competitors, as such customers could not benefit directly from the concentration. [26]
Both decisions were challenged based on an alleged breach of the principle of good administration under Article 41 of the EU Charter and the associated duty of the EC to assess matters diligently and impartially in relation to third party responses. [27]
In Wieland, the parties argued that the EC’s questions to third parties were irrelevant, biased or too technical and that the EC was late in consulting market participants and did not investigate sufficiently the third set of commitments proposed. The GC rejected the allegations of bias and lack of relevance; it also indicated that the EC was not bound to conduct a final market test following the last round of commitments submitted. Finally, the GC confirmed that the fact that a large part of customers did not respond to the EC’s questionnaire does not affect the credibility of the received responses.
Thyssenkrupp argued that the EC had made a procedural error by not using its powers to force the market participants to reply to the request for information by issuing a decision pursuant to Article 11(3) of Regulation 1/2003, which would have allowed the EC to impose financial penalties in case of non-response to the RFI. The GC simply observed that this step is incompatible with the strict time limit necessary for the adoption of the final decision. In any event, the GC also pointed out that the rate of responses was sufficient, as it was above 50%, which gives a useful indication of which level of market test response may be sufficient. [28]
Analysis of remedies
The two judgments provide interesting indications of the standard to which the EC is held when analysing the sufficiency of remedies, especially in cases (as these two) involving upfront buyers rather than fix-it-first remedies, where the remedy taker is identified at the time the clearance decision is adopted. The judgments show that fix-it-first remedies could be successfully used to address some of the concerns relating to the viability and competitiveness of the divested business.
In the Wieland case, none of the commitments were considered sufficient to remedy, fully and permanently, the potential SIEC identified by the EC. The final commitments were (i) structural, including the divestiture of some ARP’s manufacturing plants, slitting centres, and sales team; and (ii) behavioural, e.g., the conclusion of transitional services and supply agreement. Wieland proposed an “up-front buyer” (i.e., the transaction cannot be closed before the parties enter into a binding agreement with a purchaser for the divestment business, approved by the EC), rather than the ‘fix-it-first’ method (i.e., the notifying parties must identify a purchaser for the divestment business and enter into a binding agreement already during the EC’s review).
In the Thyssenkrupp decision, the EC considered that the divestment of several downstream finishing plants for both metallic coated and laminated steel for packaging and automotive Hot-dip Galvanizing (“HDG”) businesses would not be sufficient to resolve the SIEC that the concentration would create. The EC was, in particular, concerned with the viability and competitiveness of the divested business, due to the lack of vertical integration and likely future reliance on the merging parties.
Legal test: removing the SIEC
In Thyssenkrupp, the parties criticised the EC for having applied the wrong legal test and required (in particular in State of Play meetings) to remove the overlap entirely. The GC made it clear that the standard was not to remove the overlaps (which hopefully will deter the EC from making such broad requests), but that the commitments must be sufficient to remove the competition concerns. It found that the EC had not erred in its analysis by rejecting the commitments on the basis of the divestiture business’s insufficient viability, in particular in light of the lack of production capabilities for some of the products involved, the quality of the divested assets (some of them had been dormant), the lack of vertical integration of the divested business, and the risks attached to the potential purchasers (in particular in the Brexit context).
The GC also ruled in Thyssenkrupp that the EC should be able to determine “with certainty” that the commitments can be implemented and will have sufficiently lasting effects to remove the competition concerns. It found that the EC has “broad discretion” in its assessment and that the EU court must only check if the decision is vitiated by a “manifest” error of assessment. [29] This confirms that EU courts are reluctant to second-guess the EC’s assessment of remedy packages, especially if they involve complex economic assessments.
Remedies beyond the overlaps
In the Wieland judgment, the GC confirmed that the EC is always bound to verify the viability of the divested assets, particularly in concentrated markets where only a few players would be able to operate profitably a steel plant. The GC found that considering that the entity to be divested was vertically integrated within ARP, the EC was fully entitled to test whether, once separated, that business would be viable and standalone (and to go beyond an assessment of the mere question of whether the divestment business was of sufficient size to constitute a competitive force and to remedy the horizontal anti-competitive effects of the proposed transaction). [30] Thus, even if the divested business could have remedied the horizontal overlaps, the GC backed up the EC’s reasoning that it did not resolve the vertical concerns, especially as the divested business would likely remain reliant on supplies from Schwermetall.
Likewise, in Thyssenkrupp, the GC upheld the EC’s rejection of the commitments because the divestment excluded any upstream capability. [31] The GC thus fully endorsed the EC’s practice (and Remedies Notice) according to which some products or assets in an upstream market may have to be included in the divestment otherwise solving the overlaps to ensure its viability.
Upfront buyers
In Thyssenkrupp, the parties had tried to address the issue of vertical integration by including, in the commitment, a requirement that the buyer be active in the steel industry and have “sufficient capabilities to operate successfully the Divested Business”. The GC agreed with the EC that such broad-brush requirements were insufficient to guarantee the viability of the commitment (and, in particular, the issues related to the lack of upstream integration of the divested business). The GC also held that in an upfront buyer scenario (where the buyer is not known), the EC cannot be criticised for rejecting a remedy involving the divestment of a business that is not viable on a stand-alone basis without the examining resources of a possible or even presumed future purchaser. According to the GC, such resources cannot be taken into account at the stage of assessing the viability of the remedy package.
This underscores a drawback of an upfront buyer remedy (in which the buyer’s identity is unknown) compared to a fix-it-first remedy (in which the buyer is approved by the EC prior to clearance). In the fix-it-first scenario, the buyer’s resources can be taken into account to assess the viability and competitiveness of the divestment business.
In both cases, the upfront buyer remedies were found insufficient by the EC to guarantee the appropriateness of the purchaser and hence the viability of the remedy package. In Thyssenkrupp, the parties argued that this was wrong given that the EC would always have to approve the purchaser. The GC rejected the argument and confirmed that the upfront buyer requirement was equivalent to a fix-it-first remedy only where the commitments contain “specific requirements” as to the future purchaser, which was not the case here (only general requirements, such as being present in the steel industry, instead of requirements to have existing upstream capabilities). The GC underlined that the EC does not have unlimited discretion when it approves the purchaser: it can only verify that the purchaser meets the requirements foreseen in the commitments. This is why it is particularly important that such the specific requirements relating to the buyer in the commitment be clear and sufficient to ensure the viability of the business.
Conclusion
Appeals following merger prohibitions are very rare. This is primarily because merger prohibitions themselves are rare. Also, when the EC prohibits a concentration the parties have usually very little interest in a protracted court battle which, even in the winning scenario, would not have any meaningful practical effect (indeed, in both the Thyssenkrupp and Wieland cases, the GC rendered its judgment only around three years after the EC decision). For this reason, the judgments in Thyssenkrupp and Wieland are important contributions to scarce case law and help to delineate the legal framework applicable to merger control assessment. The next milestone which will have a significant impact on the substantive assessment of merger cases will come from the Court of Justice’s judgment in CK Telecoms.
________________________
[1] Gap cases are transactions in oligopolistic markets that may not create a dominant position but eliminate competitive constraints that the merging parties exert on each other and thereby reduce significantly the competitive pressure on the remaining competitors. In fact, in this case, the EC also found that in some markets the merger would have even created a dominant position post-transaction.
[2] It has been reported that Thyssenkrupp has appealed the judgment before the European Court of Justice.
[3] Tilman Kuhn, The 15th anniversary of the SIEC test under the EU Merger Regulation – where do we stand? (Part 1), ZWeR 2020, 1-51.
[4] CK Telecoms judgment, para. 118.
[5] Thyssenkrupp judgment, para. 280.
[6] Wieland decision, para. 136.
[7] Wieland judgment, para. 73.
[8] EC Guidelines on the assessment of horizontal mergers under the Council Regulation on the control of concentrations between undertakings, para. 28.
[9] Wieland decision, Section 6.3.2.
[10] Wieland judgment, para. 129.
[11] Wieland judgment, para.130.
[12] CK Telecoms judgment, para. 247.
[13] CK Telecoms judgment, para 228. Specifically, as in other mobile mergers, the EC put a lot of weight on the so-called diversion rations, which measure the proportion of customers switching away from one merging party that opt for the other merging party.
[14] Wieland judgment, para.66.
[15] Wieland judgment, para. 435.
[16] Wieland judgment, para. 83.
[17] Wieland judgment, paras. 87-91.
[18] Thyssenkrupp judgment, paras. 562 and 564.
[19] Thyssenkrupp judgment, para. 77.
[20] Wieland judgment, para. 48.
[21] Thyssenkrupp judgment, para. 255.
[22] Wieland judgment, para. 306.
[23] See for example the GC judgment in Intel (T-286/09 RENV) or the recent opinion of AG Rantos in Unilever (C‑680/20).
[24] Wieland judgment, para 117.
[25] Thyssenkrupp judgment, para. 104.
[26] Wieland judgment, para 164.
[27] Thyssenkrupp judgment, para.135; Wieland judgment paras. 499; 628.
[28] Thyssenkrupp judgment, paras. 954 and 956.
[29] Thyssenkrupp judgment, para. 793; Wieland judgment, para. 461.
[30] Wieland judgment, para. 482
[31] Thyssenkrupp decision, para. 1604; Thyssenkrupp judgment, para. 894.
________________________
To make sure you do not miss out on regular updates from the Kluwer Competition Law Blog, please subscribe here.
