The Chilean Supreme Court upholds a landmark decision by the Competition Tribunal condemning the country ’ s three largest supermarkets for conspiracy in fixing minimum resale prices of poultry

Chile -   | Concurrences, Antitrust Publications & Events
Mario Ybar, Counsel of the E.U. & Antitrust practice in Santiago de Chile.

The Chilean Supreme Court recently upheld a landmark decision by the Competition Tribunal (“TDLC”) in Fiscalía Nacional Económica en contra de Cencosud S.A. y otras, Rol C-304-2016, condemning the country’s three largest supermarkets chains for having engaged in a conspiracy to 1x minimum resale prices of poultry by means of communications through their common suppliers.

This is the 1rst time these courts have addressed a complaint brought by the Fiscalía Nacional Económica (“FNE”), the Chilean Competition Authority, explicitly based on a “hub-and-spoke” theory. Following these decisions, Chile has become one of the few jurisdictions in which a 1nding of collusion based on a hub-and-spoke arrangement has been affirmed after going through a complete judicial review.

This article summarizes many of the important insights about this category of anti-competitive agreement that can be found in the FNE’s complaint, in the supermarkets’ defenses, and in both judicial decisions.

The facts

In December 2011, the FNE obtained a warrant to seize and search the computers of several mid-level executives of the largest supermarket chains in the country. The investigation focused on “category managers” for white meat (poultry, pork, and fish), where preliminary evidence of collusion had been uncovered.

While no evidence of direct communications between the supermarkets was found during the raid, numerous emails were discovered showing the existence of vertical RPM agreements between the supermarkets and their poultry suppliers. These agreements required the stores to set prices above the listed wholesale price, which was the same across all supermarkets.

It is important to note that the minimum threshold (listed wholesale price) was set well above the actual price at which supermarkets acquired the products [1] .

Likewise, the FNE found direct evidence showing (a) category manager complaints to supplier “key account managers” (“KAMs”) about rival supermarkets pricing their products below their own listed wholesale price for those products (A-B communications); and (b) emails from the suppliers’ KAMs to the noncompliant supermarkets transmitting the complaints (B-C communications).

Some of the A-B communications merely reported what the rival supermarket was doing, while others went one step further, demanding that the supplier intervene directly and cause the competitor to raise its prices. Several considerations prompted the FNE to further investigate a potential horizontal agreement between supermarkets rather than simply challenging the vertical RPM agreements that had become apparent in the evidence [2].

However, none of the supermarkets had applied for leniency. Conversely, they uniformly denied the existence of any agreements (vertical or horizontal) on pricing. Therefore, the FNE had to build its case grounded on the evidence it did have: (a) abundant emails showing A-B and B-C communications, and; (b) explicit RPM agreements between the supermarkets and a number of their key suppliers.

The case law on hub-and-spoke agreements

Hub-and-spoke collusion involves the use of vertical exchanges or agreements between 1rms at one level of the supply chain (the spokes), and a common trading partner on another level of the chain (the hub), to generate and maintain a horizontal agreement among the spokes regarding the terms under which they will deal with the hub. [3] - [4] What separates unlawful collusion enabled through vertical restraints from presumptively legal vertical arrangements is proof of a “rim” that connects the spokes into a horizontal agreement [5]. Without the rim, equivalent vertical arrangements organized by a single 1rm would not support the 1nding of collusion, but merely constitute independent parallel behavior. At the time of the FNE’s investigation, insights into the legal assessment necessary to determine when there was a rim that turned a set of parallel agreements into a horizontal agreement could only be found in two jurisdictions: USA and UK [6]. US courts have been dealing with hub-and-spoke arrangements for more than eighty years [7], and have consistently followed an inference standard based on plus factors to determine when an otherwise parallel vertical agreements gives rise to a horizontal agreement [8].

Among those plus factors, US courts have considered, for instance, knowledge by 1rms that their competitors are entering into similar agreements; acceptance of vertical restraints subject to the condition competitors are also adhering to the restrain; whether the restraint is against the individual independent business interest of the 1rm; and whether the restraint constitutes a significant departure from previous business practice [9]. In the past two decades, “since antitrust endorsed the distinction between vertical and horizontal agreements”, [10] US courts have found the existence of a horizontal conspiracy in two hub-and-spoke cases: FTC v. Toys “R” Us (TRU) and United States v. Apple (the eBook case) [11].

The approach followed by courts in the UK when analyzing hub-and-spoke agreements derives from the European regulation on concerted practices and information exchanges. To establish the existence of a concerted practice from the evidence of vertical communications, UK courts have required that the exchange of information through the hub satisfy a legal test based on objective conduct-related criteria, and subjective elements regarding the state of the mind of the communicating parties. [12] It is particularly noteworthy that the difference in these approaches to hub-and-spoke cases on both sides of the Atlantic only emerged at later stages of the trials in “Replica Football” and “Hasbro Toys and Games”.

Both decisions from the OMce of Fair Trading (2003) initially inferred the existence of horizontal agreements from parallel vertical arrangements based on different plus factors (more decisively, the reciprocal conditioning of the acceptance of the hub´s RPM policy subject to its adhesion by the competitors) [13]. However, the standards were subsequently changed during the arguments before the Competition Appeal Tribunal and the Court of Appeal.

The FNE's investigation and complaint

The FNE 1led a complaint against Walmart, Cencosud and SMU before the Competition Tribunal in January 2016. [14] The complaint alleged, as a factual matter, that each of the three supermarket defendants engaged in resale price maintenance agreements with their poultry suppliers (Agrosuper, Ariztia, Don Pollo). According to the authority, the market therefore was affected by a parallel net of minimum price restrictions which prevented the supermarkets from selling poultry products below the listed wholesale price. The FNE’s legal theory was that these arrangements constituted an unlawful horizontal arrangement because each vertical agreement was conditioned on adherence to the same minimum price scheme by the other competitors. [15]

The existence of a systematic mechanism to enforce the supermarkets’ compliance with the RPM [16] provided the rim to convert the respective vertical arrangements into a horizontal agreement. In support of its theory, the FNE relied on several emails written by executives from the three supermarkets in support of this theory [17]. In addition, the agency relied by additional “plus factors” as evidence: Reciprocal knowledge by each of the defendants that the same RPM policy was adopted by the other supermarkets. That was contrary to the individual self-interest of the defendants to consent the minimum price rule without requiring symmetry by the other supermarkets. Because poultry is one of the most important drivers of consumer traMc to supermarkets, the products are usually subject to heavy discounts and even prone to lossleading strategies.

Therefore, it would not make sense for any retailer with a certain degree of market power (as the defendants) to commit to price above a certain threshold if their competitors were free to attract customers by pricing the poultry more conveniently.

Regular and reciprocal active monitoring by each supermarket chain regarding the prices of the different poultry products exhibited on their competitors´ premises. The FNE’s theory of the case was supported with legal opinions from two well-known scholars, Spencer Weber Waller [18] and Okeoghene Odudu, [19] which were incorporated into the process.

The supermarkets' defenses

The supermarkets each adopted very a similar defense strategy. The defendants uniformly (a) challenged the FNE’s allegations about the existence of vertical agreements with the poultry producers, arguing instead that these were policies unilaterally imposed by the poultry suppliers (which was demonstrated in another cartel case by the FNE against the suppliers during this period); (b) provided alternative explanations for the emails sent to the suppliers, arguing that those communications merely sought discounts from their suppliers to compete more aggressively, or were aimed at prevent possible supply problems due to the rivals’ promotional pricing, and; (c) denied that the complaints received from the suppliers had any effect or inQuence on their own unilateral pricing decisions.

Most relevant for this article, Walmart not only questioned the factual grounds of the FNE’s complaint but also contested its legal basis.

As stated in the company’s response, there was a “true theory of the case” which the FNE failed to assess. According to Walmart, the evidence presented by the FNE needed to be judged under the legal framework applicable to information exchanges. Namely, a horizontal conspiracy could only be found if all the objective and subjective criteria adopted in the UK hub-and-spoke case law were established: (i) retailer A discloses to supplier B its future pricing intentions; (ii) A may be taken to intend that B will make use of that information to inQuence market conditions by passing that information to other retailers (of whom C is, or may be, one); (iii) B does, in fact, pass that information to C; (iv) C may be taken to know the circumstances in which the information was disclosed by A to B; and (v) C does, in fact, use the information in determining its own future pricing intentions. [20] Walmart argued that because the factual basis of the case largely related to information exchanges about current prices, the FNE’s theory did not satisfy the UK framework, and therefore the FNE’s complaint should be dismissed. The FNE rejected Walmart´s position and responded that the case was not about future information exchanges between competitors, but rather was about a horizontal conspiracy that could inferred from a set of parallel agreements. The agency also argued that hub-and-spoke collusion could not be limited to the hypothesis of future information exchange.

The competition tribunal's decision

The TDLC issued its 1nal decision on February 28, 2019. In its judgment, the Tribunal resolved the dispute concerning the appropriate legal test under which the FNE’s complaint should be scrutinized. Rejecting Walmart’s argument that the UK framework needed to be satisfied, the Tribunal declared: “That, even when [information exchanges through a common party] is the most common type of hub & spoke agreement, it is not the exclusive form these kinds of agreements can take. More critically, in no part of its complaint does the FNE refers to an information exchange of the A-B-C type as the basis of its accusation. In light of its irrelevance for the present dispute, it is unnecessary to refer to the European case law about the hub & spoke agreements of the A-B-C type and the legal requirements established therein; Therefore, what is relevant in this case is to establish, from the set of evidence provided in the proceeding, whether the merely vertical agreements existing between the supermarkets and their poultry suppliers had a common element that “horizontalized” them. In particular, as we have indicated, if each chain complied with its respective agreement on the understanding that the others would also do so and required such compliance by the latter”. [21]

After settling the controversy about the applicable legal test, the TDLC con1rmed the existence of a horizontal agreement between the supermarkets by looking to the plus factors that had been proven during the trial: “The evidence provided in these proceedings and previously analyzed, consisting of emails, witness depositions, legal and economic reports and other statements made before the FNE, considered as a whole, allows us to conclude that the three supermarket chains adhered to a rule of behavior consisting of a vertical restriction to sell poultry at a minimum price in its various premises. Besides, adherence was conditioned on the fact that the rest of them also did so, which is especially supported when observing that each supermarket required suppliers to force their competitors to comply with the rule”. [22] The Tribunal imposed 1nes on the defendants equal to 20% of the total poultry sales affected by the unlawful collusion. The judgment considered affected sales to include only those that implied an operating margin lower than 10% (which on average were about 7% of the total poultry sales made by the supermarkets during the relevant period). [23] However, the TDLC reduced Walmart’s 1ne by 15% for having a “serious, credible, and effective” competition compliance program.

The 1nal 1nes imposed were: US$5.29 million (Cencosud); US$3.15 million (SMU); and US$4.35 million (Walmart).

Appeal to the supreme court

The FNE and the defendant supermarkets all appealed the TDLC’s decision to the Chilean Supreme Court. The supermarkets sought an acquittal before the high court, which the agency sought an increase in the 1nes by challenging the TDLC´s “affected sales” estimation. On April 8, 2020, the Supreme Court issued its decision upholding the TDLC’s judgment as to the supermarkets’ liability. The Court wrote: “The disapproval on the supermarkets’ conduct lies in that their promotional prices were not freely established since they were subject to a rule that they followed under the understanding that their competitors would also follow, complaining through their common poultry suppliers when this did not happen” (…) .

All this took place, at least, between the years 2008 and 2011, and meant that the price of sale of fresh chicken meat to the 1nal consumer was not determined in light of the supply and demand laws (…) with prejudice to consumers concerning a highly demanded and hard to substitute product”. [24] Despite aMrming the judgment on liability, the Supreme Court objected to the TDLC’s methodology in setting the 1nes. As stated in its decision, a 1ne equal to the 20% of what the lower court considered to be the affected sales “does not to fully comprehend the damage caused by the conduct since it does not consider the effects that the latter had on the demand for other products and the market in general”. [25] Consequently, based on the gravity of the offense, the Supreme Court decided to raise to double all three fines imposed by the TDLC. This document is protected by copyright laws and international copyright treaties.

Non-authorised use of this document constitutes a violation of the publisher's Furthermore, the Supreme Court eliminated the 15% 1ne reduction granted to Walmart by the TDLC based on the company’s compliance program. According to the Court, for a compliance program to be “serious, credible and effective” it has to necessarily prevent the anti-competitive conduct from being committed. “[T]he veri1cation of anti-competitive conduct for a period of four years,” the Court concluded, “shows that the compliance program was neither suitable nor effective in ful1lling its preventive purpose.

This reveals, on the one hand, that the program needs to be improved and, on the other hand, the merit of the sanction”. [26]

Concluding remarks

Hub-and-spoke collusion is a broad concept that under no circumstances should be limited to the information exchange hypothesis. Both the TDLC and the Chilean Supreme Court agreed on that point in deciding to follow a longstanding tradition in the US that has allowed horizontal conspiracies of different sorts to be inferred from the existence of parallel vertical agreements and the competitors’ awareness of those as long as there are corresponding plus factors. Among those plus factors, the proof of each constituent vertical agreement being conditioned on the participation of others —as was found in the Chilean supermarkets case— provides the most convincing evidence in order to meet the inference standard.