When do information exchanges affect competition? Lessons from the tobacco cartel in Spain regarding Hub-and-Spoke practices

Back in April 2019, the Spanish Competition and Markets Authority (“CNMC”) sanctioned major tobacco manufacturers with a €57 million fine for alleged anticompetitive conduct. This conduct consisted in the exchange of commercially sensitive information through a common distributor, Logista, which was the sole wholesale distributor of 99% of cigarettes in Spain (see CNMC's Decision on Case S/DC/0607/17 Tabacos). 

Logista and all the tobacco companies involved in the cartel lodged an appeal against the CNMC's Decision challenging the existence of a competition infringement, and the sufficiency of evidence regarding the alleged anticompetitive effects, seeking the annulment of the Decision.

On 3 November 2025, the National High Court upheld Logista's appeal and annulled the CNMC’s Decision by rigorously addressing the evidentiary standards for proving facilitator liability in hub-and-spoke arrangements (see Ruling 4663/2025).

1. The sanctioned conduct

As a distributor of tobacco products, Logista offered competing manufacturing companies a series of information services, including access to a digital storage server with sales data by product and province. 

According to the CNMC, these daily exchanges were strategic in nature, had maximum market coverage (99%), and concerned detailed, recent and non-public information. Therefore, the CNMC considered that the constant exchange of this information could have helped tobacco companies adapt their commercial strategies and cause serious anticompetitive effects in the market.

2. Lack of evidence of anticompetitive effects

Adopting the effects-based approach applied by the CNMC, and in line with the case-law of the Court of Justice of the European Union, the National High Court emphasised that where a conduct does not inherently reveal a sufficient degree of harm to constitute a restriction by object and requires contextual analysis, such analysis cannot be limited to a superficial assessment of legality. Instead, the analysis must be comprehensive and rigorous.

The CNMC classified the conduct as a competition restriction by effect under Article 101 TFEU and Article 1 of the Spanish Competition Act. Although the information exchange had occurred since 1997, the CNMC considered that anticompetitive effects only materialised from 2008 onwards. The CNMC argued that the sanctioned conduct provoked (i) stabilisation of market shares among the four main manufacturers despite falling demand; (ii) parallel increases in tobacco prices beyond what tax rises alone would justify; and (iii) elimination of competitive incentives in other commercial variables such as promotions and product launches.

Yet, the Court stresses that the three anticompetitive effects alleged by the CNMC were insufficiently proved. First, in the absence of any direct or indirect contact between the manufacturing competitors, the Court focused on analysing the nature of the information exchanged and found that, given the regulatory scheme of the tobacco market (i.e., a regulated oligopoly with certain transparency requirements on prices), manufacturers can only compete on price, product launches, and promotions. On the contrary, the information exchanged involved daily provincial-level sales from Logista to retailers, which would not affect the competitive variables identified, and therefore could not be considered strategically important or sufficiently relevant to modify a competitor’s commercial strategy.

In this regard, the Court distinguished between the “sell-in” data exchanged (Logista's sales to retailers) and “sell-out” data (actual retail sales to consumers). The Court noted that manufacturers would prefer "sell-out" data to capture real demand, and the fact that Logista charged for such data while providing “sell-in” data free of charge reinforced that the latter had lower strategic value for competitive purposes.

Secondly, the Court criticised the CNMC's failure to construct a proper counterfactual scenario capable of proving the negative effects of the sanctioned conduct over a scenario where the infringement did not occur. In this regard, the Court identified multiple empirical indicators that undermined the alleged causal link between the sanctioned conduct and the alleged associated anticompetitive effects:

• No effects from lack of access to the data: One of the sanctioned manufacturers ceased accessing the relevant exchanged information in 2013, while neither its market share nor overall market dynamics changed, directly contradicting the theory that access to the data was essential for coordinating conduct or maintaining stability.
• Declining revenues and margins: Economic evidence showed that net revenues and margins per €100 sold fell during 2008–2017, consistent with increased competitive intensity rather than collusion between competitors.
• Market share stability: The stability observed during 2008–2017 was no greater than in previous periods or in comparable countries, again rendering the exchange effectless, as falling structural demand (mostly due to regulation and change of general dynamics on tobacco consumption) provided a better explanation for the observed dynamics.

• Price parallelism: Crucially, price parallelism did not diminish after the data exchange ceased in 2018; it actually intensified, fundamentally weakening the alleged causal link.

   

3. The role of facilitators and implications for Hub-and-Spoke cartels

This ruling raises a critical question for regulated or transparent markets: can competitors leverage third-party tools to reduce commercial uncertainty provided that no demonstrable anticompetitive effects materialise?

In this case, the National High Court has confirmed that third-party information exchanges require more than theoretical harm to be contrary to competition law, opening the door for enhanced algorithmic collusion to stretch the limits of antitrust rules.

Based on this judgement, going forward, correlations alone may not suffice for AI-driven and algorithmically mediated markets, and data exchanges will have to be evidenced as effectively altering competitive conducts in the market. However, the focus must remain on ensuring that data architectures and algorithmic processes serve legitimate business purposes and do not facilitate competitively sensitive information exchange.

If you need more information or further guidance in this area, please contact Candela Sotés and Pablo Rodríguez.

VISIT OUR COMPETITION LAW HOMEPAGE