A non-compete obligation which is imposed on the seller in the context of a M&A transaction can be permissible when it is ancillary to the transfer of the relevant business, that is, when it is directly related and necessary to the implementation of the deal. In order to enjoy the fruits of the purchase of the transferred business, the buyer must be able to benefit from some protection against competition from the seller. However, non-compete clauses only comply with antitrust/competition laws when their geographical scope, duration, subject matter and the persons subject to them do not exceed what is reasonably necessary to achieve the legitimate objective of implementing the transaction.

In two judgments dated 28 June 2016, the EU General Court upheld the European Commission's strict approach to non-compete clauses in M&A transactions. This example is European, but the conduct at issue can raise antitrust risk around the world.

There are five key takeaways:

  1. The fact that a non-compete clause is concluded as part of a legitimate M&A transaction does not automatically validate that clause under the antitrust laws.
  2. It is therefore critical that any non-compete clause envisaged as part of a M&A agreement is checked to ensure that it complies with the applicable competition law rules. These limit the acceptable duration and product/geographic scope.
  3. Antitrust authorities can and do investigate non-compete provisions in response to complaints and, importantly, on their own initiative. Moreover, these investigations can lead to very heavy fines. In the present case, the European Commission fined the parties over €66 million and €12 million, respectively, though the Court has now asked the European Commission to recalculate the fines taking into account only the sales of the parties relating to the services covered by the non-compete clause.
  4. Liability can be strict – it may not be necessary for antitrust authorities to prove that the provision actually prevented competition that would otherwise have occurred. In the present case, the Court confirmed that the non-compete clause amounted to a market sharing agreement and could be classified as a restriction of competition "by object" with no need to assess the concrete effects of the clause on the relevant markets. Similarly, in US antitrust law, non-compete agreements between competitors that are not sufficiently ancillary to a M&A transaction can be per se illegal under Sherman Act §1.
  5. Inserting language that a non-compete clause applies only to the extent permitted by applicable law is very unlikely to offer any defence against antitrust liability. In the present case, the parties had used the wording "to the extent permitted by law," but this did not protect them.

The Telefónica/Portugal Telecom case

In 2013, the European Commission fined Telefónica and Portugal Telecom about €66.9 million and €12.3 million, respectively, for, in its view, entering into a market sharing agreement by way of a non-compete contractual arrangement to exclude or limit competition on each other's home markets (Spain and Portugal, respectively). Remarkably, the relevant clause was included in a share purchase agreement dated 2010 whereby Telefónica had acquired sole control over the Brazilian mobile operator Vivo, which was previously jointly owned by the parties.

The clause stated: "To the extent permitted by law, each party shall refrain from engaging or investing, directly or indirectly through any affiliate, in any project in the telecommunication business (including fixed and mobile services, Internet access and television services, but excluding any investment or activity currently held or performed as of the date hereof that can be deemed to be in competition with the other within the Iberian market". That clause was to apply between September 2010 (the date of the closing of the transaction) and December 2011.

In its two judgments of 28 June 2016, the General Court upheld the European Commission's strict approach, and stated that the non-compete clause amounted to a market sharing agreement and classified as a restriction of competition "by object".

First, the Court confirmed that the non-compete clause had the potential to restrict competition by its very nature or "by object" because it was entered into by two potential competitors in the markets for the provision of electronic communication services and television services. The Court added that these markets were liberalised and did not have insurmountable barriers to entry which would rule out any potential competition. With the clause classified as a restriction by object, it was not necessary for the European Commission to show any concrete anti-competitive effects on the market.

Second, the Court upheld the European Commission's assessment that the non-compete clause was not ancillary to the main transaction. The clause referred to the Iberian market, whereas the main transaction referred to an operator (Vivo) whose activity was limited to Brazil. The Court clarified that, for a non-compete agreement to be "ancillary" to a transaction, it is necessary to establish: (a) whether the non-compete restriction is objectively necessary for the implementation of the main operation; and (b) whether it is proportionate to it. The assessment of the objective necessity must be an abstract analysis, which does not require an assessment of the competitive situation on the relevant market, the commercial success of the main operation or the business strategy of the parties. The requirement for a direct and necessary link should be analysed from an objective perspective, that is, the restriction of competition must be, in both product and geographical scope and duration, strictly limited to what is necessary to implement the transaction. The Court did not view as relevant Telefónica's belief that the clause was considered essential by the Portuguese government to protect Portugal Telecom in Portugal. According to the Court, the application of the competition law rules would have been excluded only if the parties were legally obliged by the Government to adopt the relevant conduct.

Nor was the non-compete restriction ancillary to other clauses of the main agreement, namely a call option (the right of a party to buy back its shares held by the other party) and a provision for the resignation of the members of the board appointed by one company in the other company. According to the Court, the ancillary nature of a restriction should be determined by reference to the transaction or operation as a whole, rather than by reference to an artificial division of the transaction or operation into independent provisions. Ancillary restraints cannot be of more economic importance than the operation or transaction that may justify them. In any event, a non-compete commitment should have been strictly limited to what was necessary to implement the unilateral call option or resignation, whereas the non-compete clause applied to both parties.

Third, there was nothing to indicate that the non-compete clause contained a self-assessment obligation on which the entry into force of the non-competition obligation actually depended. The parties had claimed that the wording "to the extent permitted by law" in connection with other elements (such as, for example, the circumstances of the negotiations, the parties' behaviour after the signature of the agreement, and their intent) meant that the clause should be interpreted as not imposing any obligations without a prior self-assessment of the legality of the non-compete arrangement. But the Court noted that the non-compete clause did not establish any terms and conditions that would govern this self-assessment exercise, for example the day on which the alleged self-assessment exercise was to be completed. When commenting on the wording and possible interpretation of the non-compete clause, the Court warned that there is a need to obtain "sophisticated legal advice" both in relation to the negotiation of the main transaction and in relation to any non-compete restrictions entered into in that context.

Finally, the Court held that the European Commission will have to recalculate the fines imposed on the two companies, making a more specific determination of which sales were linked to the infringement and which were not.

Best practices for non-compete clauses in corporate deals

Appropriately limited non-compete clauses in M&A transactions are clearly procompetitive because an agreement for the sale of a company often cannot be achieved if the seller will compete with the transferred company immediately after the transfer. The seller, with its detailed knowledge of the transferred business, would be in a position to win back its former customers immediately after the transfer, and thereby deprive the buyer of the value of the business that it paid for.

However, to be permissible, non-compete clauses must be necessary and proportional to the implementation of the main deal. Below we suggest four best practices for crafting defensible non-compete clauses, although specific legal advice should be taken in every case.

  • First, seek guidance from antitrust/competition counsel on the specific non-compete clause. The fact that a non-compete clause is concluded as part of a legitimate M&A transaction does not necessarily remove that clause from the scope of global antitrust laws. Non-compete obligations will be assessed on the basis of whether their duration, geographic scope, subject matter and participants do not exceed what is reasonably necessary to achieve the legitimate objective of implementing the concentration – as determined by the antitrust rules in the territory affected. Different jurisdictions may have different approaches. Regarding the duration of the clause, for example, in Europe non-compete obligations are justifiable for up to three years in the case of a transfer of goodwill and know-how, and two years if only goodwill is transferred. It is sometimes possible to justify non-competes of up to five years, for example where it is proven that customer loyalty will persist for longer, or where the specific know-how transferred justifies an additional period of protection. U.S. antitrust law does not provide an iron-clad guideline, and the permissible duration depends on the facts of the case.
  • Second, remember that non-competes can also be problematic in the context of joint ventures. In the EU, non-compete obligations placed on controlling JV parents are acceptable where they relate to the products, services and territories covered by the JV agreement. But this does not apply to non-controlling JV parents, and other jurisdictions may have their own approach.
  • Third, consider whether any competitor or customer might complain to the relevant competition authorities or initiate a private action. Antitrust authorities can and do investigate non-compete provisions in response to complaints and, importantly, on their own initiative. M&A agreements are often public and/or subject to disclosure, and therefore any third party may see a non-compete clause and flag its existence to the relevant authorities. It is even possible that one of the parties might in the future decide to bring the issue before a competition authority in order to escape from its contractual commitments. Competition authorities might also initiate an investigation simply because they read something in the media, or because they are alerted by other authorities (as occurred in the Telefónica case).
  • Lastly, do not rely solely on a caveat or qualification that the clause will only be valid "to the extent permitted by law". Such caveats will not necessarily protect the parties from the application of the antitrust rules.