Brussels à jour

Our (Ware)House, in the Middle of Our Street

The General Court handed down on 18 May its judgment in Canon v Commission, clarifying how the Court of Justice's landmark ruling in Ernst & Young applies to warehousing structures in M&A deals.

The Canon/Toshiba saga is one of those examples that, in life, you cannot have the cake and eat it. Although the merger was cleared, the Commission took issue with a warehousing structure the parties had put in place to address Toshiba's urgent need to receive the purchase price before the required regulatory approvals had been obtained. The Commission considered the interim transaction, which placed the target company in the hands of a vehicle specifically created for that purposes (the "parking transaction"), as gun jumping and imposed an aggregate fine in the amount of EUR 28 million on Canon for a failure to notify the deal prior to implementation, Article 4(1) of the EU Merger Regulation, and for implementing the deal prior to clearance, Article 7(1) of the EU Merger Regulation.[1] In the wake of the Court of Justice's landmark ruling in Ernst & Young,[2] there was hope within the antitrust and M&A communities that there might be some (or more) wiggle room for parties to design warehousing structures that would survive the Commission's scrutiny. The General Court's ("GC") judgment Canon v Commission,[3] upholding the Commission's decision, frustrates these hopes for now… subject to what the Court of Justice might have to say on appeal.

The Merchant of Venice

On 12 August 2016, Canon notified the Commission of its plan to acquire Toshiba Medical Systems Corporation ("TMSC") from Toshiba. The transaction was unconditionally cleared by the Commission on 19 September 2016. At the time, Toshiba experienced financial difficulties. In order to avoid having to report negative results for the 2015 financial year, the company had initiated an accelerated bidding process for the sale of TMSC and had proposed to bidders a so-called "80/20 structure" pursuant to which the buyer would acquire 20% of the shares of TMSC minus one share (“20% minus one”) by the end of March 2016, and simultaneously pre-pay the price for the rest of the shares ("80% plus one"). Toshiba would have continued to own the remaining 80% plus one shares (which would have been pledged to the purchaser as collateral), and therefore control TMSC, until all competition authorities had cleared the transfer of the remaining shares. If the required merger control clearances would not have been obtained by 14 June 2016, the purchaser would have retained the 20% minus one TMSC’s shares and terminate the pledge over the 80% plus one. In turn, Toshiba would have refunded to the purchaser the provisional payment made for the 80% plus one TMSC’s shares, minus a reverse break-up fee of YEN 35 000 million (EUR 29.1 million).That structure was intended to allow Toshiba to remove TMSC from and include the purchase price for TMSC in its accounts for the 2015 financial year, irrespective of the time it would take to obtain merger control approval.

During the bidding process, Canon proposed an alternative two-step transaction (warehousing) structure involving an interim buyer (a special-purpose vehicle) that would allow the parties to achieve the same results. In accordance with Canon's (winning) proposal, TMSC converted its common shares into three different categories of shares, namely 20 class A shares (voting shares), one class B share (non-voting share) and 100 class C shares (voting shares with a redemption option exercisable by TMSC).

  • As a first step, Canon and Toshiba entered into a share transfer agreement whereby Canon acquired the class B non-voting share (which gave it veto rights over any decision to sell the target to any other purchaser), and 100 class C voting share options for EUR 5.28 billion (representing the entire value of TMSC). On the same day, the interim buyer (MS Holding) and Toshiba entered into a share transfer agreement, whereby MS Holding acquired the remaining 20 class A voting shares of TMSC for EUR 800. MS Holding had not existed prior to the bidding process and had been set up  by a Japanese law firm which selected its three (independent) shareholders without seeking the approval of Canon or Toshiba. This first step was carried out prior to notification to or approval by the Commission.
  • As a second step, following approval of the merger by the Commission, Canon exercised its 100 class C share options, acquiring the underlying voting shares of TMSC, while TMSC acquired Canon's class B non-voting share and the remaining class A voting shares of MS Holding. The ability to invoke the options was dependent on having secured the necessary regulatory approvals.

In June 2019, the Commission concluded that Canon had breached the notification requirement and standstill obligation imposed by the EU Merger Regulation, and fined the company for both infringements a hefty total of EUR 28 million. In September 2019, Canon appealed the fine before the GC, saying the transaction structure was “in line with existing and recently reconfirmed jurisprudence of the European courts” (meaning E&Y). On 18 May 2022, the GC confirmed the Commission's decision.

House of Cards

M&A dealmakers have from time to time turned to warehousing structures where the parties thought they could not—or simply did not want to—wait to close their transaction until all required merger control approvals have been obtained. Warehousing structures allow for a quick sale and payment of the purchase price, which may be come in handy if the seller is in financial difficulties. They also allow the parties to shift the merger control risk entirely to the buyer, which will typically need to pay the purchase price before merger control clearances have been obtained and will not be entitled to a refund if the deals gets prohibited or cleared only to subject to remedies.

In the past, publicly available guidance on warehousing structures has been fairly scarce, raising the question whether and to what extent the Commission and other antitrust regulators might be willing to accept them at least in certain circumstances.

  • The Commission's Consolidated Jurisdictional Notice of 2008 explains that, if an interim buyer acquires shares on behalf of the ultimate acquirer, the Commission will treat the interim warehousing transaction as the first step of a single concentration comprising the acquisition of control by the ultimate buyer. Consequently, implementation of that warehousing stage will amount to gun-jumping in the eyes of the Commission, even if it does not itself confer control over the target.[4]
  • The other piece of guidance that practitioners turned to was the GC's judgment in the Editions Odile Jacob case, which saw Odile Jacob challenging a Commission clearance decision of Lagardère's acquisition of Vivendi's book-publishing business (VUP), which involved Natexis Banques Populaires (NBP) as a middleman between Lagardère and Vivendi to accelerate the deal.[5] As VUP was seeking a quick sale and payment, prior to merger control approval, Lagardère requested NBP to act in its place and, by means of a subsidiary set up for that purpose, to acquire the target assets from VUP, to hold them provisionally and to sell them to Lagardère once the latter had obtained clearance. The GC's reply to the applicant's allegation that the first step of the transaction formed, together with the second step, a single concentration, was to examine whether the ultimate buyer—Lagardère—could exercise control over the target at the first stage. The GC looked into whether the interim acquisition by NBP had, as such, transferred control to the ultimate purchaser, i.e. Lagardère. The GC concluded that it had not because, according to the underlying purchase agreement, title to VUP was to pass to Lagardère only at the date of clearance by the Commission, and not before. Pursuant to this agreement, NBP's subsidiary only took the place of Lagardère in acquiring and holding the target assets for the sole purpose of transferring ownership thereof to it, once Lagardère has obtained authorization from the Commission. Lagardère did not have any influence / control over the target during the parking operation. Furthermore, the parking structure fell within the scope of Article 3(5)(a) of Regulation 4064/89. Therefore, the standstill obligation had not been infringed by Lagardère.[6] However, on appeal, the Court of Justice stressed the GC's on the warehousing aspects of the case were no more than an obiter dicta, missing the opportunity to provide guidance on such deal structures.

As with so many things, the devil may well be in the details when it comes to warehousing structures and their prospects of surviving the Commission's scrutiny from a gun-jumping perspective. Hence, the antitrust and M&A communities' eagerness to see whether the GC might shed more light on what's possible and what's not.  

Domino Effect

The crux of the Canon case is whether the first step of the warehousing transaction required merger clearance or not. In its assessment, the GC looked at whether the two steps amounted to one single concentration and the first step could be considered "partial implementation of a concentration", as the Commission had labelled it. The GC upheld the Commission's assessment on both aspects. According to the GC, the first transaction contributed to the acquisition of final control over TMSC by Canon, which occurred with the second transaction. In the context of the transaction structure chosen by the companies, the first transaction was necessary for Canon to gain control over TMSC. In reaching that conclusion, the Commission (confirmed by the GC) mainly relied on three elements. First, the first step was only carried out in view of the second step. Secondly, the sole purpose of MS Holding was to facilitate the acquisition of control of TMSC by Canon. Thirdly, Canon was the only party able to determine the identity of the ultimate purchaser through its veto share (although it did not control TMSC during the first step), and assumed the economic risk of the entire transaction from the time of the first step.

Throughout the judgment, the GC further elaborates on these and touches upon a number of other aspects that lead it to confirm the Commission's decision, although it is not always entirely clear to what extent these aspects relate to the first prong of the test (single concentration), the second prong of the test (partial implementation), or both.

  • The GC confirmed that Canon did not control TMSC during the interim period (which was undisputed). However, that did not save Canon from the allegation of having "jumped the gun" because, as the Court of Justice had said, Article 7 of the EU Merger Regulation also prohibited the partial implementation of a concentration, which will occur in warehousing scenario if the first step (the "parking transaction") contributes, in whole or in part, to the change of control over the target.[7]
  • According to the GC, the first and second step can qualify as a single concentration, and the first step might thus amount to gun jumping, even if the interim buyer does not formally act "on behalf of" the ultimate purchaser (which is the language used in the Consolidated Jurisdictional Notice).
  • The interim buyer, MS Holding, was a special purpose vehicle that lawyers had specifically set up on behalf of the parties. Apparently, the lawyers had also brought in the investors that held the shares in MS Holding.
  • The arrangements between the parties were essentially of a tripartite nature. While the initiative for a warehousing arrangement was initially taken by Toshiba (the "80/20 structure"), the structure was ultimately put in place in the interest of both parties: Toshiba's—to remove TMSC from its accounting by end-March 2016 and recognise the sale price, and Canon's—to win the Toshiba tender to acquire TMSC.
  • Canon—and not MS Holding—bore the entire financial risk associated with TMSC's performance in the interim period. At the closing of the first step, Canon paid a purchase price of EUR 5.28 million to Toshiba, representing essentially the entire value of TMSC. Upon the closing of the second step, title of the Class A would pass to Canon at a price that was fixed ex ante and, therefore, did not take into account TMSC's performance. Notwithstanding the fact that MS Holding was entitled to receive some dividends, MS Holding's financial interest in the transaction was limited to receiving the pre-agreed purchase price.
  • While it held 100% of the voting rights in TMSC during the interim period, MS Holding's freedom to exercise these voting rights was limited in several ways, including by virtue of certain veto rights that Canon enjoyed and the fact that MS Holding was not free to choose for how long it would own the shares (which depended on Canon exercising the options attached to the Class C shares). Apparently, there was also evidence that MS Holding had exercised its voting rights very little. MS Holding's ownership of the voting rights in TMSC was only temporary, a fact that TMSC's management was well aware of. While there was no evidence, according to the GC, that MS Holding had exercised the usual function of a controlling shareholder, Canon had apparently been in touch with TMSC's management during the interim period.
  • Even if the Commission had not cleared the merger, Canon retained exclusive right to determine the identity of the (alternative) ultimate purchaser of TMSC, for which it had already paid the full purchase price to Toshiba.
  • The warehousing structure was not saved from the gun jumping verdict by virtue of the second step being structured as an option, as opposed to a binding share purchase agreement. According to the GC, Canon did not get a "genuine" option affording it the right to buy TMSC at a later stage, but had already paid the full price at closing of the first step in exchange for a "de facto automatic mechanism" to acquire TMSC in the second step.

What the GC makes abundantly clear is that it will always look at the economic reality of what's happening rather than the formalities of how the parties have chosen to set up their structure. And that look became more restrictive in the Canon judgment, as compared to Odile Jacob.

In the Following Episode…

So what does the GC's judgment mean, going forward? Is this the end of warehousing? The judgment indicates a distrust of the GC regarding warehousing, in line with the Commission's equal lack of enthusiasm for parking structures, already laid out clearly in the 2008 Consolidated Jurisdictional Notice. Technically, a warehousing structure would not infringe the standstill obligation[8] if (i) the two steps of the warehousing transaction do not constitute to a single concentration, and (ii) the first step of the transaction (i.e. the "parking") does not amount to partial implementation of a concentration. In theory, there might be deal structures that avoid either the first or the second requirement, or both. The parties will likely need to consider ways to ensure, at the very least, that:

  • The interim buyer, not the ultimate purchaser, bears the financial risk associated with the target's business operations while "parked" at the warehouse. At the very least, that risk should not be borne by the ultimate purchaser;
  • The ultimate purchaser does not get to decide the identity of an alternative buyer if merger control approval should not be obtained;
  • The interim buyer should have generally unrestricted control over the target, while the ultimate purchaser should not have any governance or even minority protection rights that might work to limit the interim buyer's control (both de jure and de facto).
  • The second step should preferable be structured as a "genuine" option, neither as pre-agreed purchase agreement nor as an "option" that is, in reality, tantamount to mechanism that will automatically transfer control to the ultimate buyer.

Devising a mechanism that would allow the seller to receive the purchase price upon closing of step one would seem to be particularly challenging. Even if the parties take these precautions, there is likely to remain a meaningful risk that the Commission might take an issue with their structure. Moreover, in practice a "low-risk" structure may not be overly attractive for the parties and it might prove difficult, if not impossible, to find a "warehouse" that would be willing to help in these circumstances. Warehousing has never been easy, and Canon certainly has not made it easier.

On that positive note, don't forget to wear sunscreen and to follow us on LinkedIn for your favorite EU competition law topics!

[1]           For each of these two separate infringements, the Commission imposed a fine of EUR 14 million.

[2]           Case C-633/16, 31st May 2018, Ernst & Young P/S v Konkurrencerådet.

[3]           Case T‑609/19, 18 May 2022, Canon v Commission.

[4]           Commission's Consolidated Jurisdictional Notice under Council Regulation (EC) No 139/2004 on the control of concentrations between undertakings, 16 April 2008, para 35.

[5]           Case T‑279/04, 13 September 2010, Editions Odile Jacob v Commission, para 13.

[6]           Case T‑279/04, 13 September 2010, Editions Odile Jacob v Commission, para 170-172.

[7]           The GC also confirmed that the concept of "partial implementation" also applies in the context of obligation to notify a concentration prior to implementation, Article 4(1) of the EU Merger Regulation.

[8]           That is, as far as the ultimate purchaser is concerned. Whether or not the interim buyer needs merger control approval is a separate issue.


» DOWNLOAD Newsletter (PDF)


Don’t miss any edition of our Brussels à Jour Newsletter.
You can simply follow the hashtag #Brusselsajour on LinkedIn to make sure you receive our updates in your feed.