Skip to main content
Publication

A subsidiary question - how to define group members in contracts

21/04/2011

Locations

United Kingdom

A subsidiary question - how to define group members in contracts

The Supreme Court has ruled that a company ceased to be a subsidiary where the shares held by its holding company were pledged to a bank and the bank's nominee became the registered holder.

Consequently, that company was unable to benefit from an indemnity in a charterparty which adopted the statutory definition of a subsidiary. This is a common drafting technique for group definitions, and we examine the implications for commercial contracts generally.

The issue

To be entitled to the indemnity, the claimant company (E) needed to prove that it was a subsidiary of another company (A). The charterparty expressly incorporated the definition of a subsidiary found in Section 736 of the Companies Act 1985. This provision has since been replaced by Section 1159 of the Companies Act 2006, but the wording is substantively identical.

Section 736(1) sets out three tests for establishing a holding company / subsidiary relationship. It would be sufficient for E if any of the following were true at the relevant time:

(i) A held a majority of the voting rights in E;

(ii) A was a member of E, and had the right to appoint or remove a majority of the board of directors of E; or

(iii) A was a member of E, and controlled alone, pursuant to an agreement with other shareholders of E, a majority of the voting rights in E.

This may seem straightforward, but the group structure of A and E had complications. Initially, A owned 50% of the issued shares in E, with the other 50% being held by a joint venture partner (S). Test (i) above was not satisfied - however, A had entered into an agreement with S regarding voting rights which meant that test (iii) was satisfied. At that point, E was a subsidiary of A.

Subsequently, A pledged, charged and assigned its shares in E to a Scottish bank by way of security. The bank insisted, in accordance with standard practice in Scotland, upon a bank nominee becoming the registered holder of those shares (E being incorporated in Scotland). A’s name therefore no longer appeared in E's register of members.

Decisions of the courts

The trial judge held that it "departs from any kind of commercial reality or business sense" when considering external parties to say that E was no longer a subsidiary of A merely because of A entering into a security arrangement of this nature with a bank.

However, the Court of Appeal and the Supreme Court both disagreed and found unanimously against E.

A key factor was that the separate statutory definition for a "subsidiary undertaking" (used principally for accounting purposes) deemed one undertaking to be a member of another in certain circumstances even where it was not registered as such, but no equivalent deeming provision appeared in Section 736. There was no clear basis to be sure that this was a drafting error which the court could correct, even if a mistake seemed a likely explanation. Section 736 was not ambiguous, and E’s claim required "an impermissible form of judicial legislation" - albeit that on the facts of the case the outcome was "odd and possibly absurd".

The appeal courts also rejected the argument that the reference in test (iii) to membership could mean anything other than formal entry in the register of members of E.

A further argument that the statutory definition should operate differently when imported into the specific charterparty similarly failed. The court was not in a position to re-write the contract for the parties, and it was not even clear what the parties would have agreed had they addressed their minds to the issue.

Implications of the case

The facts of the case are unlikely to recur, but of greater interest is whether companies might be able to exploit the outcome in order to escape from liability. Could a group which is subject to an onerous contractual provision (such as an indemnity or perhaps a restrictive covenant) deliberately break a holding company / subsidiary relationship so that a certain group member is no longer caught?

It is always important to check carefully contractual definitions of subsidiaries and holding companies, and related terms such as "affiliates" or "associates". This should be done by reference to the group structure of each party, and it should not be assumed that "standard" wording will suffice.

If required, a deeming provision could be added as an express variation to the imported statutory definition to cover situations such as that highlighted by the case, and the use of nominees more generally. Admittedly, it is not entirely easy to frame this both precisely and concisely.

Alternatively, "subsidiary/parent undertaking" definitions could be adopted instead so as to incorporate the statutory deeming provision referred to above. Considerable caution needs to be exercised with this approach as these are much wider concepts and may be inappropriate for the contract. Also, the statutory deeming provision would cater for nominee arrangements but possibly not for security arrangements, though the courts appeared to assume it did.

Nevertheless, intentionally breaking a group relationship could be extremely risky. The gain under one contract might be greatly outweighed by disadvantages, whether arising under other contracts or, for example, through inadvertently ceasing to be a member of a VAT group. Although the appeal courts acknowledged the potential for abuse, they were not persuaded that E’s argument must be upheld on "policy" grounds. The following should be borne in mind:

  • a subsidiary relationship is most frequently created by test (i) which does not depend on membership;
  • English lenders do not usually take registered title to shares at the outset of a security arrangement (though lenders in other jurisdictions may follow the approach in Scotland, which arises because of differences in Scottish security laws);
  • despite the statutory definition being in its current form for 20 years and the issue being identified in academic texts, the court knew of no evidence of exploitation; and
  • the statutory definition can be amended by regulations (it remains to be seen whether BIS will consider this necessary).

The case discussed is Farstad Supply A/S v Enviroco Limited [2011] UKSC 16

Sign up to our email digest

Click to subscribe or manage your email preferences.

SUBSCRIBE