Follow us

A landmark UK High Court judgment approves a scheme of arrangement despite attempts by a shareholder to defeat the scheme by engaging in ‘share splitting’. The decision is to be welcomed to help avoid such attempted forms of manipulation in the future. The events that occurred in this transaction also highlight the need to abolish the anachronistic ‘head count’ test.


  • In the recent Dee Valley case, the UK High Court approved a scheme despite a deliberate attempt by a shareholder to defeat the scheme by share splitting.
  • While the Court in Australia has the power to disregard the outcome of the vote required by the head count test, this power has yet to be used by an Australian Court and the recent UK High Court decision highlights the issues associated with the anachronistic head count test.
  • In this article we outline the reasons for the abolishment of the head count test.


On 8 February 2017, the UK High Court approved a scheme between Dee Valley Group plc and its members to implement the takeover of the company by Severn Trent Water Limited.

The Court did so despite a deliberate attempt by a shareholder to defeat the scheme by transferring shares to others to create a majority of members opposed to the scheme where (absent such 'splitting') a majority in number of members would have voted in favour of the scheme. It was the first time that such share splitting had come before the Court in the UK.


The Court (in Australia and the UK) may approve a scheme if, at the scheme meeting, the scheme is approved by (a) 75% or more of the shares voted and (b) a majority in number of the members present and voting either in person or by proxy. It is this second 'head count’ test that was in issue in the Dee Valley case.

In November 2016, Severn Trent announced a recommended scheme to acquire the issued share capital of Dee Valley.

In early January 2017 (prior to the scheme meeting), it came to the attention of Dee Valley that one employee had acquired 461 shares in the company and gifted one share to each of 443 individuals. This had the effect of more than doubling the number of members on Dee Valley's shareholder register over a two day period. This plan appeared to be an organised attempt to defeat Severn Trent's scheme by preventing the head count test being satisfied.

Before the scheme meeting, Dee Valley applied to the Court for a directions order that, in light of the evidence of potential manipulation of the register of members, the Chairman of the scheme meeting should be given discretion to exclude the 'split votes', that is the votes of the new shareholders.

The Court gave the Chairman the discretion to exclude the votes of the new shareholders and report back with what the outcome would have been had those votes been included. This was on the basis that failure to exclude the votes of the new shareholders at this stage would have meant that the required statutory majority would not be obtained and the Court would have no jurisdiction to approve the scheme.

The Chairman duly exercised that discretion at the scheme meeting and (on that basis) the vote to approve the scheme passed the statutory tests.

At the final court hearing, the Court considered whether, in the light of all this, to approve the scheme.


Ultimately, the Court decided that the scheme should be approved and indicated that, in future, it will not be necessary to apply to the Court before the shareholder vote for orders permitting the Chairman to exclude ‘split votes’.

The Court concluded that:

  • share splitting undermines the spirit of the scheme legislation and is objectionable;
  • members voting at a class meeting directed by the Court must exercise their power to vote for the purpose of benefiting the class as a whole and not merely individual members only (noting that a scheme meeting is different from a company’s general meeting);
  • the Chairman had the power to reject the votes on proper grounds; and
  • the Chairman was right to disallow the votes of the transferee shareholders because they had each acquired a single share where the objective can only have been to manipulate the voting at the court meeting to defeat the scheme – they could have given no consideration to the interests of the class of members they had joined.

The new shareholders that came onto the register as a result of the share splitting represented 0.01% of the shares in total and 0.03% of the shares voted at the scheme meeting. Over 98% of the shares voted at the scheme meeting (excluding those held by a competing bidder) were in favour of the scheme and 92% of those members attending and voting (in person or by proxy) were in favour (excluding the transferee shareholders). Absent the share splitting, the scheme would clearly and comfortably have met the statutory majority thresholds.


In Australia, in 2007, Parliament sought to address the practice of share splitting by giving the Court a statutory power to dispense with the head count test.

Whilst the primary intention of this statutory amendment was to address share splitting, in pSivida Ltd v New pSivida, Inc. [2008] FCA 627 and in Re Plantic Technologies Ltd [2010] VSC 484 there was a suggestion that this power could be used by the Court to dispense with the head count test in circumstances where a large number of the shares in the scheme company are held indirectly in the form of depositary interests through a nominee company. In neither of these cases did the Court have to decide the issue.

The significant policy issues arising from the head count test were also starkly revealed during the 2009 Hong Kong scheme involving the acquisition of PCCW Ltd, where share splitting (or ‘vote manipulation’ as the Court described the practice) was used with the intention of helping to satisfy the head count test. In this instance, the Hong Kong Court of Appeal ultimately declined to approve the scheme on the basis of the share splitting that had occurred. The head count test has been removed from the Hong Kong scheme of arrangement provisions. There were also wide-spread reports of share splitting in the 2003 Australian scheme of arrangement involving MIM – however, these were never proved.


Whilst the Court in Australia has the power to disregard the outcome of the vote required by the head count test, the authors recommend the removal of the head count test altogether in Australia and the UK for the following reasons:

  • Inconsistent with economic precept underpinning corporations legislation: for the purposes of the head count test, each registered member has one vote irrespective of the number of shares held. This is inconsistent with the economic precept underpinning the corporations legislation in Australia and the UK – that is, one share one vote – and neither the takeover bid provisions nor the capital reduction provisions in Australia or the UK contain an equivalent member agreement threshold to the head count test.
  • No policy support: the head count test was inherited from the original scheme provisions in the English Companies Act 1862, which only applied to arrangements between companies which were being wound up and their creditors. It has been speculated that the head count test was initially included “to place a check on the ability of creditors with large claims to carry the day”, which provides no policy support for the retention of this test today in connection with members’ schemes.
  • Confers disproportionate power: the head count test places significant power in the hands of members with small shareholdings, which power bears no resemblance or proportionality to their economic stake in the target company.
  • Share splitting can be very difficult to prove: as exemplified by the Dee Valley case, the head count test provides a temptation for share splitting and other devices designed to manipulate the outcome of a vote on a scheme. Even in Australia, where the Courts have the power to disregard the head count test, it appears that Parliament intended that this power would, generally speaking, only be used if there was cogent evidence that share splitting had occurred. However, as it can be difficult to prove, this power may be unlikely to completely remove the temptation to engage in share splitting.

The only arguments for retaining the head count test are that it provides a mechanism for seeking to ensure that a scheme only proceeds if the consideration proposed to be paid by the bidder is regarded as acceptable by a majority of target shareholders. However, the prevalence of nominee and custodian holdings today means that even these ideological objectives are often unattainable given the focus of the head count test on ‘registered’ members.

Equally, even proponents of the head count test would presumably regard it as anomalous (and perhaps even unfair) that a nominee or custodian who may hold shares on behalf of many beneficial owners only has one vote for the purposes of the head count test.

We are not alone in our view that the head count test should be abolished – law reform bodies in both Australia and the UK have made similar recommendations (being the Australian Corporations and Markets Advisory Committee in 2009 and the UK Company Law Review Steering Group in 2000).

Herbert Smith Freehills is acting for Severn Trent on the Dee Valley scheme.

Key contacts

Andrew Rich photo

Andrew Rich

Partner, Sydney

Andrew Rich