Article

It’s a dividend, Jim…

Key points

• Was a dividend, partly in specie, a distribution of income or capital?
• Does the source of the dividend affect its nature?
• Questions of foreign law are questions of law not fact.
• A review of relevant tax and non-tax cases.
• Could an appeal to the Supreme Court result in a different decision?

The Court of Appeal (CA) has recently published its judgment in Beard v CRC ([2025] EWCA Civ 385 which considered the nature of distributions paid to the taxpayer.

Mr Beard owned shares in Glencore, a company that was incorporated in Jersey, and therefore largely governed by Jersey law, although the company was tax resident in Switzerland, where its head office sits. During tax years 2011-12 to 2015-16, Mr Beard received dividends from the company with a total value of about £150m, part of which was represented by distribution in specie of some shares in Lonmin, a subsidiary of Glencore. The dividends were all paid out of Glencore’s share premium, as permitted by Jersey law, and Mr Beard said that the sums received were therefore capital and subject to capital gains tax. HMRC disagreed and assessed him to income tax.

The relevant UK legislation is ITTOIA 2005, s 402, ‘Charge to tax on dividends from non-UK resident companies’. The relevant sections state:

‘(1) Income tax is charged on dividends of a non-UK resident company’; and

‘(4) In this chapter “dividends” does not include dividends of a capital nature.’

Previous decisions

The questions before the First-tier Tribunal (FTT) and the Upper Tribunal (UT) were therefore, whether the payments were dividends and, if so, were they nevertheless excluded from income taxation as being of a capital nature? Both tribunals decided in favour of HMRC, that the payments were dividends and were not of a capital nature.

In the FTT ([2022] (TC8460)), both sides accepted that the in specie distributions of Lonmin shares should be treated the same as the cash dividends. However, in the UT hearing ([2024] STC 786), Mr Beard argued that this was a capital receipt, or a dividend of a capital nature, but the tribunal could see no reason to distinguish the in specie dividend from the cash dividends.

A major factor in the decision of the FTT was based on Jersey company legislation, which allows a company’s share premium to be reduced and paid out in two different ways. Part 12 of Companies (Jersey) Law 1991 (CJL 1991) allowed the share premium to be repaid by way of a reduction of capital, whereas Part 17 allowed it to be repaid as a distribution. All the payments to Mr Beard had been made under Part 17, which was influential in the decision that the payments were all dividends. The FTT then concluded, on the basis of the authorities before it, that while the ultimate source of the payments was a share premium account, this did not mean that the dividends paid out under Part 17 were dividends of a capital nature.

Court of Appeal decision

The CA was asked to consider three questions:

• Generally, whether the dividends were of a capital nature (described as ground 1); and
• in respect of the in specie dividends, were they correctly categorised as dividends; and
• if so, were they nevertheless of a capital nature (described as grounds 2A and 2B)?

On all three questions, the court found for HMRC.

The arguments

The contention for Mr Beard was phrased in diametrically opposed ways. On the one hand, the argument was that ITTOIA 2005, s 402 is a new piece of legislation bearing no obvious relationship to its predecessor, which simply charged income tax on income from non-UK possessions. Thus, all previous jurisprudence was irrelevant. Therefore, since the payments were all made out of the share premium of Glencore, even if they were dividends, they were clearly ‘dividends of a capital nature’, and therefore not chargeable to income tax.

Mr Beard’s alternative argument was that the tax law rewrite did not make any material difference to previous legislation, but that none of the previous jurisprudence had any bearing on the specific facts of his case. Again, therefore, since the payments were made out of share premium, they were clearly ‘dividends of a capital nature’.

HMRC’s counter argument was simply that the legislation was fundamentally unchanged, that pre-rewrite jurisprudence remains applicable and that the FTT (and UT) were right in finding both that the payments were all dividends and that they were not of a capital nature.

Foreign law – fact or law?

The CA first had to consider whether findings relating to foreign law were questions of fact or questions of law. While questions of law are always open to review by an appellate court, questions of fact cannot be impugned unless, as stated by Lord Radcliffe in Edwards v Bairstow and Harrison 36 TC 207, ‘the true and only reasonable conclusion contradicts the determination’. This is relevant because, in this case, the earlier decisions had been largely based on the fact that the dividends had been paid out under Part 17, CJL 1991 (distributions), not under Part 12 (reduction of capital).

The CA reviewed a number of cases, which suggested that questions of foreign law were questions of law in cases where foreign law was based on UK law (as would have been the case with Jersey law), whereas for legal systems that were not related to UK law, the findings were questions of fact. The CA, however, disagreed with this approach and decided that the interpretation of any foreign law was a question of law, and therefore, open to their review. Nevertheless, it accepted that the FTT’s decision had been based on the evidence given by expert witnesses for both sides, and that the CA must be very careful in deciding to overturn the FTT’s decision, as that evidence was not available to the CA to review directly.

Role of foreign law

The next step was to review the role of foreign law in interpreting UK tax statute or, in the context of this case, ‘the proper role of foreign law in determining whether an amount falls to be charged under s 402’.

The seminal case in this area is Rae v Lazard Investment Co Ltd 41 TC 1, where Lord Pearce said: ‘The factual situation (which includes the foreign law) has to be examined in order to apply the English law.’ Thus, the CA in Beard said, ‘foreign law must be considered in order to determine the nature and characteristics of the transaction’ and ‘the tax legislation must then be construed as applied to the facts’, which ‘requires the application of English (strictly here, UK) law’. The meaning and application of expressions such as ‘dividend’ and ‘of a capital nature’ are all questions of UK law, regardless of the fact that another legal system might refer to something as being capital in nature. Indeed, ‘the dividing line between income and capital is drawn by UK law, and the task of the fact-finding tribunal is to apply that law to the facts as found’.

Relevance of pre-rewrite authorities

But can we use the previous authorities, which looked at the old Schedule D, Case V charging ‘tax in respect of income arising from possessions out of the United Kingdom’, when construing a very different looking rule in ITTOIA 2005, s 402, which simply charges tax on dividends from a non-UK company, so long as they are not of a capital nature? The CA decided that we can.

The general principle behind interpreting legislation is that the words of the legislation itself should be used for interpretation, generally without reference to external sources. Nevertheless, it is accepted that some external sources can be useful in providing background and context. In this case, the CA referred to the preamble to ITTOIA 2005, which reads: ‘An Act to restate, with minor changes, certain enactments relating to income tax on trading income, property income, savings and investment income and certain other income; and for connected purposes.’ The CA says that this means that ‘parliament intended a continuity of approach’ with the previous legislation, which is also made plain in the explanatory notes to the Bill. Indeed, paragraph 189 in volume 2 of the explanatory notes (the case report incorrectly refers to paragraph 192) actually refers to previous case law and explains why, although no change in the law is intended in the rewrite, the wording is very different.

The CA’s conclusion was that it is relevant to consider the previous authorities, because it was clear from the explanatory notes and the preamble to the Act that there was no intended change in the law, so that those precedents remain relevant. This assertion is given further force in the specific case of s 402 by the fact that those precedents are specifically mentioned in the explanatory notes.

The precedents

The CA’s views on the relevance of pre-rewrite authorities effectively disposed of part of the argument for Mr Beard, that ITTOIA 2005, s 402 is a new piece of legislation bearing no obvious relationship to its predecessor. Therefore, the precedents relevant to Schedule D, Case V – which charged ‘tax in respect of income arising from possessions out of the United Kingdom’ – are relevant in deciding whether the payments by Glencore were dividends or were capital payments, and therefore not dividends (as the phrase ‘dividends of a capital nature’ was not invented until the rewrite of the rules in ITTOIA 2005).

The CA’s review of the precedents started with CIR v Reid’s Trustees 30 TC 431, where a South African company sold a fixed asset property and paid a dividend out of the capital profit. The House of Lords all agreed that, first, a dividend was prima facie income and, second, it was very relevant that the corpus or nature of the shares remained intact, implying that the payment was not capital. As Lord Normand said: ‘A distribution of money to shareholders out of profits realised by the sale of the company’s assets without any alteration of the share capital is normally a payment of the nature of income.’

While the mechanism by which the money was returned to shareholders was not obviously part of the reason for the decision in that case, it becomes more important in later precedents. It is also of more than passing interest to note the parallels with UK company law, because a reduction of a company’s capital (including share premium) under CA 2006, s 641 et seq, is either paid out directly to the shareholders, in which case it is a capital receipt, or it is taken to the distributable reserves (Companies (Reduction of Share Capital) Order SI 2008/1915), in which case any payment to the shareholders out of that result would, of course, be a dividend, and treated as income, regardless of the fact that the source of the payment was ultimately capital subscribed for the shares.

Non-tax and foreign cases

In the non-tax case of In re Duff’s Settlement v Gregson [1951] Ch 923 (CA), part of a company’s share premium account was repaid to shareholders, shortly after UK company law was changed (in 1947) so that share premium of UK companies was no longer distributable. (Before that change, the share premium was effectively seen as a profit on the subscription for shares, which could be paid out as a dividend. This was a total surprise to me, although it makes an odd kind of sense.) The CA in that case took the view that the new rule meant that the share premium was part of the company’s paid-up share capital, so that the repayment reduced the capital, and thus the corpus did not remain intact, and the payment was capital.

A similar decision was reached in Courtaulds v Fleming 46 TC 111, regarding a payment out of the share premium of an Italian company. Although 80% of the share premium was technically distributable, this was ‘only apparently on the footing that the distribution is treated as a return of capital’. Although the Italian law could not determine the UK tax treatment, the decision by Buckley J was that share premium was ‘an accretion to the tree and part of it’, not a distributable profit. This showed the importance of the foreign law – how the payment was treated domestically – and the importance of the mechanism – because Italian law demanded that the payment be treated as a return of capital.

Rae v Lazard Investment Co Ltd 41 TC 1 concerned a transaction by a Maryland company, called a ‘partial liquidation’, which looks quite similar to what we would now call a demerger by reduction of capital. This transaction could not have been effected as a dividend under Maryland law, which was an important finding of fact. The House of Lords decided that the transaction was a division of capital assets, not a dividend. Lord Reid said: ‘In deciding whether a shareholder receives a distribution as capital or income our law goes by the form in which the distribution is made rather than by the substance of the transaction. Capital in the hands of the company becomes income in the hands of the shareholders if distributed as a dividend, while accumulated income in the hands of the company becomes capital in the hands of the shareholders if distributed in a liquidation.’

Thus, the mechanism for a payment is essential in determining the position under UK tax law.

Does the dividend affect the shareholding?

Referring to the importance of whether the corpus remained intact (per the reason for the decision in Reid’s Trustees), the CA in Beard said, in Rae v Lazard: ‘The corpus of the asset did not remain intact: trees in Maryland could be split in two and still survive.’

The post-rewrite case of CIR v First Nationwide [2012] EWCA Civ 278 concerned distributions out of share premium by a Cayman Islands company. Under Cayman law, share premium could be paid to shareholders as dividends, subject to the company being able to pay its debts as they fell due (much like the current UK company law, as CA 2006, s 641 requires a statement of solvency before a reduction of capital can be implemented). The CA in that case said that it was well established by the jurisprudence that the mechanics of distribution determined whether a payment was income or capital. Thus, dividends out of share premium were still dividends, as they would be in the UK (see above). This case is similar to Beard, where the Jersey law allowed payment of share premium as a distribution (Part 17) or as a return of capital (Part 12), and the relevant payments were distributions.

An important element in the First Nationwide decision was the rejection of a previous description of a dividend as ‘a payment-out of a part of the profits for a period in respect of a share in a company’ (Harman J in Esso Petroleum Co Ltd v Ministry of Defence [1990] Ch 163, 165, adopted in the CA decision in Memec ([1998] STC 754).

Overall, the CA reached the conclusion that, ‘while mechanism is key, it is not necessarily conclusive in all cases. For example, a payment that takes the form of a dividend is prima facie income’, although the contrary is seen in Sinclair v Lee [1993] Ch 497 (see below). It was also important that the source of the funds is irrelevant – the capital profit in Reid’s Trustees was still a dividend when paid to shareholders, as was the payment of the share premium in First Nationwide. Finally, the labels applied under foreign law are not relevant to the UK analysis.

Decision on ground 1

Ground 1 questioned whether the payments were ‘dividends of a capital nature’. This phrase was a new legislative concept, introduced in ITTOIA 2005, s 402. While the CA found the concept ‘puzzling’, it had also explicitly rejected a previous general description of the dividend as being ‘a payment out of a part of the profits [of] a company’ and used Sinclair v Lee as an example of where a dividend might be capital in nature (although, as discussed below, I do not see that this case should have any relevance to tax legislation). However, the CA accepted that there were potentially other situations where the dividend could be capital ‘under less familiar systems of foreign law’. Indeed, the UT in Beard had explicitly stated: ‘It is impossible to envisage all the circumstances in which a company may pay a dividend, in particular when s 402 is concerned with companies incorporated under a multitude of foreign laws which may include procedures and arrangements unknown in the UK…’

The CA therefore considered it to be ‘unsurprising that the draughtsman wanted to make it clear that, if it was possible to have a dividend that was capital in nature, the charge to income tax under ITTOIA 2005, s 402(1) should not apply to it’.

The CA found that it was irrelevant that the distributions were paid out of capital subscribed for the shares, or that the share premium was described (labelled) as a capital account. Instead, it said: ‘The proper focus for our purposes is … on the fact that distributions may be made from share premium under Part 17, and that was the mechanism that was in fact used to make the distributions.’ The court further noted: ‘What is important is not the source but the mechanism. The wording of s 402(4) reinforces this, because the relevant question is whether the dividend is of a capital nature.’ Quoting the UT decision: ‘The focus of s 402(4) is on the character of the dividend, not of the funds from which the dividend is made.’

Although not explicitly stated in the judgment, the UT decided that the distributions were not of a capital nature because they were distributed using the Part 17 mechanism, as distributions, not under Part 12, as returns of capital. The mechanism ‘is critical to the UK tax analysis’.

Decisions on grounds 2A and 2B

These grounds of appeal concerned whether the Lonmin shares were ‘dividends’ and ‘of a capital nature’.

On ground 2A, the UT simply said: ‘No meaningful distinction can be drawn between it [the distribution of Lonmin shares] and the cash distributions in that respect. The Lonmin distribution was made using precisely the same mechanism as would be used, and was used, for ordinary cash distributions.’ It was not relevant that the shares did not represent a profit, or that they had been inherited by Glencore on a previous acquisition. And even if this were incorrect, ‘the Lonmin distribution was actually debited to share premium in exactly the same way as the cash distributions’, so there was no reason to distinguish the distribution of Lonmin shares from the cash distributions as being, for the purposes of s 402, dividends.

In considering whether they were ‘dividends of a capital nature’ (ground 2B), the CA turned to the case of Sinclair v Lee, which was a case involving trust law, not tax law. The trust in Sinclair v Lee was an interest in possession trust, so the income of the trust went to the life tenant and the capital accrued to the benefit of the remainderman. The trust held shares in ICI and ICI demerged Zeneca by way of a distribution in specie (an exempt distribution demerger, with special tax rules in CTA 2010, s 1074 et seq). Therefore, the shares in Zeneca were received by the trust as an income distribution, for tax purposes. However, the question was whether the receipt was capital or income for trust purposes.

It was noted that the value of Zeneca was some 55% of the pre-demerger value of ICI, which was very different to the demerger in Rae v Lazard, which was of a small part of the whole, and, indeed, in the current case, where the shares distributed by Glencore were a minor part of its assets. The decision was that treating the transaction as an income distribution for the purposes of the trust ‘would be to exalt company form over commercial substance to an unacceptable [extent]’. The commercial purpose of the demerger was ‘to replace a single head company with two head companies’ and the presumed intention of the settlor was that the shareholdings would constitute the capital of the trust, accruing to the benefit of the remainderman. Therefore, for trust purposes, the receipt of the Zeneca shares was found to be capital.

In contrast, the CA in Beard said that the Lonmin distribution was not ‘of a capital nature’. Rather, it was a distribution in specie by Glencore of (in relative terms) a fairly minor asset to its shareholders, which for tax purposes has the same income character as a distribution under Part 17 of an equivalent amount of cash.’

So, the CA found for HMRC on all grounds.

What’s next?

Mr Beard has applied for leave to appeal to the Supreme Court, so there may be another chapter to come in this tale. However, since the FTT, the UT and the CA have all found in favour of HMRC, and comprehensively so, it is difficult to see how the Supreme Court could come to any different decisions, even if permission is granted. Although it would not be unprecedented, my personal view is that this last appeal is simply a reflection of the amounts of money involved, given that the dividends concerned exceeded £150m in aggregate.

Comments

I was, firstly, intrigued by the diametrically opposed arguments put forward for Mr Beard: that ITTOIA 2005, s 402 was a completely new piece of legislation, so the old authorities do not apply; and that s 402 was intended to restate the previous legislation, but that the old authorities do not apply, even so.

When I first read about this case, I was intrigued about why the charge to income tax on non-UK dividends specifically excludes dividends of a capital nature, whereas the charge to income tax on ‘dividends and other distributions of a UK resident company’ (ITTOIA 2005, s 383(1)) has no such exclusion. Indeed, section 383(3) tells us: ‘It does not matter that those dividends and other distributions are capital.’

Fortunately for me, the matter was explained by the CA at the beginning of the judgment. The court noted that ITTOIA 2005 was enacted as part of the Tax Law Rewrite Project whereby ITTOIA 2005, s 402 replaced Schedule D, Case V, which charged ‘tax in respect of income arising from possessions out of the United Kingdom’. The explanatory notes say that the original Cases IV and V of Schedule D covered a variety of income from foreign sources and that the intention of the rewritten legislation was simply to charge those types of income to income tax in the same way as if they had arisen from a UK source.

However, it was difficult to create a precise overlap between UK sourced dividends and non-UK sourced dividends. This is because of the complex definition of a distribution for the purposes of UK taxation, which can include dividends or distributions of capital nature, or treat interest as distribution, while ‘any charge on distributions from non-UK resident companies must be confined to income only’ (paragraph 186 of the explanatory notes). This is why a separate charge for non-UK dividends was required.

I feel that I am better informed, but in a sense, none the wiser. Surely it would have been a simplification to tax income arising from the ownership of foreign companies in exactly the same way as analogous income from UK companies would be treated. But perhaps there is a nuance that I am missing here.

My major concern with this judgment is the proposition that Sinclair v Lee can be seen as an example of where a dividend can be ‘of a capital nature’. This is particularly an issue, since the case was about trust law, and was therefore decided partly on the basis of the presumed intentions of the settlor. The decision for trust law purposes was diametrically opposite to the tax law position, so that we have a distribution that was treated as capital in terms of trust law but is clearly an income distribution for the purposes of UK tax law. This, in itself, is a difficult circle to square.

I find it difficult to see how that case can translate into a comment on the impact of UK tax legislation on a non-UK transaction. As the CA said in Beard, it is important to understand the legal context of a foreign law transaction and to use that context to inform the decision as to how UK law applies to that transaction.

On that basis, and taking into account the authorities in this area, if Sinclair v Lee had been a tax case involving a non-UK company, and the legal form of the transaction had been a distribution in specie of the shares, out of the company’s profits, my instinct is that the distribution would be a dividend that is not ‘of a capital nature’ for UK tax purposes. The implication of the decision in Beard, however, is that the CA might have found that it was a distribution of a capital nature.

The other area I struggle with is the element of the amount in the analysis of Sinclair v Lee. It was noted in that case that the Zeneca shares represented about 55% of the value of ICI before the demerger. In contrast, it was noted that the shares in Rae v Lazard, and in Beard, constituted relatively small proportions of the assets of the distributing company. But a distinction based on proportionality is fraught with difficulties, particularly in borderline cases. Where does the border lie: is a distribution of 50% or more a distribution of a capital nature? Or is it 30%? Or some other number?

Equally, case law tells us that every case is decided on its own unique facts, so what other elements should be taken into account in deciding whether a distribution in a demerger is capital or income in nature? I guess only time will tell. 

‍