Altrad Services Limited & Anor. v The Commissioners for HMRC

The FTT found that both appellants entered into their transactions on arm’s length terms. In particular, the price that they received from SGLJ for the sale of the Assets was within a range of market values estimated by a third-party valuer (“American Appraisal”) ([44] and [59]).

At [52], [65] and [66], the FTT found that both Leases expired by effluxion of time, and that SGLJ exercised its Put Option, with the result that some three to four weeks after selling the Assets each appellant re-acquired the Assets.

At [258] the FTT concluded that, as a matter of law, both appellants gave up legal and beneficial ownership of the Assets when they sold them to SGLJ, and that they re-acquired legal and beneficial ownership on SGLJ’s exercise of the Put Option.

At [70], the FTT found that (tax saving aside) neither appellant realised any financial benefit from the arrangements. The sums that each received from SGLJ simply provided them with sufficient money to pay the Option Price due when the Put Option was exercised and re-acquire the Assets. While they did have the use of SGLJ’s money for the short period while each Lease was in place, taking into account the fees that they had to pay, they were incurring a “very high” annualised funding cost of between 15% and 30%. From that, the FTT inferred, at [75] that:

At [253], the FTT said:

At [255], the FTT found that exercise of the Put Option was effectively pre-ordained:

At [258], the FTT made the following finding as to the effect of the arrangements as a whole:

The appellants make a number of criticisms of the FTT’s conclusions in this regard which we will address later in this decision. We will not, therefore, at this stage set out the entirety of the FTT’s reasoning.

A flavour of the FTT’s reasoning on HMRC’s Ramsay argument can be found at [260] where the FTT decided that an application of the Ramsay line of cases led to the following conclusions:

We emphasise that the FTT’s conclusion summarised in paragraph 32(2) was premised on its finding that there was no cessation of ownership for the purposes of s61 of CAA 2001. That was precisely how HMRC had pleaded their Ramsay case in their Statement of Case before the FTT. HMRC’s case hinged on the proposition that the sale of the Assets did not result in either appellant ceasing to own the Assets for the purposes of s61. If, contrary to HMRC’s assertion, the appellants did cease to own the Assets when they sold them to SGLJ, HMRC were not advancing any fall-back argument, based on the Ramsay line of cases, that sought to deny the tax treatment for which the appellants argued, although they were making specific technical arguments based on s67 of CAA 2001 and the definition of “QA”. Put another way, in order to succeed with their Ramsay case before the FTT, HMRC needed to establish that the appellants did not cease to own the Assets, for the purposes of s61, when they sold the Assets to SGLJ.

The FTT’s conclusion on HMRC’s overarching Ramsay argument was sufficient to determine the proceedings in HMRC’s favour, but the FTT went on to consider the other arguments that had been advanced.

In Part E of the Decision, the FTT dealt with HMRC’s argument, summarised in paragraph 22(1) above, that the Lease fell outside the “long funding lease” regime because it, together with the associated Put Option, was a hire purchase arrangement falling within s67 of CAA 2001. That invited a consideration of whether, for the purposes of s67(1)(b), each appellant incurred capital expenditure “under a contract providing that the [relevant appellant] shall or may become owner of the plant and machinery on the performance of the contract”. Section 67(6) supplements this by providing for two or more agreements to be treated as a single contract if the effect of doing so would be to satisfy the condition in s67(1)(b).

HMRC argued that each Lease and associated Put Option should be treated as a single contract by s67(6) and that (i) that single contract provided that each appellant “shall” become the owner of the Assets, because the exercise of the Put Option was pre-ordained or, at the very least (ii) that single contract provided that each appellant “may” become owner of the Assets because SGLJ was entitled to require each appellant to purchase the Assets by exercising the Put Option. The FTT rejected that argument, reasoning at [292] that:

In Part F of the Decision, the FTT considered the issue summarised in paragraph 22(2), to which the FTT referred as the “QA issue”. The appellants considered that the Option Price paid under the Put Option fell within the definition of “QA” in s70E on the basis that each such payment was made by the relevant appellant “under a guarantee of any residual amount”, because payment of the Option Price effectively insulated SGLJ from any risk of the Assets falling to a value below the Option Price. The FTT accepted that argument and rejected HMRC’s contrary argument that (i) the only “guarantee” provided as part of the arrangements was the Parent Guarantee and (ii) the Option Price could not be both a payment falling within QA and a payment of capital expenditure attracting allowances under s11 of CAA 2001.

Neither the appellants nor HMRC make any challenge to the FTT’s factual findings.

We need not set out the appellants’ grounds of appeal or HMRC’s Respondents’ notice in any detail since it was common ground that, in order to dispose of these proceedings, the Upper Tribunal needs to address the following issues:

In their Respondents’ Notice, HMRC advanced a further strand of argument as “Issue 3”. However, by the time of the hearing before us, they had abandoned that argument. We retain the parties’ original numbering of the issues so that this decision can easily be read together with the skeleton arguments.

Issues 2 and 4 are self-explanatory. Issue 1 is deliberately framed in a limited way. As we have noted, the FTT accepted HMRC’s argument, based on the Ramsay line of authorities, to the effect that the appellants did not cease to own the Assets for the purposes of s61(1)(a) when they sold them to SGLJ and that therefore the tax analysis for which the appellants argued necessarily did not apply. Consistent with their position before the FTT, which we have summarised in paragraph 33 above, HMRC’s Response to the appellants’ notice of appeal makes it clear that their Ramsay argument is premised on the proposition that the sale of the Assets to SGLJ did not result in the appellants ceasing to own those assets for the purposes of s61(1)(a). For their part, the appellants did not seek, either before us or before the FTT, to challenge the proposition that, if s61(1)(a) was not engaged, the tax treatment they sought was not available. It follows that the sole battleground between the parties on Issue 1 is whether an application of the Ramsay approach means that the appellants did not cease to own the Assets when they sold them to SGLJ for the purposes of s61(1)(a) of CAA 2001.

We start with Issue 4 because that issue brings sharply into focus how the planning was intended to work, which will serve as a useful basis for our later discussion of Issue 1.

On the appellants’ analysis, the sale of the Assets followed by the grant of the Lease are neutral from a capital allowances perspective. The disposal value of £100 arising on the appellants’ analysis when the Assets are sold is matched by qualifying expenditure of £100 when each appellant enters into the Lease. Moreover, it is common ground between the parties that, if the appellants are entitled to any allowances at all when they enter into their Lease, those allowances are to be based on qualifying expenditure of £100.

The taxpayers’ argument for a step up in their capital allowances is therefore based on the interaction between the provisions that calculate the disposal value when each Lease terminates (s70E and s70YE) and those that determine their entitlement to capital expenditure when they reacquired the Assets following exercise of the Put Option (s11). In essence, the appellants argue that those provisions “misfire” (from HMRC’s perspective). They rely on the following chain of reasoning:

The additional qualifying expenditure that the FTT described as “magical” at [260(1)] is therefore argued to arise as a consequence of the way that the £95 Option Price enters the calculation at different points:

In Finance Act 2011, Parliament legislated to correct the potential for anomalies. Both parties were agreed that these subsequent changes cannot inform the proper construction of the legislation as in force for the periods in which the appellants implemented their transactions. We do not, therefore, rely on the Finance Act 2011 changes as an aid to the construction of earlier legislation, but simply note that, in s33 of Finance Act 2011, Parliament legislated to prevent any amount that could reasonably be assumed to constitute qualifying expenditure when paid from being taken into account either (i) in the calculation of “QE”, the qualifying expenditure under a long funding finance lease, or (ii) in the calculation of “QA”. Therefore, if the appellants had implemented their arrangements when s33 of the Finance Act 2011 was in force, on their analysis the arrangements would have produced no net benefit. This is because there would be a disposal value of £100 when the Assets were sold to SGLJ, allowances of just £5 under the Lease, no disposal value on termination of the Lease and qualifying expenditure of £95 under s11 when the Option Price was paid.

Before us, as before the FTT, HMRC argue for a different outcome. They accept that £95 is to be included within the calculation of “QE”, though they do not agree with the appellants’ analysis of why that is the case. However, they argue that the £95 should be excluded from the definition of QA for either or both of the following reasons:

We do not accept HMRC’s first argument, summarised in paragraph 47(1) above. While we acknowledge that the word “guarantee” can refer to the tripartite arrangement described in that paragraph, and indeed may generally do so when used in a legal sense, that is not the sense in which the word is used in s70E(2C)(b). That section asks whether the lessee itself has made a payment under a “guarantee”. The “person” referred to is clearly the lessee: see s70E(1)(a), which refers to a person who is the lessee, as well as s70E(2C)(a), which refers to payments by the person under the lease. The reference to the lessee demonstrates that Parliament cannot have had in mind a tripartite situation. A lessee could not guarantee its own obligations in the sense put forward by HMRC. That in turn demonstrates that the word “guarantee” is used in a different sense, as meaning an “assurance”, as the FTT held at [303].

We would observe that if HMRC’s argument were correct then the effect would be that whether any residual amount would be taken into account as part of the capital expenditure at inception of the lease under s70C would depend entirely on the happenstance of whether or not any residual value commitment was guaranteed by a connected person. This is because HMRC say that the reason that the £95 could count as part of the qualifying expenditure at inception of the lease, and as part of QE on termination, is that it was guaranteed by a connected person, within s70YE(1)(a) (read with s70C(4)). It would follow from this that, if HMRC were correct, then where there was a connected party guarantee allowances would be available under the long funding lease rules in respect of the present value of residual amounts, but allowances would not be so available if there was no such guarantee. Apart from lacking logic (as discussed further below), HMRC’s interpretation also ignores the reference to the lessee in the definition in s70YE(1)(a) (“guaranteed by him”, that is the lessee).

HMRC’s interpretation would have the effect that allowances in respect of residual value commitments would not be available in an entirely straightforward situation, namely where a lessee – whose creditworthiness may be such that no other entity is required to underwrite it – agrees to compensate the lessor at the end of a lease for any shortfall in value as compared to the amounts received by the lessor under the lease. Further, any payment actually made pursuant to that obligation would not be taken into account in determining disposal value.

HMRC’s second argument, summarised in paragraph 47(2) is, on close inspection, simply an assertion about anomalies that can arise if the Option Price does “double duty” both as a component of QA and as qualifying expenditure under s11. We accept that those anomalies could arise and Parliament evidently felt sufficiently concerned about the potential for such anomalies to enact the Finance Act 2011 provisions to which we have referred in paragraph 46 above. However, simply referring to the potential for anomalies does not demonstrate that Parliament intended QA to exclude any payment that could be treated as qualifying expenditure under s11 in circumstances where the legislation does not hint at such a qualification.

There are a number of real-world scenarios, not related at all to tax avoidance, where a person might transition from claiming allowances under the rules applicable to long funding finance leases to claiming allowances on the basis of being the outright owner of plant and machinery. At its simplest, a lessee of plant and machinery might purchase the equipment from its lessor for commercial reasons, either terminating the lease early or on expiry of the lease. In such circumstances, the scheme of the legislation suggests that there needs to be a “closing” figure under the long funding finance lease rules that helps to determine any disposal value under that regime, and an “opening” figure under the general rules applicable to outright owners of plant and machinery that enables allowances to be determined going forward. A fundamental difficulty with HMRC’s analysis is that, until the Finance Act 2011 changes to which we have referred, Parliament did not prescribe any link between these “closing” and “opening” values. HMRC’s submission that “QA” could not include items taken into account under s11 is a rationalisation based no doubt on what they consider to be the unwelcome effects of transactions of the kind effected by the appellants, but which has no obvious basis in the statutory language that Parliament has used.

In the scenario set out in paragraph 52, if the purchase was made pursuant to a self-standing offer made during the currency of or on expiry of a long funding finance lease, rather than pursuant to a commitment previously entered into, then the expenditure incurred would count only towards the “opening” value under s11, because it would not be a payment under a guarantee of any residual amount. There is no question of the treatment of this sum as qualifying expenditure under s11 giving rise to any step up in total allowances since no amount in respect of that sum would have been brought into account as qualifying expenditure under the long funding lease rules at the inception of the lease either (because it would not form part of commencement PVMLP), and it would also not be reflected in the calculation of disposal value because it would form part of neither QE nor QA in the “(QE – QA) + R” formula in s70C. Therefore, in a straightforward situation where a lessee purchases equipment leased under a long funding finance lease from its lessor, there is nothing objectionable about the purchase price being treated as qualifying expenditure under s11.

Alternatively, if – again in an entirely commercial situation – a put option had been granted by the lessee as part of the terms on which the lease was entered into, as a means of ensuring that the lessor would be able to ensure that it would be able to make its expected return, then in a similar way the price paid on exercise of that option would in principle qualify as expenditure under s11. As in the previous scenario, it is capital expenditure on the provision of plant and machinery for the purposes of the trade, and ownership is acquired as a result of incurring it.

However, on HMRC’s case (and prior to the enactment of Finance Act 2011), in this scenario the potential commitment to incur the expenditure would also have counted as part of the qualifying expenditure at inception of the lease, and therefore as part of QE, if it happened to be the subject of a guarantee from another entity. Therefore, on HMRC’s case, a special approach to QA, which involves excluding from it sums that count as capital expenditure under s11 even where they have contributed to the definition of QE, is to be applied in the case of the purchase of equipment by a lessee from a lessor under a long funding finance lease on exercise of a put option, where the put option in question has been guaranteed by a person connected with the lessee.

Quite apart from the fact that the legislation does not contain any provisions applying such a special approach to QA, there are two reasons why that approach is illogical and goes against the grain of the legislation:

We also agree with the appellants that the question posed by s70E(2C)(a) is simply whether the payment is made “under a guarantee of a residual amount”. Other than insofar as it incorporates the concept of residual amount (a concept which focuses on the financial effect of the transaction for the lessor rather than the lessee) that question is not otherwise concerned with the purpose of the payment. Prior to the enactment of Finance Act 2011, there was nothing in the statutory formula that excluded amounts that would otherwise qualify for relief either from QA or QE.

In addition, the effect of s70C of CAA 2001 is that the amount of qualifying expenditure under a long funding finance lease is determined on inception of the lease by reference to the discounted values of payments to be made under that lease. That is a forward-looking exercise that raises the possibility of a taxpayer being credited with too much qualifying expenditure if payments actually made under the lease are lower than those taken into account in the determination of qualifying expenditure. For example, a lease might terminate early before all scheduled payments due under the lease have actually been made. A lessee might not actually make all payments due under the lease. The “(QE – QA) + R” formula in s70C is a “truing up” provision that recognises that the disposal value arising on termination of the lease should claw back the benefit of capital allowances on anticipated payments which are taken into account in the determination of qualifying expenditure (QE) to the extent that those payments are not actually made so as to fall within the definition of QA.

The FTT thought that it was an “oddity” that the definition of QE included payments under residual value guarantees anticipated to be made by persons connected with the lessee, whereas QA embraces only payments made by the lessee itself (see [305]). We respectfully see no such oddity. Parliament is content, in the first instance, to give allowances by reference to expenditure that may fall to be incurred by a person connected with the lessee. However, if such a payment is not actually made by the lessee itself, the payment does not fall within QA, with the result that the truing up mechanism we have outlined results in the allowances given up front being clawed back. In our judgment, that is entirely consistent with the scheme of the capital allowances code generally, which typically provides for a taxpayer to obtain allowances by reference only to sums ultimately paid by the taxpayer itself.

We therefore see no justification for HMRC’s argument that, although the £95 anticipated payment falls within QE, it is not part of QA despite the fact that it was in practice actually paid. In our judgment, that interpretation is at odds with clear purpose of the “(QE – QA) + R” formula being to serve as the kind of “truing up” provision that we have explained in paragraph 58 above.

Conceptually, HMRC could have put their case on Issue 4 in a different way, arguing that the up-front arrangement for the appellants to purchase the Assets on expiry of the Lease was not an assurance (or “guarantee”) of a residual amount and so fell within neither QE nor QA. Such an argument would have been to foreshadow the outcome of the Finance Act 2011 challenges. We express no view on the strength or otherwise of such an argument, which was deployed neither before the FTT nor before us. For the reasons we have given, the FTT made no error of law in rejecting the arguments that HMRC did deploy, which we have summarised in paragraph 47 above. We accordingly determine Issue 4 in the appellants’ favour.

There is sometimes a tendency to refer to a Ramsay “principle” as if it were a judge-made principle applicable only in cases of tax avoidance. As the House of Lords noted at paragraph [34] of its judgment in Barclays Mercantile Business Finance Ltd v Mawson [2005] STC 1 (“Barclays Mercantile”):

However, since Barclays Mercantile the highest courts have repeatedly stressed that the approach simply involves the application of orthodox principles of statutory interpretation. In paragraph [32] of Barclays Mercantile itself, the House of Lords summarised the principle as follows:

The two-stage nature of the approach as involving both the proper construction of the statutory provision and the proper evaluation of the facts is brought out in the observation of Ribeiro PJ at [35] of Collector of Stamp Revenue v Arrowtown Assets Ltd [2003] HKCFA 46 (“Arrowtown”):

At paragraph [9] of Rossendale Borough Council v Hurstwood Properties (A) Ltd and others [2021] UK SC 16 (“Rossendale”), the Supreme Court emphasised that the approach is not particular to tax and is instead based on the “modern purposive approach to the interpretation of all legislation”. That said, the approach has a particular application in cases of tax avoidance. Not infrequently, tax avoidance schemes may seek to bring a transaction without any business or commercial purpose within the scope of a statutory provision intended to apply to a commercial transaction. Tax avoidance schemes may include transactions, or elements of composite transactions, which have no purpose other than tax avoidance. Where that is the case, the transactions in question will frequently be disregarded for tax purposes. However, any such disregard does not take place simply because the taxpayer’s purpose in effecting them is regarded as objectionable. As explained in paragraph [11] of Rossendale:

That statement in turn derived from decisions such as that of Lord Reed in UBS AG v Revenue & Customs Commissioners [2016] UKSC 13, where he said at [64]:

Therefore, transactions without a business purpose will be disregarded where that is consistent with a proper construction of the statute, or a proper ascertainment of the facts. However, the disregard of such transactions is not invariable. As Ribeiro PJ commented at [35] of Arrowtown:

There are a number of decisions of high authority that approach the issue by considering how composite transactions involving a number of individual steps should be taxed. Such an approach remains entirely permissible, though recent authorities explain it as an aspect of the two-stage approach we have already explained, rather than as a free-standing principle. For example, in paragraph [12] of Rossendale, the Supreme Court said:

Two such authorities relied on by HMRC in this case are Carreras Group Ltd v Stamp Commissioner [2004] STC 1377 (“Carreras”) and Scottish Provident v Inland Revenue Commissioners [2005] STC 15 (“Scottish Provident”).

Carreras concerned Jamaican transfer tax. The transfer tax was normally chargeable on “disposals” of shares but, by the terms of legislation that was based on the “reorganisation” provisions used for the purposes of UK capital gains tax, an exchange of shares for “debentures” was treated as involving no “disposal”. The taxpayer sold shares in return for a short-dated debenture, redeemable just a few days after issue, and claimed to have avoided the Jamaican transfer tax successfully because there was no “disposal” on which the charge could bite. The Privy Council disagreed. It concluded that the relevant transaction for the purposes of the Jamaican legislation was a combination, comprising both the issue of the debenture and its redemption. Once that was appreciated, there was no exchange of shares for debentures (viewed in isolation) and so the planning failed. In reaching that conclusion, the Privy Council noted at [13] that:

At [15] and [16], the Privy Council tested its conclusions by asking what precisely it was that led to the debenture and its redemption being treated as a single transaction. It answered that question as follows:

HMRC rely on this statement as applicable to the analysis of the combination of the Lease and the Put Option, submitting that the Lease also had as short a maturity “as was thought to be decent in the circumstances” and that the FTT found that the Put Option was pre-ordained to be exercised. That, HMRC submitted, supported their case that when each appellant sold their Assets to SGLJ, they did not “cease to own” those Assets for the purposes of s61(1)(a) of CAA 2001. We will address that argument in the sections that follow but, for the time being, simply note that Carreras does not give any guidance on what it means to “cease to own” assets for the purposes of s61(1)(a) or indeed on the meaning of “ownership” more generally.

Scottish Provident concerned a specific tax avoidance scheme that was intended to take advantage of a perceived shortcoming in new legislation relating to the taxation of options over debt instruments. As part of that scheme, the taxpayer granted a call option over particular gilts to a counterparty, Citibank, entitling Citibank to call for delivery of gilts at any time between 30 August 1995 and 1 April 1996. Citibank granted the taxpayer a call option entitling the taxpayer to call for delivery of the same gilts also at any time between 30 August 1995 and 1 April 1996. The transaction depended for its success on the proposition that Citibank had an “entitlement” to gilts under the call option that it held.

The Special Commissioners had found as a fact that differentials in the way the two options were priced meant that there was an outside, but commercially real, possibility that the two options would not be exercised so as to cancel each other out. The House of Lords declined to interfere with that finding of fact. However, they concluded that the outside chance that the Special Commissioners had identified was an “anti-Ramsay” device included only to avoid what would otherwise have been an inevitable conclusion that the two options would both be exercised so as to leave Citibank with no “entitlement” to gilts. That “commercially irrelevant contingency” could be disregarded and the composite transaction analysed in the same way as if it were not present.

HMRC rely on Scottish Provident in a similar way to that in which they rely on Carreras. They argue that the theoretical possibility that the Put Option might not be exercised was a “commercially irrelevant contingency” that should be disregarded. On that basis, they submit that all that happened was that the appellants sold the Assets on terms that they would reacquire them a few weeks later following a temporary and entirely tax-motivated transition in the appellants’ status from owner of the Assets, to lessee and back again.

As we have noted in paragraph 41 above, HMRC put their Ramsay argument in a specific and limited way both before the FTT and before us. That argument hinges on the proposition that the appellants did not cease to own the Assets for the purposes of s61(1)(a) when they sold them to SGLJ. We agree with the appellants that this argument can only be investigated by considering the meaning and purpose of s61(1)(a). Relevant facts must then be ascertained in order to decide whether the sale of the Assets fell within s61(1)(a) as purposively construed. That is not to say that the process of construing s61(1)(a) must take place in complete isolation from an appreciation of the facts, or vice versa. However, the FTT’s purely factual findings of the kind that we have summarised in paragraphs 27 to 30 above cannot on their own demonstrate the accuracy of HMRC’s Ramsay argument because they are not directed at the proper construction of s61(1)(a).

The FTT performed a lengthy and detailed analysis of a number of authorities relating to HMRC’s Ramsay argument, running to over 30 pages from [32] to [229]. Neither party made any criticism of this analysis. However, in our judgment, the length of the analysis suggests that the FTT may have viewed its task as being more difficult than it actually was. As we have explained, dealing with HMRC’s Ramsay argument involved simply deciding what s61(1)(a) meant, construed purposively, and then deciding whether in the light of all relevant facts, the appellants had ceased to own the Assets when they were sold to SGLJ.

There were clear signs in [236] to [244] that the FTT recognised the centrality of the exercise in statutory construction. At [240], it said that “The question is always one of what the statutory provision requires” (the emphasis being that of the FTT). At [242], the FTT directed itself that “There is simply no substitute for a close analysis of what the particular provision requires”.

However, in our judgment, the FTT lost sight of the fact that the “particular provision” it needed to construe, given the way that HMRC put their Ramsay argument, was s61(1)(a). While it did consider issues relevant to a purposive construction of s61 at [246] to [248] and in doing so made some pertinent observations, its overall conclusion at [249] that s61 was concerned with “real” disposal values did not answer the question. In particular, it did not address the appellants’ core submission, which they supported by reference to the judgment of the House of Lords in Melluish v BMI (No. 3) Ltd and related appeals [1995] STC 964 (“Melluish”), to the effect that any cessation of legal and beneficial ownership, whether temporary or not, was sufficient to cause the Assets to cease to be owned in the requisite sense. Indeed, there are indications that the FTT misunderstood the argument that the appellants were putting forward based on Melluish. Although Melluish was referred to at [90] and [97] in summarising the appellants’ submissions, the Decision contains no analysis of what, if any, light Melluish sheds on the interpretation of s61(1)(a).

Moreover, at [261], the FTT recorded its view that the appellants’ arguments required the FTT to apply a “formalistic step by step analysis in assessing the tax effects of the transaction” and to “focus narrowly solely on the legal effects of the sale of ownership rights in respect of the assets and, correspondingly, the re-acquisition of those rights”. However, the appellants were not arguing for any such “formalistic” approach to the analysis of the facts. Their point was that, properly construed, s61(1)(a) provided for the appellants to “cease to own” the Assets once legal and beneficial ownership was lost. The FTT concluded at [264] to [266] that HMRC’s Ramsay argument succeeded because of (i) the absence of any commercial purpose to the transactions, (ii) the fact that the appellants lost ownership of the Assets for just a few weeks, (iii) the fact that the appellants immediately on selling the Assets retained the right to use them under the Lease and (iv) the fact that exercise of the Put Option was effectively pre-ordained. There can be no objection to the FTT’s factual findings on points (i) to (iv). However, without an appropriate grounding in a conclusion as to the true construction of s61(1)(a), those factual findings are insufficient to support the FTT’s determination that HMRC’s Ramsay argument succeeded and the Decision contains an error of law in that regard.

We must, therefore, reach our own conclusions as to the proper interpretation of s61(1)(a). Before doing so, we note something of an oddity in HMRC’s approach to their Ramsay argument. As we have noted in our high-level summary of the capital allowances code in paragraph 5 above, where a taxpayer brings into account a “disposal value” that operates to the detriment of the taxpayer in the sense that the taxpayer’s entitlement to capital allowances is reduced and it may even become subject to a balancing charge. Where the disposal value arises because plant and machinery is sold to another person, it is true that the other person may be entitled to claim allowances on the basis that the purchase price paid constitutes capital expenditure under s11. However, that will not invariably be the case: the purchaser might not have sufficient profits to be able to use the allowances fully or may even be resident outside the UK. There is, therefore, something somewhat unusual about HMRC premising their Ramsay argument on a proposition that the concept of ceasing to own plant and machinery for the purposes of s61(1)(a) is narrower than it might otherwise appear to be. In his oral submissions on behalf of the appellants, Mr Peacock QC gave relatively little prominence to this point no doubt because he accepted that all statutory provisions, whether they benefit taxpayers or HMRC, are to be construed purposively. However, the oddity to which we refer does perhaps suggest that HMRC’s Ramsay challenge is not most naturally brought by reference to s61(1)(a).

In our judgment, the following features of s61 and the statutory code of which it forms part are relevant to the construction of the phrase “ceases to own” for the purposes of s61(1)(a):

The feature to which we refer in paragraph 82(1) above was considered in Melluish. Melluish concerned predecessor legislation contained in s44(1) of Finance Act 1971 which conferred entitlement to allowances where plant and machinery “belongs” to a person, rather than, as is the case under CAA 2001, where a person “owns” the plant and machinery. HMRC suggested that the fact that Melluish was concerned with a different statutory provision and the difference in terminology as between “belongs” and “owns” of itself prevented Melluish from being of relevance to the issue of construction raised in this appeal. We disagree. The concept of “belonging” in s44(1) of Finance Act 1971 was carried through into the Capital Allowances Act 1990. The change in terminology from “belonging” to “ownership” came in Capital Allowances Act 2001, one of whose purposes was, in our judgment, to re-write tax legislation to make it clearer and easier to use. We do not detect any intention for this change in wording to effect a substantive change in law.

HMRC also submitted that Melluish was concerned with whether plant and machinery belonged to a taxpayer whereas their Ramsay argument is concerned with the direct converse, namely when a person ceases to own plant and machinery. We reject that submission as in our judgment a consideration of what it means for assets to be owned by a person, or to belong to a person, will shed some light on the circumstances in which assets can be said to have ceased to be owned.

Melluish concerned a situation where a taxpayer (BMI) acquired plant and machinery and leased it to local authorities. The plant and machinery was affixed to land owned by the local authorities so that, as a matter of law, the local authorities owned the plant and machinery legally and beneficially. BMI argued that the plant and machinery nevertheless “belonged” to it for capital allowances purposes. It argued that it had paid for the plant and machinery, received rent from the local authorities for its use and that the leases with the local authorities gave it the right to require the plant and machinery to be returned on determination of the leases. Lord Browne-Wilkinson (with whom all of their Lordships agreed) rejected that argument noting (at 973g) that:

Lord Browne-Wilkinson also made the following points about the “belonging” condition at 973j to 974a:

Lord Browne-Wilkinson went on to conclude at 974d, fortified by the earlier decision of the Court of Appeal in Stokes v Costain Property Investments Ltd [1984] STC 204:

Those statements were not made in the context of the kind of transaction with which we are concerned. However, in our judgment, they provide a clear indication that, when determining whether the appellants ceased to own the Assets when they sold them to SGLJ, s61(1)(a) focuses on whether the appellants lost legal and beneficial ownership at that point. Moreover, the statement that the “belonging” condition should be “capable of a ready answer” points against HMRC’s argument that the existence or otherwise of a future “pre-ordained” plan for the appellants to re-acquire ownership of the Assets prevented them from ceasing to own them. On that interpretation, a taxpayer facing a disposal event under s61(1)(a) could assert to HMRC that, in fact, no such disposal value should be brought into account because, even though legal and beneficial ownership had been lost, there was a pre-ordained plan for the taxpayer to reacquire ownership. If that were right, the focus of the enquiry would become whether the arrangement truly was pre-ordained, with further difficult and uncertain questions arising if the intention of one party or another to participate in the pre-ordained plan changed.

HMRC, submitted, by way of analogy with the judgment of Megarry J in Sargaison v Roberts 45 TC 12 and the judgment of the Supreme Court in Balhousie Holdings Ltd v HMRC [2021] UKSC 11, that Parliament cannot have intended a transaction where “for a fleeting moment, a person’s technical status changes from owner to lessee before reverting to owning the asset” to fall within s61(1)(a). Like the FTT, we derive little assistance from Sargaison v Roberts, which was concerned with a very different statutory regime from CAA 2001. For similar reasons we derive little assistance from Balhousie.

HMRC argue that the point made in paragraph 82(2) is of no significance. They argue that, whether express reference is made to a person’s purpose or not, s61(1)(a) is concerned only with “real world” transactions and that in Melluish itself, the House of Lords based its reasoning on a need to avoid “anomalous results” (see 974d). While we accept that the point in paragraph 82(2) provides a more slender indication of the correct purposive construction, we do not accept HMRC’s submission that it is devoid of weight. We are prepared to accept HMRC’s argument that s61(1)(a) can, at a high degree of generality, be said to be concerned with “real world” transactions. However, HMRC have not explained why precisely Parliament provided in s61(1)(a) for the sale of the Assets not to be “real world” in the requisite sense. We quite understand that HMRC object to the tax analysis that is said to arise as a consequence of the Lease, the Put Option and its subsequent exercise. However, that does not, in itself, explain why, as a matter of construction, s61(1)(a) invites a consideration of the consequences of those transactions when deciding whether the appellants ceased to own the Assets when they sold them to SGLJ, in circumstances where those other transactions did not in fact prevent a cessation of legal and beneficial ownership.

Moreover, the “anomalies” to which HMRC refer arise as a consequence of the enactment of the regime conferring capital allowances on lessees under long funding finance leases many years after s61(1)(a) was enacted. We do not consider that these anomalies are a permissible aid to the construction of s61(1)(a). As Lord Neuberger said at [23] of Boss Holdings Ltd v Grosvenor West End Properties Ltd [2008] 1 WLR 289:

The indication we have summarised in paragraph 82(3) is perhaps of the least weight. However, it is consistent with other indications that we have drawn as to the purpose of s61(1)(a). The combination of the indications to which we have referred leads us to the conclusion that, construed purposively, s61(1)(a) provides for plant and machinery to cease to belong to a person where that person loses legal and beneficial ownership of the plant and machinery.

In the light of the conclusion we have just expressed, the relevant question for the FTT, given the way in which HMRC chose to put their Ramsay argument, was whether the appellants lost legal and beneficial ownership of the Assets when they sold them to SGLJ. The FTT answered that question in the appellants’ favour at [258]. In the light of that finding, it should have concluded that the requirement of s61(1)(a) was satisfied. The FTT wrongly considered that its findings as to the composite nature of subsequent transactions, and the appellants’ purpose in effecting those transactions, led to a different result. The problem with the FTT’s conclusion was not that its factual findings in this regard were wrong. Rather, they were not relevant to the enquiry that s61(1)(a), construed purposively, required.

HMRC support the FTT’s approach by reference to the judgments in Carreras and Scottish Provident. We accept that there are some analogies between the facts of those cases and the facts of the case before us. In Carreras, a particular tax result was said to flow from the interposition of a “debenture” with a short maturity. In our case, the short-term Lease is said to produce a result the FTT described as “magical”. In Scottish Provident, the House of Lords disregarded the “commercially irrelevant contingency” that an option over gilts would not be exercised. In this case, HMRC argue that the small chance of the Put Option not being exercised should similarly be ignored.

However, we do not consider that these cases justify the outcome for which HMRC argue. First, they were concerned with different statutory provisions. In Carreras, the Privy Council reached its conclusion by reference to the purpose of the relevant Jamaican statute imposing transfer tax. That gave rise to very different purposive considerations from those arising under s61(1)(a). In Scottish Provident, the House of Lords was considering the interpretation of the word “entitlement” whereas in this case we are concerned with the question of whether a person “ceases to own” plant and machinery in circumstances where other parties may be affected by the answer to that question, with a corresponding need for it to be “capable of a ready answer” as Lord Browne-Wilkinson put it in Melluish. In any event, the nature of the analysis in Scottish Provident was very different. In that case, the two cross options were both exercisable at the same points of time. Any gilts that Citibank sought to acquire on exercise of its option could immediately be taken away from it by the exercise of Scottish Provident’s option. The House of Lords held that, once commercially irrelevant contingencies were ignored, there was no point in time at which Citibank had an “entitlement” to the gilts. Our case is different. For a period of three or four weeks, the appellants did not have legal or beneficial ownership of the Assets, but SGLJ did. HMRC’s argument is that the sale of the Assets should not be treated as falling within s61(1)(a) because a later sale back of those Assets on exercise of the Put Option was pre-ordained. We see no reason why the very different analysis in Scottish Provident should apply in our case given the different statutory provisions with which we are concerned.

It follows that we determine Issue 1 in the appellants’ favour. We understand that some readers of this decision may find it surprising that an artificial series of transactions which, on the unchallenged findings of the FTT, were devoid of business purpose and effected only to achieve a “magical” increase in qualifying expenditure should survive a challenge based on the Ramsay line of cases. We stress that we have reached our conclusion based on the Ramsay argument that HMRC chose to put forward. Just as with Issue 4, it is not for us to comment on other ways in which the Ramsay argument could have been advanced, or the conclusions we might have reached if different arguments had been put forward.

Issue 2 raises a short point of statutory construction. Under each Lease, the appellants incurred capital expenditure on the provision of plant and machinery. The question is whether, for the purposes of s67(1)(b), that expenditure was incurred:

That is supplemented by s67(6). Since the appellants were both also party to a Put Option, it is necessary to consider whether, if the Put Option and the Lease were a single contract, the above requirement would be met. If it would, the Put Option and the Lease are to be regarded as a single contract. In effect, s67(1) and s67(6) requires an analysis of two questions, all to be performed by treating the Put Option and the Lease as a notional single contract:

For completeness, we note that both parties were agreed that the presence of the Call Option was not material to the analysis since neither appellant itself had rights under the Call Option. Instead, each Call Option was granted in favour of a company other than an appellant, albeit in that appellant’s group. It was common ground that s67 does not concern itself with rights that companies connected with the appellants had to acquire the Assets or with the likelihood of such connected companies acquiring the Assets in practice.

By the time of the appeal before us, there was no dispute about the answer to the question in paragraph 98(1).

Therefore, our focus must be on the question in paragraph 98(2). Importantly, the question asked is not about the general likelihood, or even certainty, of the appellants becoming the owner of the Assets. The FTT was right to emphasise, at [292(1)] that the question is about what the notional single contract provides and whether it provides that the appellants “shall or may” become the owner of the Assets.

HMRC argue that, once the FTT found as a factual matter at [255] that “barring wholly unexpected events” the Put Option would be exercised and the appellants would reacquire the Assets, it necessarily followed that the “shall” limb of the test was satisfied. We reject that submission since the notional single contract did not “provide” that the appellants would necessarily re-acquire the Assets. Rather, it gave SGLJ an option to sell the Assets back to each appellant. The practical certainty that SGLJ would exercise that option came not from the notional single contract but was a product of SGLJ’s own intentions and motivations, as the FTT found at [257].

HMRC also make the point that, if and when the appellants re-acquired the Assets, they would do so on “performance” of the Put Option and so on “performance” of the notional single contract. We agree, but that does not dispose of the issue. The “shall” limb of the test is not concerned with whether the reacquisition is on performance of the notional single contract. The test is whether that contract provides that the appellants “shall” become owners of the Assets. For the reasons we have given, it did not.

HMRC next submit that, at the very least, each notional single contract provided that each appellant “may” become the owner of Assets because each such contract contained a Put Option which, if exercised, would have had that result. That submission has more force since, as a matter of ordinary English, the word “may” can mean “might” and no stretch of language is involved in saying that the notional single contract “might” result in each appellant becoming owner of the Assets. However, we consider that this is not the sense in which the word “may” is used in s67.

That is clear from the scheme of s67 read as a whole. The purpose of s67 is to treat a person who is using plant and machinery and has incurred capital expenditure on its provision as an owner where the relevant contract provides either (i) that the user will become owner or (ii) gives the user an option to become owner. That is how typical hire-purchase contracts operate and the heading to Chapter 6 of CAA 2001 refers expressly to “hire-purchase and similar contracts”.

This point is brought out not just in the headings to Chapter 6 but also in the operative provisions. By s67(2) and s67(4) a user is treated as owner for so long as it has the “benefit of the contract”. That demonstrates that the focus is on rights held by the user of the asset. Such a focus also makes sense in policy terms. The clear intention is to permit owners of assets, and those who have the ability to secure ownership, to claim allowances on capital expenditure they incur for that purpose. Moreover, it is consistent with the need for the legislation to define clearly who is, or is not, to be treated as an owner of an asset and entitled to claim capital allowances on it. On HMRC’s approach to the word “may”, it might be necessary on each day during the life of a hire purchase agreement to assess the prospects or possibility of a put option being exercised to check whether it can still be said that the put option “may” result in the user becoming an owner. If the user and the person hiring the asset have different perspectives on this issue, the provision could be unworkable with both the user and the hirer of the asset asserting that they are “owners”. Alternatively, it might mean that allowances could be claimed by an entity that has no real ability to secure ownership, rather like the taxpayer in Melluish. By contrast, a focus on rights conferred on the user gives rise to no such difficulties: for so long as the user itself has a contractual right to acquire the asset, or itself has the option to purchase the asset, there is no difficulty in determining who is to be treated as owner.

For these reasons, which are in no material respect different from the FTT’s analysis set out at [292(1)-(3)], we determine Issue 2 in favour of the appellants.

For the reasons that we have given, the appellants have succeeded on Issue 1. The FTT erred in law in accepting HMRC’s Ramsay argument in the particular way that HMRC chose to formulate that argument. It should have rejected the Ramsay argument that HMRC put forward. The FTT made no error of law of the kinds that HMRC allege in respect of Issue 2 and Issue 4.

It follows, therefore, that we set the Decision aside. We remake it so as to lead to the result that the appellants’ appeals against HMRC’s closure notices are allowed.

In this appendix, we reproduce relevant provisions of CAA 2001 as in force in periods material to these appeals.

Section 11 of CAA 2001 provides the primary entitlement to plant and machinery allowances as follows:

Section 61 of CAA 2001 deals with disposal events as follows:

Section 67 of CAA 2001 deals with hire-purchase arrangements as follows:

Section 70A deals with the entitlement to capital allowances on long funding leases as follows:

It was common ground that, to the extent that the appellants were entitled to capital allowances for expenditure under a Lease, the rules applicable to long funding finance leases, set out in s70C were applicable. Section 70C provides, so far as material, as follows:

Section 70E provides for a disposal event to take place on termination of a long funding lease. It determines the amount of disposal value to be brought into account as follows:

Section 70J(3) provides for leases forming part of hire purchase and similar arrangements to be excluded from the definition of “funding lease” which, in turn, means that such leases cannot constitute long funding finance leases, as follows:

Section 70YE supplies a definition of concepts that are used both for the purposes of computing the amount of capital expenditure under s70C and for the purposes of computing “QA” in s70E as follows: