Home » United Kingdom » Supreme Court Cases » Scottish Widows plc No 2 v Commissioners for Her Majesty’s Revenue and Customs (Scotland)

Scottish Widows plc No 2 v Commissioners for Her Majesty’s Revenue and Customs (Scotland)

Scottish Widows plc v Commissioners for Her Majesty’s Revenue and Customs (Scotland). Scottish Widows plc v Commissioners for Her Majesty’s Revenue and Customs (Scotland)

Table of Contents

before

Lord Hope, Deputy President
Lord Walker
Lady Hale
Lord Neuberger
Lord Clarke

JUDGMENT GIVEN ON

6 July 2011

Heard on 16 and 17 May 2011

Appellant John Gardiner QC David Johnston QC Philip Walford (Instructed by Maclay Murray & Spens LLP)	Respondent Andrew Young QC Kenneth Campbell (Instructed by Office of the Solicitor to the Advocate General for Scotland)
Cross Appellant Andrew Young QC Kenneth Campbell (Instructed by Office of the Solicitor to the Advocate General for Scotland)	Cross Respondent John Gardiner QC David Johnston QC Philip Walford (Instructed by Maclay Murray & Spens LLP)

LORD HOPE

This is an appeal from an interlocutor of the First Division of the Inner House of the Court of Session (the Lord President (Hamilton), Lord Reed and Lord Emslie) in a joint referral to the Special Commissioners by Scottish Widows Plc (“the Company”) and Her Majesty’s Revenue and Customs (“HMRC”) under para 31 of Schedule 18 to the Finance Act 1998: [2010] CSIH 47, 2010 SLT 885, 2010 STC 2133. The question that was referred to the Special Commissioners for their determination was in these terms:

“Whether in computing the Case 1 profit or loss of Scottish Widows plc for the accounting periods ending in 2000, 2001 and 2002, amounts described by the company as ‘transfers from Capital Reserve’ and included as part of the entries at line 15 of Form 40 for each period fall to be taken into account [as receipts] in computing the profit or loss as the case may be.”

It is agreed that the words “as receipts”, which were not in the question as referred, may be understood as following after the words “into account.” The Special Commissioners answered that question in the affirmative. The Company appealed against that decision and HMRC cross-appealed. The Court of Session by a majority (Lord Emslie dissenting) refused the appeal and unanimously refused the cross-appeal. Both sides have appealed against its decisions to this court.

The essence of the dispute between the parties is whether, in each of the three consecutive years in question, part of the entry in line 15 of the Company’s form 40 must be taken as falling within the scope of either section 83(2) or section 83(3) of the Finance Act 1989, as substituted by paragraph 16 of Schedule 8 to the Finance Act 1995 and paragraph 4 of Schedule 31 to the Finance Act 1996. If it falls within the scope either of these two subsections, the sum concerned will fall to be treated as a chargeable receipt for the purposes of Case 1 of Schedule D in ascertaining whether, and if so to what extent, the Company made a loss during those years.

The Company carries on business as a life assurance company. Insurance business is a trade within the meaning of Case 1 of Schedule D: Income and Corporation Taxes Act 1988, section 18. The amounts to be taken into account in computing its profits include its investment income from its long term business fund and any increase in the value of its assets during the accounting period. Those profits may be computed for tax purposes in one or other of two ways. They may

be computed on the Schedule D, Case 1 basis, the actuarial surplus being a suitable starting point for dealing with cases of this description: see Scottish Union and National Insurance Co v Inland Revenue (1889) 16 R 461, 475, per Lord President Inglis. Or they may be computed on the basis of the income which the insurer receives on its investments less management expenses, known as the “I – E” basis. HMRC are entitled to elect to charge tax on the investment income, and they almost invariably do so as it nearly always pays the Crown to take the interest on the investments and not to trouble with the profits: Revell v Edinburgh Life Insurance Co (1906) 5 TC 221, 227 per Lord President Dunedin. But a Case 1 computation is nevertheless required in every case.

The dispute between the parties arises from the demutualisation in 2000 of the Scottish Widows Fund and Life Assurance Society (“the Society”) and the transfer, under a scheme sanctioned by the Court, of its business to the Company. The scheme came into effect on 3 March 2000. In para 22.1 it was provided that on or after the effective date the Company was to maintain a memorandum account within its long term fund, designated as the capital reserve, which was to represent the amount of the shareholders’ capital held within the long term business fund. The capital reserve was to be divided between the Company’s with profits fund and its non participating fund. Para 22.4 provided that the Company was to maintain records of the capital reserve and of the parts of it allocated to each of these two funds. While the funds comprised identifiable assets with all the qualities that attach to items of that kind, the capital reserve was a device or, as Lord Walker says in para 55, an abstraction. It was created for accounting purposes only and had no real life of its own.

At the time of the transfer to the Company the value of the Society’s assets was substantially in excess of its liabilities. But the Company sustained trading losses in each of the relevant accounting periods. The market value of its assets decreased from the inception of its long term business fund, due principally to falls in the value of the stock market. The Company claims that account should be taken of its commercial losses in its non participating fund during the relevant accounting periods. It has included Case 1 losses in its tax returns and computations for those periods equal to £28,689,437, £612,583,866 and £431,261, 757 respectively.

HMRC maintain that, on a proper construction of section 83(2) of the 1989 Act, which failing of section 83(3), and having regard to entries in the Company’s statutory returns for the relevant periods in which it was required to show that it had a surplus in excess of its liabilities for regulatory purposes, these claims should be disallowed. The Company brought various sums into account, described as transfers from the capital reserve, in the years in which they sustained losses. HMRC submit that each increase in the regulatory value selected by the Company falls to be treated as an increase in the value of its assets within the meaning of

section 83(2)(b). In any event the amounts brought into account and recorded as transfers from the capital reserve fall to treated as receipts under section 83(3) because they were amounts which had previously been added to the long term business fund as part of the transfer of business to the Company from the Society.

The statutory provisions

Section 83, as amended, so far as relevant to this case provides as follows:

“(1) The following provisions of this section have effect where the profits of an insurance company in respect of its life assurance business are, for the purposes of the Taxes Act 1988, computed in accordance with the provisions of that Act applicable to Case 1 of Schedule D.

So far as referable to that business, the following items, as brought into account for a period of account (and not otherwise), shall be taken into account as receipts of the period –

the company’s investment income from the assets of its long term business fund, and

any increase in value (whether realised or not) of those assets.

If for any period of account there is a reduction in the value referred to in paragraph (b) above (as brought into account for the period), that reduction shall be taken into account as an expense of that period.

In ascertaining whether or to what extent a company has incurred a loss in respect of that business in a case where an amount is added to the company’s long term business fund as part of or in connection with –

a transfer of business to the company, or

a demutualisation of the company not involving a transfer of business,

that amount shall (subject to subsection (4) below) be taken into account for the period for which it is brought into account, as an increase in value of the assets of that fund within subsection (2)(b) above.

Subsection (3) above does not apply where, or to the extent that, the amount concerned –

would fall to be taken into account as a receipt apart from this section,

is taken into account under subsection (2) above otherwise than by virtue of subsection (3) above, or

is specifically exempted from tax.”

In section 83(8) of the 1989 Act, as amended it is provided that the word “add” in that section, in relation to an amount and a company’s long term business fund, includes transfer (whether from other assets of the company or otherwise). Section 83A(1), which was inserted by section 51 of and paragraph 16 of Schedule 8 to the Finance Act 1995 and amended by paragraph 6 of Schedule 31 to the Finance Act 1996, provides that “brought into account” in sections 83 to 83AB (as inserted by paragraph 5 of Schedule 31 to the 1996 Act) means brought into account in an account which is recognised for the purposes of those sections. Section 83A(2) provides:

“Subject to the following provisions of this section and to any regulations made by the Treasury, the accounts recognised for the purposes of those sections are –

a revenue account prepared for the purposes of the Insurance Companies Act 1982 in respect of the whole of the company’s long term business;

any separate revenue account required to be prepared under that Act in respect of a part of that business.

Paragraph (b) above does not include accounts required in respect of internal linked funds.”

The revenue account that section 83A(2)(b) refers to is the regulatory return in form 40: see para 12, below.

Section 431 of the 1988 Act contains a list of interpretative provisions relating to insurance companies. They include a definition of the word “value”: see section 83(2)(b) of the 1989 Act. It is in these terms:

“‘value’, in relation to assets of an insurance company, means the value of the assets as taken into account for the purposes of the company’s periodical return.”

Insurance companies are under an obligation to submit annual returns to the Financial Services Authority (“FSA”) for regulatory purposes. The purpose of these returns is to demonstrate that the insurer meets the regulatory standard of solvency. They are required to show the results of a statutory actuarial investigation, which calculates the value of the insurer’s liabilities and identifies the amount of surplus in excess of those liabilities. They must show that there is a sufficient surplus to cover any declared bonuses. At the time of the demutualisation the relevant regulations were to be found in the Insurance Companies Act 1982, the Insurance Companies Regulations 1994 (SI 1994/1516) and the Insurance Companies (Accounts and Statements) Regulations 1996 (SI 1996/943).

Section 17 of the 1982 Act provides that every insurance company carrying on insurance business in the United Kingdom must prepare a revenue account for each financial year of the company, a balance sheet and a profit and loss account, the contents of which are to be such as may be prescribed by regulations. Section 18 provides for an actuarial investigation once in every period of twelve months of every insurance company which carries on long term business. Section 28 provides that it must maintain an account of the assets and liabilities attributable to its ordinary long-term insurance business. Regulation 45(6) of the 1994 Regulations provides that an insurance company may, for the purposes of an investigation to which section 18 of the Act applied, elect to assign to any of its assets the value given to the asset in question in the books or other records of the company. This was already a practice of long standing in the insurance industry.

For a detailed description of the background to these requirements and to the provisions of the Finance Act 1989 about the taxation of the life assurance business of an insurance company, reference may be made to Lord Reed’s opinion in the Court of Session to which, like Lord Walker, I would pay tribute: 2010 SLT 885, paras 91-126. A useful summary of the FSA regime that was in force at the relevant time is to be found in Lord Emslie’s opinion, para 198 (ix) to (xii); see

also Lord Walker, at paras 49 – 53, below. In short, the regulatory returns which a company carrying on long term life insurance was required to complete and submit included a series of numbered forms. Form 13 set out an analysis of all the company’s admissible assets, entered at a value which broadly corresponded to their year end market value. Form 14 set out in line 51 the amount by which the net admissible assets exceeded the company’s long term business liabilities. Form 40, which was headed “Revenue account”, set out revenue flows and expenditure for the Company’s long term business fund, its with profits fund and its non participating fund, and the amount of each fund to be carried forward to form 58. Form 58, which was headed “Valuation result and distribution of surplus”, determined the amount of the actuarial surplus by comparing the value of the insurer’s liabilities under the policies that it has issued with the fund shown on form 40.

The data that were used to prepare these regulatory returns were the same as those used to prepare the Company’s statutory accounts. The amounts that were calculated by the Company as the commercial losses of its non-participating fund were derived from decreases in the market value of the admissible assets less liabilities in that fund during the relevant accounting periods. For each of these periods, however, there were included in the figure entered as “other income” in line 15 of form 40 amounts described as “transfers from capital reserve” which reduced the capital reserve by an equivalent amount. They should perhaps have been included as an increase in the value of assets brought into account in line 13. But it is agreed that the way they fall to be treated does not depend on whether they were entered there or in line 15. The Special Commissioners said there is no difference in principle, as both lines are brought into account in the total shown at line 19: para 51. The amounts in aggregate for the relevant accounting periods were £33,410,000, £472,724,000 and £370,000,000 respectively.

The approach to construction

It is well understood that statutory provisions which bring profits and gains into charge to tax are to be construed as directed towards profits and gains in their natural and proper sense – in a sense which no commercial man would misunderstand – and that those words are equally applicable whatever the commercial concern may be: Gresham Life Assurance Society v Styles [1892] AC 309, 315, per Lord Halsbury LC. The objective is to ascertain and charge the true profits and gains of the business in question. The requirement that there should be a true and fair view involves the application of a legal standard. The courts are, in general, guided as to the content of the computation by expert opinions of accountants as to what the best current accounting practice requires: Revenue and Customs Commissioners v William Grant & Sons Distillers Ltd [2007] UKHL 15, 2007 SC (HL) 105, [2007] 1 WLR 1448, para 2, per Lord Hoffmann. The special rules that section 83 of the 1989 Act lays down for the calculation of the profits of

life assurance companies in respect of their life insurance business for the purposes of corporation tax are, in this respect, no different from the rules that apply to companies generally. They provide a legal standard according to which these profits are to be ascertained.

As has already been noted, that section has been amended more than once. But I do not think that it is helpful to look back into the legislative history. Lord Wilberforce said in Farrell v Alexander [1977] AC 59, 73 that self-contained statutes, whether consolidating previous law or so doing with amendments, should be interpreted, if reasonably possible, without recourse to antecedents, and that the recourse should only be had when there is a real and substantial difficulty or ambiguity which classical methods of construction cannot solve. Further amendments to section 83 were introduced by section 170 and paragraph 2 of Schedule 33 to the Finance Act 2003. In Inland Revenue Commissioners v Joiner [1975] 1 WLR 1701, 1715-1716 Lord Diplock said that it was a legitimate purpose of legislation by Parliament to clarify the law by making it clear in which of two alternative meanings the ambiguous language of an earlier statute was to be understood, but that it would only be if the language of a provision in an existing statute was ambiguous that it would be legitimate to infer that a purpose of the subsequent statute was to remove doubts as to what the law had always been. So the proper approach is to concentrate on the wording of sections 83(2) and 83(3) as they were at the relevant accounting periods.

With that background, and with the benefit of the much more comprehensive description of the facts that Lord Walker has provided and of the carefully reasoned opinions of all the judges in the Inner House, I now turn to the provisions of the 1989 Act which are under scrutiny in this case.

Section 83(2)

This subsection directs that there must be taken into account as receipts of the period for the purposes of Case 1 of Schedule D (a) the Company’s “investment income” from the assets of its long term business fund and (b) any increase in value of “those assets”, in so far as these items have been “brought into account” by the Company. The question is whether its language permits the Company to claim that it in fact sustained an allowable loss during the relevant period when the values “as brought into account” for that period indicate the contrary.

It is common ground that the reference to “investment income” in paragraph (a) of the subsection is a reference to actual income from assets actually comprised in the long term business fund. The Company submits that, by parity of

reasoning, the reference to “any increase in value” in paragraph (b) must be taken to be a reference to something that can be recognised on a commercial basis as a real increase in real assets. So the word “assets” in both paragraphs meant assets of the long term business fund which had the capacity to earn income and to grow in value. The fact was that its assets had decreased, not increased, in each of the relevant accounting periods. The amounts included in line 15 of form 40 were there for regulatory purposes only. They were book entries which had no commercial validity. The fact was that the assets of the long term business fund had decreased, not increased, in each of the relevant accounting periods. The mere fact that an amount, such as interest on unpaid tax, was entered in form 40 did not mean that it was taxable. To arrive at a true and fair view it was necessary to go behind the entries on the forms and look at the facts.

In the Court of Session the judges of the First Division were unanimous in accepting this argument and rejecting the argument for HMRC. The Lord President said that he was unable to accept that the contents of the revenue account that had been prepared for regulatory purposes had the definitional character for which HMRC contended. The fact that the investment income was inevitably an actual receipt suggested that the increase in value should be an actual sum, as opposed to an accounting element: para 54. Lord Reed made the same point in para 181, adding that the words “whether realised or not” were a strong indication that section 83(2) was concerned with real gains rather than a change in notional values. Lord Emslie said that, consistent with the long established distinction between “assets” and “fund”, the reference to an increase in value of assets should be taken as reflecting commercial reality in the form of actual increases in the value of assets: para 204.

As Mr Andrew Young QC for HMRC pointed out, however, that section 83(2) is a special rule for the computation of the profits of an insurance company in respect of its life assurance business. The general rules for the computation of profits and gains for the purposes of Case 1 of Schedule D must be taken to have been modified to the extent provided for in this subsection. The Company was being taxed on the I minus E basis and it is this subsection, not the rules that are generally applicable, that must be construed. An insurance company is entitled to elect, under regulation 45(6) of the 1994 Regulations, to assign to any of its assets the value given to the asset in the books or other records of the company. Section 83(2) can be taken to have been drafted in the light of the fact that insurance companies almost always, if not invariably, choose to use book values (in the sense indicated by regulation 45(6)) to arrive at the necessary balance in form 40 to demonstrate solvency to the regulatory authority.

Once this point is grasped, it seems to me that the meaning to be given to section 83(2)(b) falls fairly easily into place. The wording of the subsection follows that of the forms. While the investment income in paragraph (a) is real

income, the increase in value referred to in paragraph (b) may or may not be a real increase. The assets which gave rise to this increase in value may or may not be the same assets as those referred to in paragraph (a). It depends on the content of the amounts shown in lines 13 and 15 of form 40. Amounts taken from its long term business fund were used by the Company to supplement its trading income in each of the three years in question. It chose to use its own book values, not values computed according to the current value of the assets of its long term business fund, to arrive at the final values that were brought into account on form 40. In the absence of further directions in the statute as to how the increase in value is to be computed in cases where that option has been chosen – and there are none – I would hold that the increase in value referred to in paragraph (b) must be taken to be the amount which has been brought into account on the form.

The phrase “as brought into account for a period of account” in the opening words of the paragraph lies at the heart of this interpretation. It was suggested by the Company that this phrase determined the period for which items were to be treated as taxable receipts but not the items which were taxable. But this interpretation of the phrase does not, I think, give full weight to the word “as”. Linked to the words “the following items” which precede it, the phrase indicates that the computation must proceed on the basis of the way the items have actually been entered on the forms. If values shown in the books or other records of the company have been used, instead of market values, it will be the book values that will determine whether or not there has been any increase in value during the relevant period and, if so, how much that increase is.

The phrase “and not otherwise” was said to support the Company’s interpretation of paragraph (b) because it indicated that it was being assumed that the items that were being brought into account when any increase in value was being assessed were items that could be realised, not notional ones. But I think that their purpose is to make it clear that the basis of computation referred to in subsection (2) is the only basis that is relevant for the purposes referred to in subsection (1). The words “whether realised or not” are there to indicate a change from the computation indicated by the original wording of section 83. If the company chooses to bring unrealised increases in value into account, those increases in value must be taken into account as receipts for the period in the same way as increases that have been realised.

For these reasons I am unable to agree with the judges of the Court of Session as to the meaning and effect of section 83(2)(b). But in para 205 of his opinion Lord Emslie made some further points which, as they were attractively put, need to be answered too. He said that a factor which favoured the Company’s construction of section 83(2) was that it accorded well with the general principles

(1) that the ascertainment of receipts or gains for tax purposes should prima facie reflect commercial reality; (2) that income or gains to be taxed should prima facie

be the taxpayer’s and not those of a third party; and (3) that the ordinary recognition of shareholders’ capital to cover actual trading should not prima facie be a chargeable receipt. He described these principles in more detail in paras 197 and 198, and I agree with him that prima facie they can be taken to be a reliable guide as to how tax legislation ought to be construed. But his use of the phrase “prima facie” indicates, if I may say so quite correctly, that these are not absolute rules that are incapable of being disapplied by the statute. In this case we are dealing with special rules that have been designed to take account of the unique nature of the business carried on by life assurance companies. That in itself suggests that it is the language of the statute, rather than these general rules, that should be the determinative factor in this case.

Taking Lord Emslie’s three points in turn, I would hold, firstly, that the language of section 83(2) shows conclusively that, if the insurance company chooses to use book values to arrive at the final values shown on form 40, it is on those values that the computation referred to in section 83(1) must be based. This can be said to reflect the commercial reality of the life assurance industry, as the Company’s taxable receipts were based on its own figures as submitted to the regulatory authorities to justify the surplus of assets that it wished to recognise. Secondly, there is no question, in this case, of taxing the income or gains of a third party. The values brought into account on form 40 are the product of assets that were vested in the Company when it established its long term business fund. Their link with the Society was entirely broken when the transfer under the scheme took effect.

As to Lord Emslie’s third point, it must be appreciated that the capital reserve was not, as he said in para 202, ordinary shareholders’ capital. The words themselves might be taken as suggesting otherwise, but I think that the name that was given to what the scheme described as a memorandum account is a distraction. The reality is that the reserve had no life of its own separate from the long term business fund. It was an accounting mechanism which the Company had established for its own internal accounting purposes as part of its long term business fund. It did not consist of particular assets but was a financial structure which was subject to all the statutory restrictions and requirements to which that fund was subject. In para 205 he said that, as the capital reserve was shareholders’ capital, its ordinary recognition to cover actual trading receipts should not prima facie be deemed a chargeable receipt. But, as the capital reserve had no life of its own, amounts that were described as transfers from the reserve fell to be treated in the same way as any other assets comprised within the long term business fund for regulatory purposes and, in consequence, for the purposes of section 83(2) too.

For these reasons, and those given by Lord Walker, I would allow the HMRC’s cross-appeal.

Section 83(3)

As I would answer the question in the reference in favour of HMRC on the ground that the amounts in question fall to be taken into account as receipts under section 83(2) with the result that there was a corresponding increase in the assets of the long term business fund for each of the relevant accounting periods within the meaning of paragraph (b) of that subsection, the question whether section 83(3) applies to those amounts does not arise. This is because section 83(4) provides that subsection (3) of that section does not apply where, or to the extent that, the amount concerned is taken into account under subsection (2). But, as the judges of the Court of Session were divided in this issue and out of respect for the care which they took to examine it, I would like to make these few brief comments.

The exercise to which section 83(3) is directed proceeds in two stages which are, as Lord Reed said in para 191(1), conceptually distinct from each other. First, there is the question whether an amount has been added to the company’s long term business fund as a part of or in connection with a transfer of business to the company. Section 83(8) provides that the word “add” includes “transfer”. As for the facts of this case, amounts were added to the Company’s long term business fund when the scheme took effect as part of the transfer of the Society’s business to the Company. The whole of the amount that was to be treated as the capital reserve for accounting purposes was added or transferred to the Company’s long term business fund as an integral part of the scheme. It seems to me to be plain, having regard to the terms of the scheme, that the addition to the Company’s long term business fund was as part of or in connection with the transfer of the Society’s business to the Company. The fact that it was only later that some amounts were brought into account by way of what were called transfers from the capital reserve does not matter.

The second stage is the bringing of the amount into account for the period in question. It seems to me that this occurs as and when, and indeed whenever, the amount is brought into account as an increase in value to reduce or eliminate a loss that would otherwise have occurred during the relevant period. As Lord Reed said in para 191(2), there are understandable reasons why Parliament might consider that the use of amounts acquired on a transfer of business to offset liabilities resulting from normal patterns of trading which were not otherwise chargeable to tax should be disallowed. I agree with him that, when the amounts were subsequently brought into account on form 40, they would – but for the fact that they were already caught by section 83(2) – have fallen to be treated by section 83(3) as chargeable receipts for the purpose of ascertaining whether or to what extent the Company had incurred a loss in each of the relevant periods. Lord Emslie’s point, which he made in para 228 of his dissenting opinion, that the transfers were made in a non-chargeable context is answered by the two-stage nature of the exercise to which section 83(3) is directed. It is not the context in

which the transfer was made at the outset that determines the way in which the amounts fall to be treated when, at some later stage, they are brought into account.

Had it been necessary to do so to arrive at an answer to the question that was referred to the Special Commissioners, I would have affirmed the decision of the majority in the Inner House on this issue and dismissed the Company’s appeal.

Conclusion

I would recall the interlocutor of the Inner House of the Court of Session, allow the cross-appeal by the HMRC and answer the question referred to the Special Commissioners in the affirmative.

LORD WALKER

Introduction

On 3 March 2000 Scottish Widows plc (“the Company”), a new company within the Lloyds TSB banking group, acquired the principal assets and liabilities of the life assurance business of the Scottish Widows’ Fund and Life Assurance Society (“the Society”). The Society had a long and distinguished history. It was established in Edinburgh in 1814 “upon the principle of mutual assurance”. It was incorporated by statute in 1861 as a company without a share capital and it remained a mutual life office – that is an entity owned by its members, the policyholders, with no outside shareholders – until the change in 2000, which has been referred to as ‘demutualisation’.

The process of demutualisation was achieved by a scheme of transfer approved by the Court of Session under section 49 of and Schedule 2C to the Insurance Companies Act 1982 (“ICA 1982”). Some of the provisions of the scheme are of central importance to this appeal. The transfers which it effected were on a very large scale: the Company acquired, in round terms, assets with a market value of the order of £25bn and became subject to actuarial liabilities of the order of £19bn. The qualifying members of the Society received compensation of approximately £5.846bn, representing the difference (with various adjustments and enhancements) between the assets and the liabilities. This compensation was paid by the Company’s holding company, Scottish Widows Financial Services Holdings Limited (“Holdings”), which owns the whole of the Company’s issued share capital.
The Society’s assets included large holdings of equities as well as fixed- interest securities, immovable property and other investments. As it happens the United Kingdom stock market reached what was (and remains) an all-time high in the new year of 2000, and in the first years of the Company’s business the market value of its holdings of equities was substantially reduced. This unexpected and unwelcome turn of events has led to a dispute between the Company and HM Revenue and Customs (“the Revenue”) as to the tax consequences.

On 11 October 2006 the Company and the Revenue joined in making a referral to the Special Commissioners under Schedule 18, para 31 of the Finance Act 1998. The agreed question to be determined was as follows (with a small agreed explanatory addition):

“Whether in computing the Case 1 profit or loss of [the Company] for the accounting periods ending in 2000, 2001 and 2002, amounts described by the company as “transfers from Capital Reserve” and included as part of the entries at line 15 of Form 40 for each period fall to be taken into account [as receipts] in computing the profit or loss as the case may be.”

It is common ground that the answer to this question depends on two issues, one turning on the meaning and application of a general provision in subsection 83(2) (read with subsection (1)) of the Finance Act 1989 as amended (“FA 1989”), and the other turning on the meaning and application of a more particular provision in subsection 83(3) (read with subsection (4)) of the same section. The Company must win on both issues in order to succeed. Conversely it is sufficient for the Revenue to succeed if it wins on either issue.

The first issue, once understood, is a short point of construction. But for the non-specialist a lot of background, some of it quite technical, is required in order to understand the point, and to be able to weigh the linguistic arguments against more general considerations based on the legislative scheme and purpose. The second issue (which arises only if the Company is successful on the first issue) is a rather more intricate point of construction.

The complex background, and the large amounts of tax at stake, help to explain why these two points of construction took four and a half days before the Special Commissioners, and no less than seven and a half days before the Court of Session. The Special Commissioners (Mr J Gordon Reid QC and Dr John F Avery Jones CBE) decided the first issue in favour of the Company and the second issue in favour of the Revenue, so that the Revenue was successful. The First Division of the Inner House of the Court of Session (the Lord President (Hamilton), Lord

Reed and Lord Emslie) reached the same conclusions on both issues, unanimously on the first and with Lord Emslie dissenting on the second: [2010] CSIH 47; 2010 SLT 885; [2010] STC 2133. The Company now appeals on the second issue and the Revenue cross-appeals on the first issue.

The historical background.

The first issue (the subject of the Revenue’s cross-appeal) comes naturally before the second issue (the subject of the Company’s appeal). But before getting to the detailed arguments on either issue it is necessary to say something about the historical background, and to cover regulatory as well as taxation aspects, since these two aspects have become closely interrelated. The background has already been covered with conspicuous thoroughness and clarity in the judgment of Lord Reed (paras 87-104), to which I gratefully acknowledge my indebtedness. This part of my judgment is largely based on the fuller description by Lord Reed, with the addition of a few points of my own.

Life assurance, in its many different forms, has played an important part in British social and economic history. Actuarial science was already developing by the beginning of the eighteenth century (one of the founding fathers, Edmund Halley, published a paper on ‘The Degrees of the Mortality of Mankind’, commissioned by the Royal Society, in 1693). The Life Assurance Act 1774 addressed the problem of insurable interest and curbed the scandal of tontines, then fashionable in some wealthy circles.

Interest in life policies was by no means restricted to the wealthy. The Society was only one (and among the most prominent in Scotland) of many mutual societies by which working men could insure against the risk of their families being left in penury in the event of the early death or disablement of the main breadwinner. In England the most prominent comparable body was probably the Friends Provident Society, founded in 1832 (it was a registered friendly society, regulated by a different statutory system). The growth of these mutual societies was remarkable: they had just over 700,000 members in 1803, over 3½ million in 1887, and over 6½ million in 1910 (there are fuller statistics in D. Green, Reinventing Civil Society, 1993). The mutual movement went into decline after Lloyd George introduced a system of compulsory national insurance in 1911. The public interest in life assurance as encouraging prudent self-reliance was reflected in its tax treatment, though for the most part the incentives were directed to policyholders rather than life offices.

In 1870, after several life offices had run into difficulties, Parliament introduced a new system of regulation. It was the foundation of the more elaborate

system that we have today. The Life Assurance Companies Act 1870 (“LACA 1870”) required life offices (whether mutual or proprietary) to keep proper accounts and to prepare annual financial statements consisting of a revenue account and a balance sheet in a prescribed form. Regular actuarial investigations were made mandatory. Lord Reed explains in his judgment (paras 91-94) how section 4 of LACA 1870 introduced for the first time the statutory concept of a ‘life assurance fund’ held as security for the rights of holders of life policies and annuities. This was the origin of what is now referred to as a life office’s long term business fund (“LTBF”).

As regards taxation, during the 19^th century and the first two-thirds of the 20^th century there was no corporation tax and no capital gains tax. Companies were subject, in much the same way as individuals, to income tax assessed and charged under the various schedules and cases defined in the Income Tax Acts. If a taxpayer received income which could be regarded as falling within more than one schedule or case, the Revenue could not claim tax twice, but could choose which schedule to apply. This choice (sometimes referred to as the ‘Crown option’) was available to the Revenue in relation to proprietary life offices, which held large reserves of income-producing investments in order to meet their actuarial liabilities and provide for unforeseen contingencies. They could be taxed either on the profits of a trade under Schedule D Case I, or on their investment income as such. It was usually more advantageous for the Revenue to make an assessment on the company’s investment income, as Lord President Dunedin noted in Revell v Edinburgh Life Insurance Co (1906) 5 TC 221, 227. The Crown option was abolished by the Finance Act 2007 and replaced by mandatory provisions. With a mutual life office the Revenue never had a choice, since mutual trading does not produce profits taxable under Schedule D Case I.

The first statutory provisions giving special tax treatment to life offices were in the Finance Act 1915. Life assurance was to be treated as a separate business. Annuity funds were to be taxed separately from life funds. Life offices taxed on their investment income were to be allowed a deduction for management expenses (including commission paid to brokers). This system of taxation is generally referred to as the ‘I minus E’ (that is, income minus expenses) basis of assessment.

It remained open to the Revenue to choose to assess a life office to tax under Schedule D Case I, but the basis of that assessment was altered (and the likelihood of its actually being adopted by the Revenue was reduced) by section 16 of the Finance Act 1923 (“FA 1923”), which gave effect to a recommendation in the report, published in 1920, of the Royal Commission on Income Tax (Cmd 615). Profits allocated to with-profits policies were to be excluded from the life office’s taxable profits. This was not unprincipled, since on allocation the profits became liabilities. This provision has been re-enacted in successive consolidating

statutes, and finally in section 433 of the Income and Corporation Taxes Act 1988 (“ICTA 1988”), the terms of which are set out in para 103 of Lord Reed’s judgment. Section 433 of ICTA 1988 was repealed and replaced by FA 1989.

The change made by FA 1923 was an important change. In practical terms it diminished the difference in tax treatment as between proprietary and mutual life offices. Its importance increased with changes in economic conditions in Britain during the second half of the 20^th century (in brief, monetary inflation and the prospect of substantial capital gains from investment in equities and property). The Society was required by its constitution and regulations (to be found in their final form in the Scottish Widows’ Fund and Life Assurance Act 1980), as the Company is required under the scheme, to allocate to the holders of its with-profits whole-life and endowment policies nine-tenths of the with-profits part of the gains which it recognised, or brought into account (the expressions mean the same), in the revenue account of its LTBF. After the introduction of capital gains tax and corporation tax on chargeable gains, realised gains made by the Society were taxed at differential rates, the details of which are not material. But unrealised gains could be recognised (or brought into account) in order to enable larger bonuses to be allocated and paid to with-profits policyholders without having been taxed in the Society’s hands.

This was perceived by the Revenue as a serious defect in the system, as appears from an official consultation document published in 1988, The Taxation of Life Assurance, (summarised in paras 123-126 of Lord Reed’s judgment). This document gives a summary of how during the 1980’s the life assurance industry was rapidly evolving into being part of a larger savings industry, in competition with unit trusts and other savings media, and itself increasingly making use of unit- linked policies rather than traditional with-profit policies (investment by small savers in authorised unit trusts and approved investment trusts was encouraged in a different way, by deferring tax on capital gains until individual unitholders or shareholders realised their gains). Section 83 of FA 1989, which is at the heart of this appeal, was part of the changes which Parliament made in consequence of this review. It makes an express link between the imposition of liability to tax (or the creation of an allowable loss) under Schedule D Case I and the regulatory regime under ICA 1982. It is therefore necessary, before coming to section 83, to give a short account of the regulatory regime in ICA 1982 and regulations made under it.

ICA 1982 and regulations under it

The regulatory system introduced by LACA 1870 had been re-enacted and modified from time to time. ICA 1982 replaced it with a similar but much more detailed system, elaborated in a number of statutory instruments, of which the most relevant for present purposes are the Insurance Companies Regulations 1994 (SI

1994/1516) (“the 1994 regulations”) and the Insurance Companies (Accounts and Statements) Regulations 1996 (SI 1996/943) (“the 1996 regulations”).

Section 17 of ICA 1982 required every insurance company to which Part II applied to prepare with respect to each financial year of the company, a revenue account for the year, a balance sheet as at the end of the year and a profit and loss account (or for a mutual an income and expenditure account) for the year. Each of these was to be in a form prescribed by regulations. Under the 1996 regulations (as amended down to the year 2000) different forms were prescribed for different types of insurance companies, and they were still required by the new regulatory system mentioned in para 55 below. The form of balance sheet prescribed (by regulation 6) for companies carrying on long term business were forms 13 (relating to assets) and 14 (relating to liabilities). These together made up the two sides of the balance sheet. The form prescribed (by regulation 8) for companies carrying on long term business was form 40; if the company had more than one LTBF a separate account was required for each LTBF, and a consolidated form for all of them. The Company has three LTBFs, a with-profits fund, a non-participating fund for business taken over from the Society, and a non-participating fund for new business.

Section 18 of ICA 1982 required every insurance company to which it applied, and which carried on long term business, to cause its actuary to make an annual investigation of its financial condition, and to cause an abstract of the actuary’s report to be made. Assets were to be valued and liabilities determined in accordance with valuation regulations, and the abstract was to be in a form prescribed by regulations. Under regulation 25 of the 1996 regulations (as under the new regulatory system) the most relevant of the prescribed forms to be included in the abstract was form 58.

Regulation 45 of the 1994 regulations (as amended down to the material time) dealt with valuation of assets. After some general provisions in paras (1) to
it dealt specifically with actuarial investigations under section 18 of ICA 1982:

“(6) Notwithstanding paragraph (1) above (but subject to the conditions set out in paragraph (7) below), an insurance company may, for the purposes of an investigation to which section 18 of the Act applies or an investigation made in pursuance of a requirement under section 42 of the Act, elect to assign to any of its assets the value given to the asset in question in the books or other records of the company.

The conditions referred to in paragraph (6) above are –
1. that the election shall not enable the company to bring into account any asset for the valuation of which no provision is made in this Part of these Regulations;

that the value assigned to the aggregate of the assets shall not be higher than the aggregate of the value of those assets as determined in accordance with regulations 46 to 57 of these Regulations.”

Section 28 and 29 of ICA 1982 required separate accounts and funds to be maintained for long term business, and for the assets representing those funds to be applicable only for the purposes of the appropriate business, except so far as the value of the assets was shown, on a statutory actuarial investigation, to exceed the liabilities attributable to the fund.

I shall have to come back to the prescribed forms. I add one comment. Lord Reed observed (para 112), and I agree, that the use of the word ‘fund’ in ICA 1982 is not entirely consistent. Lord Reed had earlier quoted an observation of Lord Greene MR in Allchin v Coulthard [1942] 2 KB 228, 234:

“The word ‘fund’ may mean actual cash resources of a particular kind (e.g. money in a drawer or a bank), or it may be a mere accountancy expression used to describe a particular category which a person uses in making up his accounts.”

This is an important distinction, although Lord Greene’s reference to ‘cash resources’ is a little surprising and may have been influenced by the context of the particular case before him (it concerned the taxation of a local authority’s general rate fund).

In the context of life assurance a LTBF is a fund of investments of various types, and it falls within Lord Greene’s first category. The investments (the assets of the fund) change from time to time, as and when the investment managers need to raise money or exercise their judgment to switch investments, and the values of the assets fluctuate constantly. But at any time it is possible to identify the assets for the time being constituting the fund, which is a continuing entity. By contrast the Capital Reserve established by the scheme approved by the Court of Session and put into effect in 2000 (and here I am putting down a marker for later parts of this judgment), if it was a fund at all, was a fund in Lord Greene’s second sense. It was an accounting abstraction and it never consisted of identifiable assets.
Before going on to the scheme it is convenient to record, out of chronological sequence, that ICA 1982 was repealed by a statutory instrument made under the Financial Services and Markets Act 2000 which came into full force on 1 December 2001 (having come into force at earlier dates for limited purposes including rule-making powers). In consequence the new regime applied to the second and third of the Company’s accounting periods relevant to this appeal that is, the calendar years 2001 and 2002; the Financial Services Authority (“FSA”) became the regulator, and the system of regulation was prescribed by rules made by the FSA rather than by statutory instrument. But the substance of the system, and the identifying numbers of the forms, were unchanged. In particular, rule 9.10(c) of the FSA’s Interim Prudential Sourcebook for Insurers Instrument 2001 reproduced the effect of regulation 45(6) of the 1994 regulations. There were some minor changes of terminology in the forms, which were set out in Appendices 9.1 (forms 13 and 14), 9.3 (form 40) and 9.4 (form 58) of the 2001 instrument.

The scheme

The scheme for the transfer of the Society’s business to the Lloyds TSB group was preceded by an agreement dated 23 June 1999 between the Society and Lloyds TSB Group plc. The agreement provided for the scheme to be approved by a special resolution of the Society in general meeting (which duly occurred) and for an application to be made to the Court of Session for sanction of the scheme under section 49 of and Schedule 2C to ICA 1982. The Court of Session (Lord Nimmo Smith) sanctioned the scheme by an order made on 28 February 2000, and the scheme took effect on 3 March 2000. The scheme also obtained regulatory approval and tax clearances.

The scheme is lengthy and in parts very technical. It runs to 41 clauses and 12 schedules. In bare outline, the bulk of the assets and liabilities of the Society were transferred to the Company; pension policies and assets and liabilities associated with them were transferred to another Lloyds TSB group company and are not relevant to this appeal. Payment of the membership compensation to qualifying members of the Society (later quantified at £5,846m) was undertaken by Holdings, which is the owner of all the Company’s issued share capital. The provisions of the scheme which call for most attention are in Part D (Fund Structure) and Part E (operation of the Funds). Clause 22 in Part E (Capital Reserve) is of particular importance.

Under Part D (Fund Structure) the most basic division was between the LBTF (defined as the Long Term Fund) and the Shareholders’ Fund. The latter fund was to have allocated to it infrastructure assets and shares in seven subsidiaries and any joint venture companies (clause 15.1 and relevant definitions

in Schedule 1). All other assets (other than pension assets as mentioned above) were to be allocated to the LTBF, which was to be divided into two separate subfunds, the With Profits Fund and the Non Participating Fund (respectively the “WPF” and the “NPF”), with an appropriate allocation of existing policies (clauses 13 and 14.1). The allocation of assets between the WPF and NPF was to be determined by the actuary in accordance with the detailed provisions of clause

15.2 to 15.6, 15.10 and 15.11. Liabilities were to be similarly matched, subject to some special exceptions (clause 16).

In part E (Operation of the Funds) clause 18 deals with allocation of surplus arising in the WPF. One-ninth of the amount of bonuses allocated to conventional (that is, not unit-linked) with-profits policies (in other words one-tenth of the gross allocation) is to be allocated to the NPF or the Shareholders’ Fund, as the board directs. All other surplus is to be applied as bonus for the benefit of holders of with-profits policies. This replicates the position under the Society’s constitution and regulations (para 47 above). In life offices’ shorthand the WPF is a ‘90/10 fund.’ The NPF, by contrast, is a ‘0/100 fund’. Following each actuarial valuation of the NPF the board may transfer to the WPF statutory surplus arising in the NPF (clause 21).

Finally I come to the Capital Reserve, provided for in clause 22. Clause

22.1 is as follows:

“On and after [3 March 2000], [the Company] shall maintain a memorandum account within the [LTBF] designated as the Capital Reserve (the Capital Reserve). At [3 March 2000] the Capital Reserve shall represent the amount of the shareholders’ capital held within the [LTBF].”

Clause 22.2 provides for the Capital Reserve to be “credited with an amount” arrived at by a complicated formula. It is common ground that this amount was

£4,455m. Clause 22.3 (headed ‘Maintenance of Capital Reserve’) provides that no more may be added to the Capital Reserve and that it may be reduced only by being brought into account in the revenue account of the WPF (up to a limit arrived at by a formula) or the revenue account of the NPF (without limit). There does not seem to have been a finding or formal agreement as to the amount of the WPF limit, but the unchallenged evidence of Mr Adrian Eastwood, the Company’s actuarial director at the material time, was that the amount was £432m. Clause

22.4 to 22.6 provided for the Capital Reserve to be notionally allocated between the WPF and the NPF. The initial division was £1,895m to the WPF and £2,560m to the NPF. Tables B and C annexed to the agreed statement of facts and issues (“SFI”) show how the Company’s opening capital of £4,769m in its LTBF can be

reconciled with the opening Capital Reserve (£4,455m) and the membership compensation (£5,846m).

Clause 22A of the scheme provided for what was described as a contingent loan, free of interest, from the NPF to the WPF, repayable as mentioned in that clause. Its purpose was to compensate the WPF for the fact that its right to future profits could not be included, for regulatory purposes, as an asset with an admissible value. This inter-subfund loan (described in SFI, para 29) hardly featured in the parties’ written and oral submissions, but it is a further complication in understanding the regulatory forms, to which I now turn.

The forms

The balance sheet consists of forms 13 and 14. Form 13 sets out the values of the assets of the fund (that is, the LTBF or a subfund of it) at their admissible values (a technical term which in practice was not less than 99% of market value). The effective bottom line of form 13 is line 89, Grand total of admissible values.

Form 14 sets out liabilities and margins. For present purposes the most important lines are: 11, mathematical reserves (that is, actuarial liabilities which have not yet been finally quantified); 13, balance of surplus (or valuation deficit); 14, LTBF carried forward; 49, total other (ie non-actuarial) liabilities; 51, excess of the value of net admissible assets; and 59, total liabilities and margins. The entries at line 89 of form 13 and line 59 of form 14 must be the same. The balancing items on form 14 are the figures entered at line 13 (balance of surplus) and line 51 (excess of the value of net admissible assets, which is also called the ‘investment reserve’: SFI para 53(2)).

The interrelation between the figures at lines 13 and 51 of form 14 is that the line 13 figure is generally a relatively small amount representing value that has been brought into account but not yet finally appropriated. The line 51 figure is the true balancing figure, and is the last figure to be entered on the form. It represents the (generally very much larger) value that has not yet been brought into account at all – the amount by which the admissible value of the LTBF assets exceeds the book value that has been brought into account. It illustrates the proposition stated (perhaps in rather question-begging terms) in para 9 of the Revenue’s written case, that life offices are treated differently from most businesses in that they can shelter profits from taxation to meet unforeseen future liabilities. This point is discussed further in paras 82 to 86 below headed ‘Bringing assets into account at book value’.
Form 40, the revenue account, shows movements during the accounting period. The most important lines for present purposes are 11, earned premiums; 12, investment income; 13, increase (or decrease) in the value of non-linked assets brought into account; 14, increase (or decrease) in the value of linked assets; 15, other income; 19, total income; 29, total expenditure; 39, increase (decrease) in fund in financial year; 49, fund brought forward; and 59, fund carried forward (39

+ 49). The entry at line 49 must be the same as line 14 on form 14 for the previous accounting period, and the entry at line 59 must be the same as line 14 on form 14 for the current period. As to lines 13 and 14, Lord Reed explains in his judgment (para 116) that any increase or decrease in the value of linked investments (line 14) is required to be brought into account automatically, but unrealised increases in the value of non-linked assets (line 13) need not be brought into account.

Form 58 (valuation result and distribution of surplus) shows the actuarial surplus (line 29), its movement during the accounting period (lines 31, 34 and 35), and its distribution as between policyholders (line 46), shareholders (line 47) and balance (line 49, this being the same as line 13 on form 14). The term ‘distribution’ as used in lines 41-48 does not imply that sums necessarily leave the Company’s hands; it refers to an allocation as between policyholders and shareholders. Three separate forms 58 were completed for the WPF, the transferred business in the NPF and the new business in the NPF.

How the forms were completed by the Company

Volume V of the papers before the Court contains over 500 pages of the Company’s regulatory returns for the three relevant accounting periods, including completed forms 13, 14, 50 and 58 for the LTBF and its sub-funds (except that form 58 was completed, as already noted, for three sub-funds and not for the LTBF as a whole). From these forms the following information as to the whole LTBF can

be extracted (in £bn, rounded to adjustments in the computations).	the	nearest	£1m, and with	some rounding
		2000	2001	2002
Form 13
line 89: total assets at admissible value		23.066	22.427	20.962
Form 14line 11: mathematical reserves		19.128	19.807	18.645
line 13: balance of surplus		0.033	0.064	0.181
line 14: LTBF carried forward		19.162	19.871	18.827

line 49: total non-actuarial liabilities	0.441	0.386	0.468
line 51: excess of value of net	3.462	2.107	1.668
admissible assets	———	———	———
line 59: total liabilities and margins	23.066	22.427	20.962
Form 40
line 11: earned premiums	2.445	2.540	2.000
line 12: investment income	0.633	0.787	0.922
line 13: increase (decrease) in value of
non-linked assets brought
into account	1.273	(1.168)	(2.254)
line 14: increase (decrease) in value of
linked assets	(0.011)	(0.031)		(0.036)
line 15: other income	16.875	0.502		0.408
line 19: total income	21.216	2.631		1.040
line 29: total expenditure	2.054	1.921		2.084
line 39: increase (decrease) in LTBF	19.162	0.709		(1.045)
line 49: fund brought forward	.000	19.162		19.871
line 59: fund carried forward	19.162	19.871		18.827
Form 58 (WPF)
line 59: distributed surplus	0.633	0.915		0.576
line 61: percentage distributed to
policyholders	94.72	96.64		97.05

It would be imprudent to attempt any sophisticated commentary on these figures. The entry on form 40, line 15 for 2000 is obviously exceptional, representing the effect of a change of ownership of a long-established business; no one has suggested that the whole sum is taxable. But we know that it included a sum of £33.410m as a transfer from Capital Reserve (see para 70 below). Taken overall, the figures illustrate the effect of bringing into account value which, for prudential reasons, has not previously been recognised.

During the three accounting periods the admissible (for practical purposes, market) value of the assets of the LTBF fell by about £4bn (the figures can be collected from SFI, Table A and the Company’s completed forms 13 for the three accounting periods). The mathematical reserves decreased by a little under £0.5bn and the recognised value of the LTBF, tracking as it did the mathematical reserve and the unappropriated surplus, went down by about £0.3bn. But the investment reserve, that is the excess of admissible value over the recognised value of the LTBF (line 51 on form 14) was reduced by almost £1.8bn. The successive entries

on line 13 of form 40 are noteworthy. In the accounting period ending on 31 December 2000 the value of non-linked assets brought into account increased by over £1.2bn although their admissible value decreased during that period. This disparity was reversed in the two following accounting periods, during which (taken together) admissible value fell further by about £1.6bn but the form 40, line 13 decrease was a good deal larger, about £3.4bn. During the whole period the Company declared bonuses of significant amounts, and allocated more than the mandatory 90% to with-profits policyholders.

Part of the form 40, line 15 amounts included sums described in the notes submitted with the statutory forms (volume V, pp1756, 2035-2036 and 2319) as transfers from Capital Reserve. The amounts were as follows (SFI, paras 56-60):

	£m
2000	to WPF 33.410	to NPF	total 33.410
2001	30.724	442.000	472.724
2002	17.000	353.000	370.000
	81.134	795.000	876.134

Whether they should nevertheless have been brought into the computation of the Company’s profit or loss under Schedule D Case I under section 83(1) and (2) of FA 1989 is the first issue. Line 15 of form 40 is, it will be recalled, specifically mentioned in the referred question (set out at para 36 above).

The statutory provisions

The provisions which this Court has to construe are in a single section, section 83 of FA 1989. A rapid survey of the landscape in which that section is found shows that in the consolidating statute, ICTA 1988, Part XII dealt with special classes of companies and businesses, and Chapter 1 of Part XII dealt with insurance companies, underwriters and capital redemption businesses. I have already mentioned section 433, which was repealed by FA 1989 and replaced by similar (but more complex) provisions in section 82 of FA 1989. Section 444A, inserted into ICTA 1988 by the Finance Act 1990, applies to a transfer of long- term business in accordance with a scheme sanctioned under section 49 of ICA 1982, but neither side placed any reliance on this section. At the time when the consolidating statute was enacted the government was engaged in a far-reaching

review of the taxation of life offices, as already noted (para 48 above). The outcome was sections 82 to 90 of FA 1989 (together with Schedule 8 to that Act, amending Part XII of ICTA 1988).

These sections, and Schedule 8, were frequently amended between 1989 and 2000, especially by the Finance Acts of 1995 and 1996. The details are set out in Lord Reed’s judgment (paras 134 to 163). But I agree with Lord Hope (in para 15 of his judgment) that it is unnecessary, and maybe unhelpful, to go into the legislative history. What matters is the statutory provisions as they were in 2000, 2001 and 2002.

During that period section 83(1) to (4) was in the following terms:

So far as referable to that business, the following items, as brought into account for a period of account (and not otherwise), shall be taken into account as receipts of the period –

the company’s investment income from the assets of its long term business fund, and

any increase in value (whether realised or not) of those assets.

In ascertaining whether or to what extent a company has incurred a loss in respect of that business in a case where an amount is added to the company’s long term business fund as part of or in connection with

–

a transfer of business to the company, or

a demutualisation of the company not involving a transfer of business,

Subsection (3) above does not apply where, or to the extent that, the amount concerned –

would fall to be taken into account as a receipt apart from this section,

is taken into account under subsection (2) above otherwise than by virtue of subsection (3) above, or

is specifically exempted from tax.”

Section 83A(1) to (3) of FA 1989 was in the following terms:

“(1) In sections 83 to 83AB ‘brought into account’ means brought into account in an account which is recognised for the purposes of those sections.

Subject to the following provisions of this section and to any regulations made by the Treasury, the accounts recognised for the purposes of those sections are –

a revenue account prepared for the purposes of the Insurance Companies Act 1982 in respect of the whole of the company’s long- term business;

any separate revenue account required to be prepared under that Act in respect of a part of that business.

Paragraph (b) above does not include accounts required in respect of internal linked funds.

Where there are prepared any such separate accounts as are mentioned in subsection (2)(b) above, reference shall be made to those accounts rather than to the account for the whole of the business.”

It is common ground that the relevant revenue accounts are forms 40 for the whole LTBF and its constituent parts, the WPF and the NPF.

The first point of construction (which I have already described as a short point, but one which takes some getting to) is the meaning of “value (whether realised or not) of those assets” in section 83(2)(b). The Company contends that it means market value, and that any reduction in their value (the form of words at the end of the subsection) is to be treated as an expense capable of giving rise to an allowable loss. The Revenue contends that section 83(2) is referring to a difference in value (whether it be an increase or a reduction) as brought into account for the relevant period of account, and that section 83(A)(2) leaves no room for doubt as to what that means. It directs attention to the appropriate regulatory account, in this case form 40. The Lord President (para 54) described this approach as ‘definitional.’

Taxing a loss?

The Company’s written case before this Court, and Mr Gardiner QC’s robust oral submissions, characterised the Revenue’s position as unnatural, uncommercial and contrary to fundamental principles of tax law. The Court was reminded of some famous judicial observations made more than a century ago, including Lord Halsbury LC in Gresham Life Assurance Society v Styles [1892] AC 309, 315:

“The word ‘profits’ I think is to be understood in its natural and proper sense – in a sense which no commercial man would misunderstand”

and Lord Macnaghten in London County Council v Attorney General [1901] AC 26,35:

“Income tax, if I may be pardoned for saying so, is a tax on income. It is not meant to be a tax on anything else.”

In this case, Mr Gardiner submitted, the Revenue was attempting to tax what was in reality a loss of capital.

These submissions call for careful consideration. The massive volume of documents and figures put before the Special Commissioners and the Court of Session, and now before this Court, creates a risk of getting lost in a labyrinth of abstractions. Actuaries, accountants and lawyers are trying to converse in the same language, but it is not easy going. It is a case in which there is a real danger, in the hackneyed phrase, of not seeing the wood for the trees. It may help to avoid confusion to start with three simple points.

The first point is that the Revenue is not seeking to exact tax from the Company under Schedule D Case I either on profits or on losses incurred by the Company; it is taxing the Company on the I minus E basis. Simultaneously the Company is seeking to establish large Schedule D Case I losses in order to have them available for surrender to obtain group relief. The second point is that it is, and always has been, standard practice for life offices to bring the assets of their LTBFs into account, not at market values that fluctuate from year to year, but at a book value (though in practice that expression is applied to LTBFs in a way that an outsider may find surprising). The third point is that the Capital Reserve is not, and never has been, a separate fund distinct from the Company’s LTBF. It has always been part of the LTBF. Each of these three points calls for some further explanation.

The Crown option as it applies to this case

The Revenue is not seeking to charge tax under Schedule D Case I on losses incurred by the Company. It is common ground (SFI, para 61) that at all material times since 3 March 2000 the Company has been taxed on the I minus E basis (the detailed computations for 2000 and 2002 can be seen in volume VII at pages 3211 and 3290; the relevant page for 2001 seems to have been inadvertently omitted). Nevertheless (SFI paras 62 and 63) the Company seeks to claim an allowable loss under Schedule D Case I which would be available for surrender to other Lloyds TSB group companies by way of group relief.

The Revenue accepts (SFI, paras 62 and 63) that if the Company succeeds in this appeal the losses available for surrender would be approximately £28.7m for 2000, £612.6m for 2001 and £431.3m for 2002 (the relevant computations are

at volume VII pages 3216, 3255 and 3295). The fact that a proprietary life office can simultaneously pay tax on the I minus E basis and have an allowable loss under Schedule D Case I shows that whatever the position a century ago, when there were no special statutory provisions, the taxation of long term life assurance business is now a very specialised area.

Bringing assets into account at book value

Regulation 45(6) of the 1994 Regulations (set out in para 20 above, and later reproduced in Rule 9.10(c) of the FSA’s 2001 instrument) allowed a life office, for the purposes of an actuarial investigation, to take the value of any of its assets as its value “in the books or other records of the company”. This had been expressly permitted by the regulatory system since 1980, when Regulation 3 of the Insurance Companies (Valuation of Assets) Regulations 1976 (SI 1976/87), was amended by the Insurance Companies (Valuation of Assets) (Amendment) Regulations 1980 (SI 1980/5). But the two expert witnesses agreed that it was a very long-standing and well-established practice, and the Special Commissioners made a finding to that effect (para 16 of their decision).

It would be potentially misleading to say that a life office is permitted to bring the assets of its LTBF into account at book value, since that is normally understood to mean historic cost. In a LTBF some assets are normally brought into account at the full admissible value, and others at nil (Special Commissioners’ decision, para 48; also para 122 of Lord Reed’s judgment). It is unnecessary to go into the reasons for this practice, as to which there was no dispute.

The reasons for maintaining an investment reserve of unrecognised value are fundamental to the way in which long term life business, and especially with- profits business, has been conducted in the United Kingdom. It is the mechanism by which the life office, relying on the professional skills of its chief actuary and his staff, can achieve a balance between competing considerations and interests. First and foremost is the overriding need for a sufficient margin of solvency. Subject to that the life office will wish to produce consistently good results for its with-profit policyholders, both in the policyholders’ interests and to preserve and enhance the company’s reputation. It must also achieve fairness between different classes of policyholders in accordance with their rights and expectations (the difficulties of which are illustrated by Equitable Life Assurance Society v Hyman [2002] 1 AC 408). Finally there are tax considerations. No company likes to pay more tax than it has to, or to pay it sooner than it has to. Before 1989 the tax system allowed life offices to defer taxation, especially on unrealised capital gains. It is common ground that section 83 of FA 1989 was intended to change that; the controversy is as to the extent of the change.
These points were well made by Mr Brian Drummond, an accountant, in an article entitled ‘Making Sense of the FSA Return in Life Company Tax Computations’ (Tax and Accountancy Review, June 2006, p6). Some changes had taken place by then (both on the regulatory front and the taxation front) but the article is nevertheless instructive. After mentioning recent changes the author gives a brief overview of the forms:

“In broad terms, however, the overall structure remains unchanged.

Form 13 remains a reasonably straightforward analysis of the total admissible value of the assets of the company by category with narratives that are commendably clear;

Form 40 demonstrates how much of the Form 13 value is brought into account for the purposes of calculating surplus;

Form 58 deals with the calculation, composition and distribution of the surplus; and,

Form 14 then links that exercise back to Form 13 by showing how much of that original Form 13 value is covering liabilities and bonuses and how much of it is being held in reserve.”

He describes form 14 as ‘an area of linguistic opacity’, and comments:

“This confusion is carried across into form 14 of the FSA return where it increases further. The first line in form 14 is described as ‘mathematical reserves, after distribution of surplus’ and in this one narrative only two of the six words (‘after’ and ‘of’) take their conventional or even accounting meaning.”

The most relevant passage is on the general philosophy of with-profits business (at pp 9-10):

“Form 40 is described as ‘revenue account’ but in conventional terms it is a very partial one. By reference to normal accounting

convention it is surprising to have a revenue account that makes no explicit reference to a movement in liabilities to third parties. The layout of Form 40 and its interaction with Form 58 reflects much more of the history of with profit funds than it reflects normal accounting principles.

In with profits funds the starting point in determining the extent to which surplus is recognised is establishing what bonus should be recommended. This will be driven by a combination of the results of the company (in terms of investment return and underwriting profit) together with policyholder reasonable expectations and the need to treat customers fairly. One of the principles of UK with profits business is smooth bonuses from year to year.

Having established what bonus it is appropriate to declare for the year it is then possible, depending on the structure of the fund, to calculate the minimum extent to which surplus must be recognised – both to meet the bonus requirement and any corresponding entitlement of the shareholders to participate in surplus as a fraction of the amount allocated to policyholders (very often one-ninth the 90:10 structure). Historically with profit funds hesitated to recognise any more surplus than was required to meet the bonus, and associated shareholder entitlement, and hence the fund would generally be approximately equal to the liabilities (after current year bonus) plus any residual surplus not allocated.”

The nature of the Capital Reserve

The third point mentioned in para 78 above is that the Capital Reserve is not, and never has been, an appropriated fund separate from the Company’s LTBF. It is, as para 22.1 of the scheme makes clear, part of the LTBF. It is an account falling within Lord Greene MR’s second category in Allchin v Coulthard [1942] 2 KB 228, 234-235 – “merely an accounting category”.

Abstract though it is, the Capital Reserve is on the Company’s case of central importance to this appeal. It is not easy to discern its purpose. The Company’s own independent actuarial expert, Mr Chamberlain, stated in his report dated 18 September 2007, para 5.1:

“The Scheme by which [the Society] demutualised established something it refers to as a Capital Reserve. This ‘Reserve’ is a

financial structure whose form and operation is defined by the Scheme, and does not meet any particular regulatory or other requirement, other than that emanating directly from the Scheme. It is a memorandum account and does not consist of particular assets.”

Mr Allen, the Revenue’s independent expert, stated in his report dated 5 October 2007, para 6.1:

“Within the notes to their returns I understand that [the Company] created a memorandum account (the Capital Reserve) with an initial balance of approximately £4.5bn. Notwithstanding that this account was referred to in the Scheme which obtained approval from the Court of Session, in my opinion this memorandum account had no meaning or relevance, other than as an item of information, as regards either the Company’s statutory report and accounts or its regulatory returns. The memorandum account did not represent any particular assets, nor did it reflect any actual profit or loss incurred by the Company, it was simply a note of a particular transaction.”

The experts did not give a further explanation of the expression ‘memorandum account’, nor did counsel offer any. The Special Commissioners made a finding that reflects the natural meaning of ‘memorandum’ (para 45): “The purpose of the Capital Reserve was to keep a record of this initial value created by [Holdings] and to distinguish it from subsequent profits.” The notion that it was an item of information that ought to be remembered comes out most clearly in the witness statement of Mr Michael Ross. He was an actuary who was employed by the Society for most of his career, becoming chief actuary of the Society in 1986 and the first chief executive of the Company in 2000. In his witness statement (paras 18 to 27, not challenged in cross-examination) he described how demutualisation involved a strategic choice between ‘ring-fencing the estate’ and ‘monetising the estate’ (the estate is a term used to describe a mutual’s excess of assets over liabilities, or investment reserve).

After careful thought the Lloyds TSB group and the Society opted for monetising the estate. This course was likely to be more attractive to the Society’s members but required the Lloyds TSB group to find a very large sum to pay the membership compensation. But the payment of that compensation gave the Lloyds TSB group the advantage that the Company had a comfortable investment reserve at the inception of its business (whereas with ring-fencing the group might have had to inject further capital). The group wanted to earmark what Mr Ross (para 27) regarded as “shareholder-owned capital, held within the [LTBF]” in order to be able, in the long term, to benefit from it.
That provides a clue, I think, to the purpose of the restriction on reduction of the Capital Reserve in para 22.3 of the Scheme (summarised in para 60 above). Only a limited amount (£432m out of £4,455m) of the Capital Reserve could be brought into account in the revenue account of the WPF, because it was a 90/10 fund and nine-tenths of the distributed surplus were to go for the benefit of holders of with-profit policies; only one-tenth (at most) could find its way to the shareholder, Holdings. There was no restriction (beyond the total amount of the Capital Reserve) on bringing it into account in the NPF, which was a 0/100 fund.

The decision of the Special Commissioners and the judgments in the Court of Session

The Special Commissioners decided the first issue in favour of the Company, and the Court of Session unanimously upheld that decision (Lord Emslie’s dissent was on the second, narrower issue as to section 83(3)). All three members of the Court delivered full judgments, so this Court has four separately reasoned routes to the same conclusion on the first issue. The reasoning can be imprecisely classified under three heads: the correct approach to the construction of taxing statues, arguments based on the legislative scheme and purpose, and detailed linguistic arguments.

The Lord President dealt most fully with statutory construction (paras 45 to 49). He cited the well-known speech of Lord Steyn in Inland Revenue Commissioners v McGuckian [1997] 1 WLR 991, 999–1000, in which Lord Steyn referred to Lord Wilberforce’s seminal speech in W T Ramsay Ltd v Inland Revenue Commissioners [1982] AC 300, 323:

“Lord Wilberforce restated the principle of statutory construction that a subject is only to be taxed upon clear words . . . To the question ‘What are clear words?’ he gave the answer that the court is not confined to a literal interpretation. He added ‘There may, indeed should, be considered the context and scheme of the relevant Act as a whole, and its purpose may, indeed should, be regarded.’ This sentence was critical. It marked the rejection by the House of pure literalism in the interpretation of tax statutes.”

The Lord President ultimately decided the issue by applying the ‘clear words’ principle in the light of his view of the statutory purpose (paras 55 and 56).

Lord Emslie relied on the same principle, and some other principles which he set out at para 197:

“Since this appeal concerns the construction of tax legislation, certain fundamental rules, principles and presumptions may be thought to apply.

First, as Lord Wilberforce explained in Vestey v Inland Revenue Commissioners [1980] AC 1148, 1172:

‘Taxes are imposed upon subjects by Parliament. A citizen cannot be taxed unless he is designated in clear terms by a taxing Act as a taxpayer and the amount of his liability is clearly defined.’

Second, in the absence of specific charging provisions, capital and capital receipts do not fall to be taxed as revenue and vice versa.

Third, corporation tax being an annual tax on the profits of a company, it is prima facie reasonable and appropriate to construe statutory charging provisions as directed towards real receipts and gains ‘ . . . in a sense which no commercial man would misunderstand’: Gresham Life Assurance Society v Styles [1892] AC 309, 315, per Lord Halsbury LC. And fourth, as reflected in countless provisions of the taxing statutes, a subject is in general assessable to tax on his own profits and gains, and not on those of any third party.’”

The second, third and fourth of these principles (and especially the second) may be what the Special Commissioners had in mind (in para 79 of their decision) in a more general reference to ‘tax principles’ as predisposing them in the Company’s favour, and in characterising the transfer from the Capital Reserve as a capital receipt (para 80).

Arguments based on the legislative scheme and purpose move from the very general to the rather more particular. What was the underlying purpose of section 83? In particular was it intended, as the Lord President stated in para 55 of his judgment, to reverse the effect of section 433 of ICTA 1988? Is there a ‘key conceptual distinction’ (Lord Emslie, para 201) between the Company’s LTBF and the assets representing that fund? Was a transfer from the Capital Reserve a capital receipt comparable to an injection of new capital (Special Commissioners, para 80)? How cogent is the argument (Lord Reed, para 183; Lord Emslie, para 197, fourth point, and para 205, second point) that one taxpayer should not be taxed on another taxpayer’s profits or gains? What practical results do the statutory

provisions produce if construed (Lord Emslie, para 200) as a one-stage or alternatively a two-stage process? I shall consider these points in turn.

Legislative scheme and purpose

It is permissible, without getting into the territory of Pepper v Hart [1993] AC 593, to look at the official consultation paper published in 1988, The Taxation of Life Assurance, to see the general nature of the problem perceived by the Revenue. The most relevant paragraphs are paras 6.2 to 6.7, 6.12 to 6.21, 6.33 and

7.1 to 7.8 A life office might have a large capital gain on a long-term income- producing investment (such as a fully-let office block or a strategic holding of shares in an oil company) as part of the with-profits part of its LTBF. Before 1989 this gain could be recognised (or brought into account) in its revenue account without being realised so as to give rise to a chargeable gain. Value representing at least nine-tenths of the gain could then be distributed (in the form 58 sense, that is allocated) to the holders of with-profits policies so as to obtain the protection of section 433 of ICTA 1988, as well as escaping income tax or capital gains tax in the policyholder’s hands on the maturity of their policies (assuming them to be qualifying policies). Section 83 of FA 1989 made the recognition of an unrealised capital gain a receipt to be brought into the Schedule D Case I computation, while section 82 of FA 1989 re-enacted the substance of section 433 in a more satisfactory form.

All this is very clearly set out, in a good deal more detail, in paras 123 to 133 of Lord Reed’s judgment (which refer to section 83 in the form in which it was originally enacted). I respectfully think that in para 55 of his judgment the Lord President was to some extent running together the functions of sections 82 and 83, and misunderstanding the purpose of the two sections in tandem. Lord Emslie referred to section 433 (para 200) but not to section 82. In my opinion Lord Reed’s analysis is to be preferred. Section 83 is concerned with the immediate implications, in making the necessary Case I computations, of bringing into account all or part of the difference between book value and market value, and section 82 is concerned with the next stage of the computations, that is adjustments in respect of the distribution of surplus to holders of with-profits policies (covered by form 58, lines 41 to 59).

The next point is the term ‘fund’. It is, as both Lord Reed (para 112) and Lord Emslie (para 199) observed, used inconsistently both in ICA 1982 and in the regulatory forms. But the two principal and relevant meanings, in this context, are clear (and here I am repeating ground I have already covered). The LTBF is an actual, appropriated fund of identifiable investments, the constituent assets of which (with their admissible values) appear in form 13. The Capital Reserve is a notional part of that fund to an initial amount of £4,455m; the independent

actuarial experts agreed that it serves no regulatory purpose. The fund for the purposes of lines 39, 49 and 59 of form 40, and for all the purposes of form 58, is the same fund, but valued in a special way (that is at book values in the sense that actuaries use that term) in order to produce the life office’s objectives – solvency and prudent preservation of the investment reserve, but at the same time smooth progress in the allocation of bonuses to with-profits policies.

I am not sure that I understand para 201 of Lord Emslie’s judgment. In that paragraph he is (as I understand it) setting out part of the submissions made on behalf of the Company. But later (para 204) Lord Emslie himself accepted that there is a significant distinction between the assets and the fund itself. Of course there is a difference, the difference between the parts and the whole. But the value of the whole is in this case the sum of the values of the parts, and the significant distinction, affecting both, is the basis of valuation.

It is common ground that if in 2001 or 2002 the Lloyds TSB Group had decided to inject fresh capital into the Company’s LTBF (as might have been done by the Company issuing new shares to Holdings, paid for in cash that was appropriated to the LTBF) it would not have been treated as a receipt under section 83(2). The new money would have appeared on line 26 of form 40 (transfer from non-technical account). The admissible value of the LTBF would have been increased, and so (if it was needed for solvency purposes) would its value as brought into account (lines 39 and 59 of form 40). A transfer from the Capital Reserve, by contrast, costs the group nothing (although it may be an indication that the state of the business is disappointing). The transfer does not increase the market value of the LTBF. Nor has it any regulatory significance, as the experts agreed. What happens is that part of the value held in the investment reserve is brought into account, a familiar event generally recorded (as Mr Allen stated, though Mr Chamberlain disagreed) on line 13 of form 40.

I respectfully consider that the Special Commissioners, and to some extent the Court of Session also, attached too much weight to the label ‘Capital Reserve’ and to the notion that capital gains ought not to be taxed under Schedule D, Case I. It could not be clearer that under section 83(2)(b) “any increase in value (whether realised or not)” of investments constituting a LTBF, as brought into account, is to come into the Case I computation.

The argument that (in the absence of very clear words) one taxpayer ought not to be taxed on another taxpayer’s profits or gains is, on the face of it, a strong one. It is not satisfactorily answered simply by pointing out (though this should not be forgotten) that this appeal is not about taxing profits. It is about allowing losses capable of being surrendered for the benefit of other group companies. But the Company acquired a long-established mutual business and a LTBF with a healthy

investment reserve. That reserve may have been built up by the Society largely by means of unrealised gains. But it was the Company and the Lloyds TSB Group that decided, for entirely understandable reasons, to bring part of the investment reserve into account, rather than making an injection of new capital. The language of section 83(3)(b) (as amended in 1996) shows that Parliament had demutualisation well in mind as a situation for which the legislation should make provision.

The last general point to be considered, before getting to linguistic arguments, is the implication of Lord Emslie’s illuminating distinction (para 200, summarising the Company’s argument, and para 204, accepting it) between a two- stage process (asking whether there are any real gains, and then how far they have been brought into account) and a one-stage process (asking simply what increase in value, if any, has been brought into account). Again, it is necessary to be reminded that this appeal is about losses, not gains; and the three accounting periods have to be considered separately, and not as a whole.

In any accounting period the operation of the statutory provisions, if analysed as a two-stage process, allows six different combinations, although some of them may be fairly improbable in practice, as follows (AV denoting admissible value, and RV value recognised and brought into account):

AV up, RV up by less

AV up, RV up by more

AV up, RV down

AV down, RV down by less

AV down, RV down by more

AV down, RV up.

It is easy to see how the competing interpretations work in situations (1), (2), (4) and (5). On the Company’s two-stage approach the lower figure (whether an increase or a reduction) will be brought into the computations; on the Revenue’s one-stage approach the difference in RV will always be taken. But it is

not so easy to see how either side’s interpretation would apply to situations (3) and (6); and the Company’s regulatory return for 2000 disclosed situation (6).

It might be thought that though neither side’s interpretation fits easily, the Company’s two-stage approach is distinctly more difficult to reconcile with the situation in which there is a reduction in admissible value, but an increase in value brought into account, in an accounting period, and the Company is seeking to establish an allowable loss during that period. But so far as I can see that submission was not made either to the Special Commissioners or to the Court of Session, nor do I recollect it being put forward in this Court. The terms of the agreed question do not positively require the point to be resolved. Indeed SFI, para 63 suggests that the point may already have been agreed between the parties. So the best course is, I think, to exclude that point, which was not argued, from any consideration of the statutory scheme and purpose. Nevertheless, unlike the Special Commissioners and (to some extent) the Court of Session, I do not approach the narrower linguistic points with any predisposition in favour of the Company’s case. I approach them disposed towards the Revenue’s case as being more in accordance with the statutory scheme and purpose.

Linguistic points on the first issue

I can take these more shortly, and it is convenient to do so by reference to the numbered sub-paragraphs at the end of para 181 of Lord Reed’s judgment.

The first point is that “an increase in value . . . of . . . assets” is said to refer most naturally to capital gains. In some contexts it might do so. In the context of a system of computation which is closely and explicitly linked to the regulatory returns in respect of LTBFs I see little force in this point. What is important is how value is to be measured, and to my mind sections 83(2) and 83A leave no doubt about that.

The second point is on the words “(whether realised or not)” in section 83(2)(b). The section was making an important change in the law in that unrealised increases in value, so far as brought into account, were to come into the tax computation. To my mind it would have been surprising if the draftsman had not inserted this parenthesis so as to leave no doubt as to the character of the change in the law.

The third point is on another parenthesis in section 83(2), “(and not otherwise)”, though these words have come out in the text of Lord Reed’s judgment before us as “or otherwise”. Again, I have to say that I think the

draftsman is being rather unfairly criticised for his efforts to leave no doubt about the intended meaning. The preceding word “as” means “in the manner that” and the parenthesis means “and in no other manner”. To my mind it is a bit hard to dismiss this as otiose.

Lord Reed’s fourth point is that the expression ‘brought into account’ is not apt to describe ‘the overall effect of those entries’. I confess that I simply do not understand this point. The critical entry is line 13 on form 40 (‘increase (decrease) in the value of non-linked assets brought into account)’. That is the only line on form 40 in which the words ‘brought into account’ are found. It was conceded that the line 15 entry could have been on line 13. The bottom lines (39, 49 and 59) show the overall position, and do not use the words ‘brought into account’.

Lord Reed’s remaining points on the first issue (in para 181(5) and (6) and para 183) are more general and I will not revisit them.

Conclusion

In my judgment the Revenue’s submissions on the first issue are correct, both as to the statutory scheme and purpose and as to the linguistic points just mentioned. I have gone into the matter at some length because I am conscious that I am differing both from the Special Commissioners and from the unanimous view of the Court of Session. But in the end I consider that it is simply a question of giving section 83(1) and (2) of FA 1989, as amended, their natural meaning. On that basis the second issue does not arise and I prefer to say nothing about it. I would allow the Revenue’s cross-appeal and treat the Company’s appeal as moot.

LADY HALE

As so often happens, what appears at first sight to be a very complicated question turns out on closer analysis to be quite a simple one. When calculating the profits of an insurance company in respect of its life assurance business under Case 1 of Schedule D to the Taxes Act, does an “increase in value” – or conversely a “reduction in the value” – of the assets of its long term business fund refer to an increase or decrease in their actual value? Or does it refer to an increase or decrease in their value “as brought into account for a period of account” in the company’s revenue account prepared for the purpose of the Insurance Companies Act 1982?
We know that the words “as brought into account for a period of account (and not otherwise)” in section 83(2) of the Finance Act 1989 (set out by Lord Walker at para 41 above) describe the words “the following items”; we know that the following items are “(a) the company’s investment income from the assets of its long term business fund, and (b) any increase in value (whether realised or not) of those assets”; we know from section 83A (set out by Lord Walker at para 74 above) that “brought into account” means brought into account in the revenue account (or accounts) prepared for the purposes of the Insurance Companies Act 1982 in respect of the company’s long term business (or part of it); so the linguistic question boils down to what is meant by “as” in section 83(2).

The Company would have it that “as” means “when”. The link to the regulatory returns is a purely temporal one. Value means real value not whatever the company chose to put in the forms. The Revenue would have it that “as” means “as”. What is taken into account in computing the company’s profits for income tax purposes is what the company brings into account in completing its revenue accounts for regulatory purposes. In my experience, if Parliamentary counsel mean “when”, they write “when”, and if they mean “as”, they write “as”. We should be slow to re-write what they have written.

The words “and not otherwise”, if nothing else, make it clear that there might have been some other way of taking items (a) and (b) into account for income tax purposes, but this is the way it is to be done. They are making a special rule for life insurance business. This is not surprising, for all the reasons that Lord Walker has so clearly and carefully explained. The words “whether realised or not” point to the real change which was being made by the 1989 Act. Otherwise it was business as usual. It was not until 1995 that these insurance companies were required to file any other sort of accounts than those which they had to file for regulatory purposes. It was natural for the Revenue to use the figures in the regulatory revenue account as their starting point.

In full agreement with Lord Walker, and Lord Hope and Lord Neuberger, therefore, I would allow the Revenue’s cross-appeal and regard the Company’s appeal as moot.

LORD NEUBERGER

I too would recall the interlocutor of the Inner House of the Court of Session and allow HMRC’s cross-appeal. Having had the great benefit of reading in draft the judgments of Lord Hope and Lord Walker, I can express my reasons shortly.
The cross-appeal raises an issue as to the meaning of section 83(2) of the Finance Act 1989 (as substituted by paragraph 16 of Schedule 8 to the Finance Act 1995). It is unnecessary for me to set out section 83, as it is fully quoted in para 7 of Lord Hope’s judgment and para 73 of Lord Walker’s judgment.

As will be clear to anyone who has read those judgments, the difficulty in this case arises from the fact that the issue of interpretation arises in the context of a very complex background. That complexity is attributable to a number of different factors, namely (i) the technical rules as regards the regulatory returns to be made by life assurance offices, (ii) the many changes in the legislation embodying those rules since they were first introduced in 1870, (iii) the many changes in the statutory provisions governing the taxation of life assurance offices,

(iv) the extensive contractual provisions in the documentation governing the scheme (“the scheme”) for transferring of the business of Scottish Widows’ Fund and Life Assurance Society to Scottish Widows plc (“the Company”), and (v) the details of the regulatory returns made by the Company in the three accounting years in issue.

When considering the application of section 83(2) to the facts of this case, I am sceptical about the value of analysing the history of the statutory provisions governing either the returns to be made by life assurance offices or the taxation of profits made by life assurance offices – i.e. what I have characterised as factors (ii) and (iii). This cross-appeal concerns the meaning of the statutory provision in force during the three relevant years, section 83(2), and its impact on the returns actually made in respect of those years, in the then stipulated form by the Company. Particularly as the provisions of the scheme, the prescribed forms for returns, and the contents of the Company’s returns for the three years in question, all require careful analysis, it seems to me that to focus in addition on the rather intricate history, as opposed to the present provisions, of the regulations, risks taking one’s eye off the ball (or, as Lord Walker puts it, not seeing the wood for the trees).

Legislative archaeology has its place in statutory interpretation, but its role is limited. Where a statutory provision, when read in its immediate statutory and practical context, has a meaning which is tolerably clear as a matter of language, and not unreasonable or unfair in terms of its consequences, it seems to me that little is to be gained, and much may be lost (in terms of time, expense and eventual confusion) by going into the genesis and development of the provision in earlier legislation. As Lord Hope points out, such an approach is consistent with what was said both by Lord Wilberforce in Farrell v Alexander [1977] AC 59, 73 and by Lord Diplock in Inland Revenue Commissioners v Joiner [1975] 1 WLR 1701, 1715-6.
Once one understands the scheme, the relevant regulations, the forms, and the Company’s returns for the three years in question, it appears to me that the answer to the question posed on the cross-appeal, namely the meaning and effect of section 83(2), is tolerably clear. I could not hope to equal the clarity of Lord Walker’s analysis in paras 49 to 53, 57 to 70, and 82 to 91, and, very gratefully, adopt the benefit of his distillation of the various complex matters which he there explains.

Turning to the central issue on the cross-appeal, the meaning of section 83(2), it may be a little glib to suggest that HMRC’s case is ultimately vindicated by a single word. However, if one was to isolate a single crucial point, it seems to me that it would involve focussing on the word “as” in that subsection, as Lady Hale suggested during argument. Section 83(2) stipulates that the “items” which should be “taken into account as receipts of [a particular accounting] period” are to be those items identified in paras (a) and (b) “as brought into account for [that] period of account”. The obvious and natural effect of the words which I have emphasised is that those items are to be “taken into account” for the period in question in the same way and to the same extent as they are “brought into account” for that period.

It was argued on behalf of the Company that the expression “as brought into account for a period of account” should be treated simply as a reference to the period in which the item was brought into account, and was not concerned with how the item was brought into account. Particularly in the light of the inclusion of the word “as”, that does not seem to me to accord with the natural reading of the expression.

Although both the Company and HMRC relied on other provisions in the 1989 Act to support their respective cases on the meaning of the expression, I am unconvinced that they are of any real assistance. Thus, it was suggested that the words “(and not otherwise)” in the subsection assisted the Company’s interpretation. I do not see how that is so: they are there simply to emphasise that an item is only to be brought into account if it falls within the expression, and therefore they can take the issue of what the expression means no further. Equally, the fact that section 83A(1) (as inserted by paragraph 16 Schedule 8 to, the 1995 Act and amended by paragraph 6 of Schedule 31 to the 1996 Act) defines the term “brought into account” does not take matters further, as the position of each party is consistent with that definition.

The conclusion that HMRC’s construction of section 83(2) is correct seems to me to be supported, rather than undermined, by the normal approach to taxation of business profits as explained by Lord Halsbury LC in Gresham Life Assurance Society v Styles [1892] AC 309, 315 and by Lord Hoffmann in Revenue and

Customs Commissioners v William Grant & Sons Distillers Ltd [2007] UKHL 15, [2007] SC (HL) 105, [2007] 1 WLR 1448, para 2. In connection with taxing business profits, the concept of a profit should normally be accorded its proper meaning, which will obviously depend on the specific context, but current accountancy practice is generally a good, and often the best, guide as to the precise quantification of any profit.

In the case of a life assurance business, HMRC’s case is that, in effect, statute requires the profit to be assessed by a rather unusual means, namely by reference to the returns in the regulatory forms. This is entirely consistent with the normal approach to assessing the profits of companies for taxation purposes, as these returns effectively take the place of the statutory audited accounts, which are relied on to define the profits of the overwhelming majority of other businesses owned by limited companies. Furthermore, although there is, at first sight, real force in the Company’s argument that HMRC’s case results in its business being taxed on some figure which cannot sensibly be said to be a profit, a closer analysis of the situation, as provided by Lord Walker in paras 81 to 90 and 95 to 105, shows that this is incorrect.

Accordingly, and in agreement with the fuller reasoning of Lord Hope and Lord Walker, I would allow HMRC’s cross-appeal.

As the cross-appeal succeeds, it is unnecessary to consider the Company’s appeal. I agree with Lord Walker that it would be better not to go into the question whether the majority of the Inner House was right in finding for HMRC on section 83(3). It is tempting to do so, given that there is a decision of the Inner House on the point. However, at least on the basis of the argument we have heard on this appeal, it does seem to me that the interpretation of subsection (3) is rather difficult, and I think it would be better to wait for a case where the issue matters, not least as it may be that some assistance would be gleaned from the facts of such a case, which may throw light on the practical consequences of the rival interpretations.

LORD CLARKE

I confess that I was initially attracted by the approach of Lord Emslie to the issues in relation to both the cross-appeal and the appeal. However, having considered the masterly judgment of Lord Walker, I have found his reasoning compelling and agree with him (and indeed Lord Hope, Lady Hale and Lord Neuberger) that the Revenue’s cross appeal on the true construction of section 83(2) of the Finance Act 1989 as amended should be allowed. Like Lord Walker,

Lady Hale and Lord Neuberger, I prefer to express no view on the issue of the true construction of section 83(3). I too would therefore allow the Revenue’s cross- appeal and treat the Company’s appeal as moot.

Scottish Widows plc No 2 v Commissioners for Her Majesty’s Revenue and Customs (Scotland)

Scottish Widows plc v Commissioners for Her Majesty’s Revenue and Customs (Scotland). Scottish Widows plc v Commissioners for Her Majesty’s Revenue and Customs (Scotland)

LORD WALKER

LADY HALE

LORD NEUBERGER

LORD CLARKE

Related Posts:

More Posts

Section 47 EFCC Act 2004: Short Title

Section 46 EFCC Act 2004: Interpretation

Section 45 EFCC Act 2004: Savings

Section 44 EFCC Act 2004: Repeal of Act No. 5 2002

Leave a Reply Cancel reply