JD Ltd v Comptroller of Income Tax
[2005] SGCA 52

Case Number	:	CA 21/2005
Decision Date	:	02 December 2005
Tribunal/Court	:	Court of Appeal
Coram	:	Chao Hick Tin JA; Tan Lee Meng J; Yong Pung How CJ
Counsel Name(s)	:	Leon Kwong Wing and Chee Fang Theng (KhattarWong) for the appellant; Liu Hern Kuan and David Lim (Inland Revenue Authority of Singapore) for the respondent
Parties	:	JD Ltd — Comptroller of Income Tax

2 December 2005

Judgment reserved.

Yong Pung How CJ (delivering the judgment of the court):

1          This is an appeal by the taxpayer company against the judgment of the High Court in [2005] SGHC 92 dismissing its appeal against the decision of the Income Tax Board of Review (“the Board”) in [2004] SGDC 245. The Board had affirmed an assessment of the Comptroller of Income Tax (“the Comptroller”), who had disallowed the deduction of interest expenses incurred in relation to non-income-producing share investments. We agree with the decision of the judge below and accordingly dismiss the appeal. We now set out our reasons.

Background

2          The appeal, as was the case before the Board and the court below, is based on an Agreed Statement of Facts. The facts have been duly amended pursuant to s 83(3) of the Income Tax Act (Cap 134, 2004 Rev Ed) (“the Act”) so as not to disclose the identity of the taxpayer.

3          The taxpayer is an investment holding company listed on the Singapore Stock Exchange. It focuses on making long-term share investments. During the years of assessment 1985 to 1996 (“the years of assessment in dispute”), the taxpayer received, as its only income, dividends from long-term share investments held in various subsidiary and associated companies.

4          The taxpayer financed the purchase of the share investments by either (a) obtaining overdrafts and loans from banks and related companies at varying interest rates, or (b) issuing its own shares or obtaining interest-free loans from related companies. These funds were mixed and banked into the same account. The funds were utilised not only for the acquisition of the taxpayer’s share investments but also for its re-financing of earlier loans and the making of advances to related companies.

5          Some of the companies in which the taxpayer had shareholdings did not declare dividends for the entire duration of the years of assessment in dispute. Various other companies also did not declare dividend income for certain of those years of assessment in dispute. The taxpayer was assessed on the basis that only interest expenses attributable to shareholdings which produced income were deductible. Interest expenses attributable to non-income-producing shareholdings were not allowed for deduction.

6          The taxpayer, objecting to the Comptroller’s method of assessment, appealed to the Board on two main issues. This court, however, as was the court below, is only concerned with the issue relating to the disallowance of interest expenses, that is: If interest expenses are incurred to maintain a portfolio of share investments in which some share investment counters do not yield dividend income, is the entire sum of interest expenses deductible, or are only those interest expenses attributable to the income-producing share investment counters deductible?

7          Both the taxpayer and the Comptroller agreed to the application of the Total Assets Formula (“the Formula”) in determining the amount of interest expenses applicable. Where they differed, however, was in the application of the Formula to the facts. This is indicated in the following table:

                                               Total Assets Formula

A     x    I   C	A (Taxpayer’s basis)	Cost of all share investments financed by interest-bearing funds for each year of assessment in dispute.
	A (Comptroller’s basis )	Cost of all income-producing share investments financed by interest-bearing funds for each year of assessment in dispute.
	C	Total cost of assets as at balance sheet date which were financed by interest-bearing funds for each year of assessment in dispute.
	I	Total interest expenses incurred by taxpayer for each year of assessment in dispute.

8          The taxpayer was of the view that the interest expenses incurred in the obtaining and servicing of loans to acquire its whole basket of share investments was the cost of earning the dividend income received. It contended that all the share investment counters formed a single source of income so that the dividend income received should be assessed as a whole. The taxpayer consequently sought to deduct all interest expenses incurred in respect of the borrowings against the entire sum of dividend income earned from its investments portfolio.

9          The Comptroller, on the other hand, regarded each share investment counter as a separate source of income and allowed deductions of interest expenses only for those shareholdings which produced dividend income at any time during the years of assessment in dispute. The Comptroller thus aggregated the interest expenses incurred only in respect of the income-producing share investment counters to determine the total amount of deductible expenses for the particular year of assessment. This figure was then used in determining the chargeable income for that year of assessment.

10        Due to the difference in approach, the taxpayer’s aggregate chargeable income for the years of assessment in dispute as computed by the Comptroller was $83,484,337 whereas the figure as computed by the taxpayer was $74,694,762. The amount of tax in dispute was $2,497,841.74.

11        The difference in positions taken by the taxpayer and the Comptroller stems from their disagreement as to the interpretation of the meaning of “source” (of dividend income) and of the deductibility provision under the Singapore scheme of taxation. The relevant provisions for present purposes are ss 10(1)(d), 14(1) and 14(1)(a) of the Act. Unless otherwise stated, we have used the 2004 revised edition of the Income Tax Act for ease of reference, as the text of the sections under review is the same as that of the previous editions in use during the years of assessment in dispute. The sections are set out below:

Charge of income tax

10.—(1) Income tax shall, subject to the provisions of this Act, be payable at the rate or rates specified hereinafter for each year of assessment upon the income of any person accruing in or derived from Singapore or received in Singapore from outside Singapore in respect of —

…

(d)        dividends, interest or discounts;

…

Deductions allowed

14.—(1) For the purpose of ascertaining the income of any person for any period from any source chargeable with tax under this Act (referred to in this Part as the income), there shall be deducted all outgoings and expenses wholly and exclusively incurred during that period by that person in the production of the income, including —

(a)        except as provided in this section, any sum payable by way of interest upon any money borrowed by that person where the Comptroller is satisfied that the interest was payable on capital employed in acquiring the income;

…

[emphasis added]

12        The Board interpreted “source” in s 14(1) according to its natural or ordinary meaning of “channel or stream of income”. Therefore, the Board regarded dividends from each of the different share investment counters as constituting a separate and distinct source of income for the purposes of deduction. This was consistent with the principle that s 14(1)(a) required a direct nexus between the expenses incurred and the income produced. The Board, having also found the Formula to be legally tenable and to have been reasonably applied on the facts, upheld the Comptroller’s position to disallow the deduction of interest expenses attributable to non-income-producing share investment counters. The judge below affirmed the Board’s reasoning and decision.

13        In the present appeal, counsel for the taxpayer and the Comptroller made various submissions on whether ss 10(1)(d), 14(1) and 14(1)(a) entitled the Comptroller to treat the share investment counters as individual sources of income for the purposes of deduction. In addition, the taxpayer contended that the subsequent inclusion of s 10E in the Act, which expressly permits the Comptroller to differentiate between income-producing and non-income-producing share investment counters for the purposes of deduction of expenses incurred, must have meant that before its inclusion, such differentiation was not allowed under the Act. The taxpayer also argued that the situations where expenses were not deductible were confined only to cases where the expenses were incurred in the gaining of tax-exempt income or where the investment itself had been extinguished.

The issues

14        The issues we are concerned with in this appeal are therefore:

(a)        whether ss 10(1)(d), 14(1) and 14(1)(a) of the Act entitle the Comptroller to treat the share investment counters as individual sources of income for deduction purposes. In particular:

(i)         what is the meaning of “source” in s 14(1)?

(ii)        what is the correct statutory interpretation of “in the production of the income” in s 14(1)?

(iii)       whether the Comptroller exercised his administrative discretion properly;

(b)        whether s 10E of the Act is intended to be a declaration of the law as it has been or a change in the law; and

(c)        whether the situations where expenses are not deductible are confined only to those where expenses are incurred in the obtaining of tax-exempt income or where the investment itself has been extinguished.

Whether the Comptroller is entitled by the Act to treat the share investment counters as individual sources of income for deduction purposes

What is the meaning of “source” in s 14(1)?

15        The judge below, noting that there is no technical definition of “source” in the Act, upheld the Board’s interpretation of the word “source” in its natural or ordinary meaning.

16        In unravelling the meaning of the word “source”, we bore in mind a previous statement by this court in Comptroller of Income Tax v GE Pacific Pte Ltd [1994] 2 SLR 690 at 697, [23], that “[i]n any question of statutory interpretation, the first and most important factor is the literal meaning of the words of the provision”. “Source” has been defined by the Oxford Modern English Dictionary (Oxford University Press, 1992), which was referred to by the Board at [13] of its decision, as “a spring [or] fountain from which a stream arises” or “a place or thing from which something originates”.  The Board thus probably likened each share investment counter to being such a spring from which dividend income arises, with each shareholding forming a separate and distinct source of income.

17        Of course, any discussion of the meaning of “source” of income would be incomplete without an examination of the charging section, s 10 of the Act, since “source” in s 14(1) is followed by the phrase “chargeable with tax under this Act”. Section 10 lists the various heads of income chargeable with tax, one of which is dividend income. From a literal reading of s 10, however, there seems to be no unequivocal indication of whether “dividend income” could refer to many sources of income arising from different shareholdings, or whether it could only refer collectively to a single source of income arising from the various shareholdings.

18        We note that in the area of statutory interpretation, we are also guided by s 9A of the Interpretation Act (Cap 1, 2002 Rev Ed), which requires the interpretation of written law such as will promote the purpose or object underlying the law. Section 9A therefore allows the courts, in appropriate cases, to have recourse to additional materials outside the four corners of a statute, including subsidiary legislation, to ascertain the meaning of a statutory provision: Raffles City Pte Ltd v AG [1993] 3 SLR 580 at 587, [19] per L P Thean J (as he then was). Reference can be made to extrinsic materials such as parliamentary reports to ascertain the legislative intent behind a certain enactment that is ambiguous or obscure or the literal meaning of which would lead to an absurdity. Such a purposive approach to statutory interpretation can be resorted to even where the provision in question is not ambiguous or inconsistent: The Seaway [2005] 1 SLR 435 following Constitutional Reference No 1 of 1995 [1995] 2 SLR 201, L & W Holdings Pte Ltd v MCST Plan No 1601 [1997] 3 SLR 905 and Planmarine AG v Maritime and Port Authority of Singapore [1999] 2 SLR 1.

19        It was perhaps in this light that counsel for the taxpayer referred us to the legislative background of the enactment of our Income Tax Act. The charging section (s 10) and the deductibility provision (s 14) were some of the core tax provisions which were included when income tax legislation was first enacted as law in Singapore. Counsel for the taxpayer first brought our attention to the Appendix to the Annual Report on the Federation of Malaya 1948 (H T Ross, 1949), at p 179, where it was stated that “[o]wing to the close economic relations existing between the Federation and the Colony of Singapore it was considered essential that the [income] tax should be operated on a pan-Malayan basis”. He then referred us to two draft Bills, the Federation of Malaya Income Tax Bill of 1947 (No 7510) and the Singapore Income Tax Bill of 1947 (GN No S 361/1947), both of which declared that cl 10, the precursor to s 10(1), “specifies the various sources, income from which is liable to tax” and that “[t]he sources of income specified in Clause 10 are exhaustive”. To the taxpayer, this was authority that dividend income, even if arising from a variety of share investment counters, collectively forms a single source of income under s 10(1)(d) of the Act, against which interest expenses are deductible regardless of whether the share counters are income-producing.

20        It is true that our tax legislation shares a common historical background with its Malaysian counterpart. After the passing of the Bills in 1947, the Income Tax Ordinance 1947 (Federation of Malaya) (Ord No 48 of 1947) and the Income Tax Ordinance (Singapore) (Ord No 39 of 1947) came into effect on 1 January 1948. For many years, the legislation in both countries was identical and income tax operated on a pan-Malayan basis. When Singapore merged with the Federation of Malaya to form the newly-established Federation of Malaysia in 1963, income tax continued to be levied and administered in Singapore as part of the Malaysian income tax laws. However, this stopped in 1965 when Singapore separated from the Federation to become a fully independent sovereign state. Income tax then became administered under the Income Tax Act (Cap 141, 1970 Rev Ed) as the law of Singapore as a republic. As befits the independent status of each country, although Singapore and Malaysia have maintained the basic structure of their respective 1947 Ordinances, both countries have since diverged in many aspects from the common origin of their tax legislation.

21        In the light of the fact that the two countries have over the years developed their own schemes and laws of taxation separately to fit the unique needs of their own economies, our shared history with the Malaysian tax legislation now bears less relevance. It would not at all be surprising, therefore, if courts in the two jurisdictions take different approaches in the interpretation of provisions in the tax legislation, even if these provisions had originated from the same mould. Consequently, we find that reference to the Income Tax Bills of 1947 for Malaysia and Singapore is an exercise of limited relevance and utility in the present appeal.

22        In any event, we are of the opinion that the statements that cl 10 “specifies the various sources, income from which is liable to tax” and that “[t]he sources of income specified in Clause 10 are exhaustive” are not reflective of whether these sources themselves may or may not be subdivided. On the contrary, we think that when s 10(1) is read in conjunction with s 14(1), the latter provision supports the proposition that the s 10(1) sources may be subdivided for assessment purposes when taking into account the deducibility of expenses. This will be elaborated upon later.

23        It is appropriate though to address, at this point, the taxpayer’s argument that this court had in fact agreed with its interpretation of “source” in light of its earlier statement in Andermatt Investments Pte Ltd v Comptroller of Income Tax [1995] 3 SLR 451 (“Andermatt”) at 457, [16] that “[s]ection 10(1) sets out the various sources of income which are subject to tax”. Such an argument is wholly unmeritorious. In our view, reliance on judicial expressions without regard to the unique factual matrix of the case is potentially misleading. In Andermatt, the dividends had arisen from only one homogenous block of shares of one company alone. There could therefore only have been one source of dividend income on the facts of that case. No discussion of whether various shareholdings could form separate and distinct sources of dividend income could have presented itself. The statement by this court should not be understood to mean that dividend income, no matter how it is derived, invariably forms a single source of income.

24        We also agree with the view of the judge below that the heads of income enumerated under s 10(1) of the Act are simply descriptions of the types of income that s 10(1) enacts as being chargeable to tax, one of which is dividend income. The key is to look at the taxpayer’s activities and/or property as the originating cause of income or the means by which the different types of gains or rewards enumerated and classified as “income” are derived.

25        Further, as noted by the Board, there are provisions in the Act which indicate that the heads of chargeable income listed under s 10(1) can comprise two or more sources of income. Section 35A(1), for instance, suggests that there may be more than two sources of income from employment. The provision states that:

This section shall only apply to any trade, business, profession, vocation or employment (except subsidiary employment which had not been treated as a new source on commencement) which commenced before 1st January 1969.

Section 35A(1) presupposes that a taxpayer may have a subsidiary employment which can be treated as a separate source of income from his principal employment. The head of “employment” income under s 10(1) can therefore encompass more than one source of income. Similarly, when this reasoning is applied for our purposes, “dividends” may thus comprise separate sources of income, instead of a single source of income.

26        In our view, therefore, for the foregoing reasons, the judge was correct to hold that the source of the dividend income must necessarily stem or originate from the particular share counter that yielded the revenue. We wholly agree with and adopt the pronouncement of the High Court of Allahabad in Seth Shiv Prasad v Commissioner of Income-Tax, UP [1972] 84 ITR 15 (“Seth Shiv Prasad”) where Pathak J said at 18:

[W]e find that income from dividends has been treated as one kind of income. What is the source of dividend income? It is the shareholding held by the assessee; and there can be as many sources of income as there are shareholdings. Shareholding[s] in different companies constitute different sources of income. The shares of one company may be treated by the assessee as a single shareholding. The assessee may also, for good reason, treat the shares of the same company as constituting a number of separate and distinct shareholdings. The shares may be divided into groups defined by reference to the circumstances in which they were acquired, or to the purpose for which they were purchased, that is, some as an investment holding and others as a share-dealer’s stock-in-trade, or to the category or class to which they belong, for example, whether they are preference or equity. There may be other criteria reasonably defining them into separate and distinct shareholdings, and, therefore, as distinct and separate sources of income. [emphasis added]

27        We shall now turn to explain why, in our view, the taxpayer’s interpretation of “source” does not comport with the deduction rules in s 14(1).

What is the correct statutory interpretation of “in the production of the income” in s 14(1) of the Act?

28        The issue that concerns us here is whether income must have been produced before its related expenses are allowed for deduction. Counsel for the taxpayer argued that “no revenue authority in the world shares the Comptroller’s opinion that capital has not been employed to acquire dividends unless the dividends are actually received in the course of the year under tax” because “capital is, for a fact, employed prospectively, that is, in the hope and expectation of income”. Counsel for the taxpayer therefore argued that interest expenses should be deducted regardless of the economic performance of the shares, since investment returns are never guaranteed owing to market fluctuations and the economy. In essence, the proposition advanced is that so long as capital is employed in the inherent expectation of income, there is no requirement that income must actually be produced in order for there to be deduction of expenses.

29        In determining whether the taxpayer’s argument is valid, it is necessary to turn to the wording of the general deduction formula in s 14(1). Section 14(1) provides that “there shall be deducted all outgoings and expenses wholly and exclusively incurred during that period by that person in the production of the income” [emphasis added]. In our opinion, regard must be had as to whether the phrase “in the production of the income” means that income must be produced before expenses are allowed for deduction. There is apparently a paucity of local case law on the interpretation of this phrase. Counsel for the taxpayer consequently introduced many authorities spanning the various jurisdictions in the Commonwealth in an attempt to substantiate his point.

The persuasiveness of foreign authorities

30        Before examining these positions, however, a word must be said about the reliance on foreign authorities.

31        It is understandable for tax practitioners to exhibit a proclivity towards relying on texts and case law from the Commonwealth jurisdictions such as Malaysia, Hong Kong, India, Australia and New Zealand. This is because our Income Tax Act is based on the Model Colonial Territories Income Tax Ordinance 1922 devised for British colonies. There is also a tendency to refer to English tax cases, owing to our British colonial heritage and the passing of the Application of English Law Act (Cap 7A, 1994 Rev Ed) which expressly makes English common law part of the law of Singapore so far as it is applicable to the circumstances of Singapore and its inhabitants and subject to such modifications as those circumstances may require.

32        However, we find it important to remind practitioners that as tax law is essentially a creature of statute, decisions from foreign jurisdictions should be treated with the appropriate degree of caution, especially where the wording of the foreign tax legislation is not identical with or not in pari materia with the local equivalent. In this respect, the pronouncement of the Privy Council in Liquidator, Rhodesia Metals, Limited v Commissioner of Taxes [1940] AC 774 is apposite. There, Lord Atkin said at 788:

Their Lordships have no criticisms to make of any of those decisions, but they desire to point out that decisions on the words of one statute are seldom of value in deciding on different words in another statute, and that different business operations may give rise to different taxing results.

It is desirable, therefore, in interpreting tax legislation, to rely on foreign authorities only if the corresponding tax statutes are identical or very similar to local legislation, and if the schemes of deduction and taxation systems are alike.

33        Our courts have consistently exhibited awareness of the need for such caution when referring to foreign tax cases. For instance, Lim Teong Qwee JC had, in the High Court in Andermatt Investments Pte Ltd v Comptroller of Income Tax [1995] 1 SLR 66, after referring to a Jamaican case which dealt with s 8(1) of the Income Tax Law 1954 (Jamaica), said at 74, [25]:

I think this case lends some support for the conclusion I have reached. However, it is not entirely desirable to do so as the question of deductibility in the case before me has to be considered by reference to the provisions of the Act which are not exactly in pari materia.

Another illustration of how our courts are careful in comparing tax provisions across jurisdictions is seen in T Ltd v Comptroller of Income Tax [2005] 4 SLR 285, where Andrew Ang J said at [55] to [57]:

55         So much then as to how the UK provision was construed in the European Investment Trust case [(1932) 18 TC 1]. Should the Singapore provision be interpreted likewise? Although counsel for the appellant argued cogently why it ought not be likewise construed, he may have been a little generous in conceding that the language of the Singapore provision is close to its UK counterpart. As noted, the UK provision prescribes that:

[N]o sum shall be deducted in respect of … any sum employed or intended to be employed as capital …

In other words, “no sum … in respect of any sum employed as capital …” is to be deducted.

56         As framed, it is possible to construe the “sum” first referred to as separate and distinct from the second. On this basis, it did no violence to the language of the UK provision for the European Investment Trust case ([54] supra) to construe the first sum as being referable to interest while the second referred to the principal amount on which the interest accrued.

57         The Singapore provision is differently worded. It states:

[N]o deduction shall be allowed in respect of … any sum employed or intended to be employed as capital.

The provision makes no mention of any sum other than the sum employed or intended to be employed as capital. Whereas the words “in respect of” in the UK provision could be read to mean “in connection with” without doing violence to the statutory provision, the same words “in connection with” could not, in my view, comfortably substitute for “in respect of” in the Singapore provision. It would immediately invite the question “Deduction of what?” To my mind, the words “in respect of” in the Singapore provision is the equivalent of “for” or “on account of”. Thus, what is prohibited is the deduction of the sum employed or intended to be employed as capital.

34        With these considerations in mind, we shall now turn to the authorities raised by counsel of the taxpayer to support its reading of s 14(1). It will be evident, from a comparison of the relevant deduction formulae across the jurisdictions raised, that none of them are identical to our s 14(1). The schemes of taxation in those jurisdictions are also dissimilar. As a result, these authorities should have correspondingly little bearing on our interpretation of our own deduction formula. Indeed, our courts should not blindly follow the interpretation adopted by other jurisdictions. As mentioned, each jurisdiction has different tax legislation and different tax and economic policies. The tax treatment cannot all be uniform.

35        Counsel for the taxpayer relied heavily on two Malaysian authorities which were very similar on the facts to the present appeal. In P Securities Sdn Bhd v Ketua Pengarah Jabatan Hasil dalam Negeri (1995) 2 MSTC 2,256, (“P Securities”) the taxpayer was an investment holding company which principal activity was investing in shares and which also financed the acquisition of the shares by interest-bearing loans. The Director-General of Inland Revenue took the view that each share investment was a separate source of income to the company, and apportioned the taxpayer’s interest costs so as to differentiate between shares which had paid dividends for that year of assessment and shares which had not paid dividends for that year of assessment. The Director-General therefore disallowed the deduction of those interest expenses attributable to the non-income-producing basket of shares. The company appealed on the basis that its only source of income was dividend income, and applied to deduct the whole amount of interest expenses against its dividend income as a whole, irrespective of whether the shares did or did not pay dividends in any particular year.

36        The Malaysian Special Commissioners of Income Tax found for the taxpayer on the basis that the dividend performance of each share counter was irrelevant because under the terms of the charging section, tax was payable on the “income in respect of dividends” as a whole. The Special Commissioners were of the view that it was wrong for the Director-General to further subdivide each source by treating each share investment counter as a separate source because to do so was to disintegrate the grouping further than that authorised by the Malaysian Income Tax Act 1967 (Act 53). This was upheld by the High Court of Kuala Lumpur, which did not issue grounds. The Director-General did not appeal against the decision of the High Court.

37        The next case is Ketua Pengarah Hasil Dalam Negeri v Multi-Purpose Holdings Bhd [2002] 1 MLJ 22 (“Multi-Purpose Holdings”), where the taxpayer was a public-listed investment-holding company. Again, the revenue authorities treated each share investment as a separate source of income to be assessed. This was overturned by the Special Commissioners of Income Tax and the High Court of Kuala Lumpur on the basis that there was no authority to “split this classification [of “dividends” in the Act] up in a manner that increase[d] the taxpayer’s liability to tax” (at 31).

38        Although the facts of P Securities and Multi-Purpose Holdings are similar to those of the present appeal, one ought to be scrupulous in deciding whether to place such heavy reliance on them because the Malaysian tax legislation on which they are based is now worded differently from the local equivalent. Indeed, the deduction formula in s 33(1) of the Malaysian Income Tax Act 1967 expressly applies to “outgoings and expenses wholly and exclusively incurred during that period by that person in the production of gross income from that source” [emphasis added]. There is not the same “the” appearing before “gross income”. Therefore, there is no specific reference to income that must be produced against which the related interest expense is deductible. This is a reflection of the fact that whilst the Malaysian tax system might once have been the same as ours, this is “no longer the case as our taxation statute had undergone substantial amendments”: see [34] of the High Court judgment below ([1] supra).

39        Counsel for the taxpayer also relied on the fact that the Indian courts in Ormerods (India) Private Ltd v Commissioner of Income-Tax, Bombay City [1959] 36 ITR 329, K Appa Rao v Commissioner of Income-Tax, Madras [1962] 46 ITR 511, P V Mohamed Ghouse v Commissioner of Income-Tax, Madras [1963] 49 ITR 127 and M N Ramaswamy Iyer v Commissioner of Income-Tax, Kerala [1969] 71 ITR 218 held that it was a well-accepted rule of income tax law that a deduction could not be disallowed by the mere happenstance that there was no income at the same time. However, one ought to note that s 12(2) of the Indian Income-Tax Act 1922 (the relevant provision in those cases) provides that “[s]uch income, profits and gains shall be computed after making allowance for any expenditure (not being in the nature of capital expenditure) incurred solely for the purpose of making or earning such income, profits or gains” [emphasis added]. From the wording of the Indian provision, under which “for the purpose” has been held to mean “with a view to” or “intending to”, it is indicative that expenses are allowable so long as they are produced in the expectation of income, and there is no need for the purpose (of earning income) to come to fruition. In the same vein as our analysis of the Malaysian authorities, the Indian provision is not in pari materia with its Singapore counterpart. Section 14(1) of the Act makes no reference to “purpose” at all.

40        The revenue authority in New Zealand has also adopted the position that income does not have to be derived in the same income year in which the deduction is claimed, as long as it relates to the derivation of gross income for any income year: Staples’ Guide to New Zealand Tax Practice (Brooker’s, 60th Ed, 2000) at para 830.50. It serves to note, however, that in New Zealand, the deduction formula is once again not in pari materia with its Singapore counterpart. Indeed, the relevant provision in New Zealand is s DA1(1) of the Income Tax Act 2004 (No 35) (a clarification of the previous formula in s  DD1(b) of the Income Tax Act 1994 (No 164)) which provides that deductions are allowed for expenses “incurred by [the taxpayers] in deriving … their assessable income”. This is different from our s 14(1).

41        As for Australia, reference must be made to the general deduction formula in s 51(1) of the Australian Income Tax Assessment Act 1936 (No 27 of 1936) (“ITAA 1936”), or its equivalent in s 8-1 of the Income Tax Assessment Act 1997 (No 38 of 1997) (“ITAA 1997”). (It ought to be noted that s 8-1 of the ITAA 1997 is not intended to amend the pre-1997 law, according to the Explanatory Memorandum accompanying the Bill that introduced the rewritten provision. An analysis of s 51(1) of the ITAA 1936 is thus directly relevant to the interpretation of s 8-1 of the ITAA 1997 as well: see The Laws of Australia (The Law Book Company Limited, Looseleaf Ed, Update 59, 8 July 1997) Subtitle 31.4 at paras 1–2. The first limb of s 51(1) provides for a deduction of “losses and outgoings … incurred in gaining or producing the assessable income” [emphasis added]. According to the editors of the Australian Tax Practice Commentary (Butterworths, Looseleaf Ed, Update 339, April 2004) vol 3 at para 51/0170, the definite article “the” is not to be interpreted literally but instead applies ejusdem generis such that it is sufficient (for deduction purposes) if the income arises in the year of income, or is reasonably expected to arise in the future. The fact that no income actually arises as a direct result of the outgoing is no bar to the deductibility of the expense provided there was a reasonable prospect, at the time the outgoing was incurred, of future income being derived: Ronpibon Tin No Liability v Federal Commissioner of Taxation (1949) 78 CLR 47 at 56 (per Latham CJ), John Fairfax & Sons Pty Ltd v Federal Commissioner of Taxation (1959) 101 CLR 30 at 35 (per Dixon CJ), Federal Commissioner of Taxation v Ampol Exploration Ltd (1986) 69 ALR 289 at 303 (per Lockhart J).

42        Although this has been the entrenched interpretation of the deduction formula in Australia, we find the following statement by the authors of The Laws of Australia at Subtitle 31.4, para 7, insightful:

There is no need … to match income and expenses in the same year of income in order to characterise outgoings as allowable deductions (despite the curious role of ‘the’ in the assessable income in the first limb). [emphasis added]

43        It is fundamental rule of statutory interpretation that Parliament shuns tautology and does not legislate in vain: Shell Eastern Petroleum Pte Ltd v Chief Assessor [1999] 1 SLR 441 (“Shell Eastern Petroleum”). In Shell Eastern Petroleum, this court, at [12], accepted the observation of Viscount Simon in Hill v William Hill (Park Lane) Ld [1949] AC 530 at 546–547, that

When the legislature enacts a particular phrase in a statute the presumption is that it is saying something which has not been said immediately before. The rule that a meaning should, if possible, be given to every word in the statute implies that, unless there is good reason to the contrary, the words add something which would not be there if the words were left out.

Therefore, “effect must be given, if possible, to all the words use [in that statue,] for the Legislature is deemed not to waste its words or to say anything in vain”: Quebec Railway, Light, Heat and Power Company, Limited v Vandry [1920] AC 662 at 676 (per Lord Sumner). This was followed by the Privy Council in Enmore Estates Ltd v Darsan [1970] AC 497 at 506 and by the Malaysian Supreme Court in Foo Loke Ying v Television Broadcasts Ltd [1985] 2 MLJ 35. In the latter case, it was held at 43 that:

The court however is not at liberty to treat words in a statute as mere tautology or surplusage unless they are wholly meaningless. On the presumption that Parliament does nothing in vain, the court must endeavour to give significance to every word of an enactment, and it is presumed that if a word or phrase appears in a statute, it was put there for a purpose and must not be disregarded.

44        In our view, there is nothing curious about the word “the” in s 14(1). It means what it says, that there must have been income produced before expenses are deductible. In any event, this court in Andermatt has cautioned against relying on s 51(1) of the Australian ITAA 1936 as its wording is not in pari materia with that of the Singapore provision, nor is the Australian taxation scheme sufficiently similar in respect of the deductibility issue. This court said, at 457, [15] in Andermatt, that:

Moreover the wording of relevant Australian provisions (s 51) are not in pari materia with the Singapore provision. There are, inter alia, these two areas of differences. First, in Australia, a taxpayer’s gross income has to be ascertained before deductions are allowed whereas in Singapore deductions are allowed on each and separate source of income of the taxpayer. In other words there has to be a matching of expenses against each individual source under the Singapore law. In Australia this is not required. Expenses which generate income can be deducted against the cumulative income of the taxpayer. Second, under the Australian Act in view of the words ‘to the extent to which they are incurred’ it is clear that the losses and outgoings are apportionable, which does not appear to be possible under our law. Therefore, we think the Australian cases should always be considered bearing the differences in mind.

45        Finally, comparison by the taxpayer with the Scottish case of Vallambrosa Rubber Co, Ltd v Farmer (Surveyor of Taxes) (1910) 5 TC 529 (“Vallambrosa”) is, in our view, also inappropriate. The judge in the High Court below, in her pre-eminence, had said at [25] of her judgment ([1] supra) that it was “[i]n the context of s 10(1)(d)” that “[the Comptroller’s counsel] was right when he said that [the taxpayer’s counsel’s] submissions contradicted s 14(1) which required the interest expense to be incurred in the same period in which income was produced”. We note that Vallambrosa dealt purely with income from trade or business (which, in the local context, falls under s 10(1)(a)), whereas the taxpayer in the present appeal received passive income from its investment holdings taxable under s 10(1)(d). The distinction is significant because certain tax restrictions are not visited on s 10(1)(a)-sourced income. For instance, only s 10(1)(a) losses can be set off against statutory income for the same year of assessment (see s 37(3)(a)), and unabsorbed s 10(1)(a) losses can be carried forward for set-off against statutory income of future years of assessment upon satisfying certain criteria in s 37(5). For non-s 10(1)(a) sources, s 14(1) requires a source-by-source concept to be applied, whereby the expenses incurred in the production of each source of income must be scrupulously matched against the source, ie, expenses are not deductible on an aggregated basis against the total of all sources of income nor can they be set off against statutory income. The treatment of s 10(1)(a) income therefore differs somewhat from that of non-s 10(1)(a) income. In the light of these circumstances, it would be more appropriate to refer to authorities that deal with s 10(1)(d) of the Act.

46        From our analysis of the foreign authorities above, we found them to be of little assistance in this appeal. We think that the right approach is to look squarely at the provision itself.

The deduction formula in s 14(1)

47        Section 14(1) does not merely refer to expenses that are incurred in the production of income, but to expenses incurred in the production of “the income” and in acquiring “the income”. The judge below therefore rightly noted that pursuant to s 14(1), the interest expenses are specific to “the income” and not “any income” chargeable with tax: see [22] of the High Court decision ([1] supra). The High Court in Andermatt Investments Pte Ltd v Comptroller of Income Tax ([33] supra) expressed a similar view when it pronounced, at 72, [18], that:

[T]he deductibility [of interest expense] is subject to a further condition. The Comptroller has to be satisfied that the interest was payable on capital employed in acquiring the income. While I accept that for the interest to be deductible from the rent income some relationship between the rent income and the money borrowed may be necessary I do not agree that any relationship, however substantial, is necessarily sufficient. There has to be a particular kind of relationship. The link is that provided in s 14(1)(a). To be deductible the interest must be payable on capital employed in acquiring and in acquiring the income and not just income or any income. [emphasis in original]

Later, the High Court also held, at 73, [21], that:

The expression ‘the income’ in s 14(1)(a) refers to the income from a specific source for a specific period which is to be ascertained and not the total income from all sources for that period. It refers to income from that source for that period in acquiring which such capital has been employed.

This was also upheld in Andermatt where this court, at [21], endorsed the view that under s 14(1)(a), there must be a direct link between the money borrowed and the income produced. In our view, this link would clearly be broken if the dividend income arose from a share investment counter which purchase was not funded by the loan for which the related interest expense was sought to be deducted.

48        We therefore fully endorse the following interpretation of s 14(1) in Butterworths’ Annotated Statutes of Singapore, vol 9(2) (Butterworths Asia, 2000 issue), at pp 573–574:

Where the taxpayer has borrowed funds and paid interest on these funds, and invested these funds, these investments must produce income for the interest expense to be deductible. Where the investment does not produce income, the interest expense on those investments would not be deductible.

Whether the Comptroller exercised its administrative discretion appropriately

49        Section 14(1)(a) of the Act provides that for interest expenses to be deductible, the Comptroller must be “satisfied that the interest was payable on capital employed in acquiring the income”. Some degree of administrative discretion is therefore conferred on the Comptroller, who has to be satisfied that the interest expended on borrowings was incurred to earn the dividend income. We are of the opinion that such discretion, as has been acknowledged by the Comptroller, does not give the latter carte blanche authority but is instead to be exercised in good faith and in the interests of good administration.

50        In our view, the Comptroller has shown himself to be measured in his approach towards exercising the administrative discretion granted under s 14(1). In the present case, the Comptroller chose to satisfy himself that the expenses allowed for deduction had been incurred to produce the dividend income by adopting the Formula to ascertain the proportion of interest expenses attributable to the income-producing and non-income-producing share investment counters. We are of the view that such use of the Formula by the revenue authority is valid provided it is not unreasonable in the Wednesbury sense (Associated Provincial Picture Houses, Limited v Wednesbury Corporation [1948] 1 KB 223; Council of Civil Service Unions v Minister for the Civil Service [1985] AC 374). The cases show that the adoption of a general policy by a body exercising an administrative discretion is perfectly valid provided that, amongst other important considerations, it is not a decision that is so outrageous in its defiance of logic or accepted moral standards that no sensible person who applied his mind to the question to be decided could have arrived at it or that no reasonable person could have come to such a view: Lines International Holding (S) Pte Ltd v Singapore Tourist Promotion Board [1997] 2 SLR 584 at [78] per Judith Prakash J. Further, Prakash J in the same case opined that in considering unreasonableness in the Wednesbury sense, the courts are not entitled to substitute their views of how the discretion should be exercised with that actually taken (see Chan Hiang Leng Colin v PP [1994] 3 SLR 662), nor is unreasonableness established if the courts merely come to the view that such a policy or guideline may not work as effectively as another since the courts are not exercising an appellate function in respect of administrative decisions. In our view, the Formula was a logical mathematical method and not some arbitrary or unreasonable measure.

51        It serves to bear in mind that the Comptroller has, in other ways, demonstrated his measured approach in exercising his administrative discretion. The Comptroller has, as a matter of practice, made some concessions as concerns the deductibility of expenses against non-s 10(1)(a) income. One such concession is that in respect of dividends, once a share investment counter has generated dividend income in a particular year, the Comptroller will deem it to be “income-producing” and the interest expense attributable to that counter will be allowed as a deduction for subsequent years even if it does not generate any more dividends in the later years. For instance, in the present appeal, one of the share investment counters declared dividends from years of assessment 1985 to 1988 and from years of assessment 1994 to 1996. During years of assessment 1989 to 1993, even though no dividends were declared from that particular share investment counter, interest expenses attributable to it were nevertheless allowed for deduction. The Comptroller, however, disallowed for deduction interest expenses attributable to share counters that have never once produced any dividend income in the relevant years of assessment 1985 to 1996. In the light of this, we are satisfied that the Comptroller is mindful of the practical tax consequences faced by taxpayers and has not been shown to have abused the administrative discretion conferred upon him by statute.

Whether s 10E of the Act is intended to be a declaration of the law as it has been or a change in the law

52        In contrast to s 14(1), s 10E, an addition to the Act, expressly declares that in the context of investment holding companies, where the expenditure incurred is not attributable to any income produced, such expenditure is not allowed for deduction. It provides:

Ascertainment of income of investment holding company

10E.—(1) Notwithstanding any other provisions of this Act, in determining the income of a company or trustee of a property trust derived from any business of the making of investments the following provisions shall apply:

(a)        any outgoings and expenses incurred by the company or trustee of a property trust in respect of investments of that business which do not produce any income shall not be allowed as a deduction under section 14 for that business or other income of the company or trustee of a property trust;

…

53        Counsel for the taxpayer then made the ingenuous argument that, the taxpayer being a company “in the business of making investments”, s 10E was applicable to it, but since s 10E only came into effect in 1996, it must have meant that prior to that, nothing in the Act permitted the Comptroller to discriminate between the year’s income-producing and non-income-producing share counters.

54        Despite the attractiveness of this argument, we remain unpersuaded by it in view of the legislative intention inferred from the history behind the inclusion of s 10E in the Act. Section 10E was inserted via the Income Tax (Amendment) Act 1995 (Act No 32 of 1995). Reference to the Explanatory Statement to the Income Tax (Amendment) Bill (Bill No 28 of 1995) does not assist, since it does not describe the inspiration for s 10E. It only states that s 10E is meant to “provide the manner for determining the income of a company derived from any business of the making of the investments”. However, the parliamentary reports have been more helpful. In his reading of the Bill for a Second Time, the then Minister for Finance, Dr Richard Hu Tsu Tau, said (Singapore Parliamentary Debates, Official Report (27 September 1995) vol 64 at col 1548):

[C]lause 6 inserts a new section 10E to the Act to clarify the circumstances under which expenses incurred by a company carrying on the business of holding investment are deductible under the Act. [emphasis added]

According to The Oxford English Dictionary (Clarendon Press, 2nd Ed, 1989), “clarify” means, inter alia, “[t]o set forth clearly, declare”, or “[t]o clear (the mind, etc.) from ignorance, misconception, or error; to rectify”. If the insertion were meant to correct a previous position of the law, however, there was no explanation of the rationale or justification behind this change. Taking into account these various factors, we are of the view that it is much more likely that the purpose of s 10E was merely meant to confirm the prevailing legal position as it is, rather than to lay down a change in the deduction formula for expenses incurred by investment holding companies. In other words, it is more probable that the legislative intention has always been to disallow deduction for expenses incurred by investment holding companies in respect of investments which do not produce any income, and neither the law nor Parliament’s intention changed with the insertion of s 10E. Section 10E was enacted to lay down this rule for the avoidance of doubt. Consequently, this contention must fail.

Whether situations where expenses are not deductible are confined only to those where expenses are incurred in the gaining of tax-exempt income or where the investment or capital itself has been extinguished

55        Counsel for the taxpayer further argued that he was unaware of any situation where interest expenses incurred on an investment would be disregarded if there was no return on the investment, with the exception of only two situations: where the income earned is tax-exempt, or where the investment itself has ceased to exist. Counsel for the taxpayer introduced the example of how in Hong Kong, investment holding companies generally pay no tax on dividend income received (as dividends are tax-exempt in Hong Kong) so that the related interest expenses are generally non-deductible. The taxpayer also cited the Indian case of Seth Shiv Prasad ([26] supra) to illustrate a situation where a taxpayer continued to service interest expenses even after the entire investment had ceased to exist.

56        We did not regard this last ground of appeal as having much merit. Whilst the instances brought up by counsel for the taxpayer where interest expenses are not deductible are valid, there is no rule that our tax jurisdiction should be restricted only to those instances of non-deductibility. Indeed, the mere fact that these precedents have occurred in other tax jurisdictions does not mean interest expenses are not deductible only in these two situations. Despite the fact that there may be tax principles common across jurisdictions, each country’s taxation system is different and one cannot blindly pick foreign precedents conveniently just to suit one’s purposes. Ultimately, one must return to the wording of the relevant tax provision to gauge whether a particular tax treatment is legally justifiable.

In camera proceedings

57        There remains to be said one more point. Shortly before the hearing, counsel for the taxpayer sought our assurance that the hearing would be held in camera, or closed to the general public. We considered the appropriate law and accepted his request. We think it opportune to set down once and for all the law relating to the hearing of tax appeals before the Court of Appeal in camera.

58        There is apparently no clear statutory provision or case law providing a direct answer as to whether a tax appeal before the Court of Appeal should be held in camera. Generally, proceedings are held in open court (see s 8(1) of the Supreme Court of Judicature Act (Cap 322, 1999 Rev Ed) (“SCJA”)), unless it is proper, in the opinion of the court, for the court to exercise its discretion to hear them in camera, or where statute provides that the hearings are to be held in camera: see Jeffrey Pinsler, Singapore Court Practice 2005 (LexisNexis, 2005) para 38/1/1. The latter situation usually pertains to actions in which the maintenance of privacy or secrecy is a primary consideration.

59        The relevant statute in the present appeal is of course the Income Tax Act (Cap 134, 2004 Rev Ed). It is apposite at this juncture to refer to the provision in the Act which stipulates when proceedings are to be held in camera. This is s 83 which is set out here as follows:

Proceedings before Board and High Court

83. —(1) Subject to subsections (2) and (3), all proceedings before the Board and in appeals to, or in cases stated for the opinion of, the High Court under the provisions of this Part, and in appeals from decisions of the High Court under section 81 (5) shall be heard in camera.

(2)        Where the Comptroller or the taxpayer applies to the Board or the High Court, as the case may be, that the proceedings be heard by way of a hearing open to the public, the Board or the Court may direct that the proceedings be so heard, notwithstanding any objection from the other party to the proceedings.

(3)        Where in the opinion of the Board or the High Court any proceedings heard in camera ought to be reported, the Board or the Court may publish or authorise the publication of the facts of the case, the arguments and the decision relating to these proceedings without disclosing the name of the taxpayer concerned.

[emphasis added]

It is clear from s 83(1) that with regard to proceedings before the Board of Review and the High Court at least, such proceedings are to be held in camera. As for proceedings before the Court of Appeal, there is no blanket rule that all such appeals must be held in camera. Indeed, s 83(1) only stipulates that “appeals from decisions of the High Court under section 81(5) shall be heard in camera” [emphasis added]. Section 81(5), in turn, provides that:

There shall be such further right of appeal from decisions of the High Court … as exists in the case of decisions made by that Court in the exercise of its original civil jurisdiction. [emphasis added].

It must mean, therefore, that s 83(1), read with s 81(5), provides that appeals before the Court of Appeal will only be held in camera if the High Court had heard the case at first instance, that is, if the High Court had heard the case in the exercise of its original civil jurisdiction. The Act remains silent as to the case where the taxpayer appeals against the decision of the High Court exercising its appellate jurisdiction, as is the situation in the present appeal where the case was first heard by the Board of Review.

60        There has apparently also not been any discussion of this in the local decisions, although it was noted in Butterworths’ Annotated Statutes of Singapore, vol 9(2) ([48] supra) at p 914, that “[i]n line with the confidentiality accorded to tax and the financial matters of the appellant, all hearings before the Board of Review and the appellate courts are to be heard in camera”.

61        Despite the lack of statutory direction, we are of the view that the special nature of tax cases and the spirit of s 83(1) of the Act ought to be taken into consideration in deciding whether the hearing of all tax appeals before the Court of Appeal should be held in camera. It has generally been observed that income tax cases should be heard in camera so as not to expose the appellant taxpayer’s taxation and financial affairs to the peril of public scrutiny: Hall v Inland Revenue Commissioner (NZ) (1979) 9 ATR 595. Indeed, such information may be particularly sensitive when it relates to a listed company, such as the taxpayer in the present appeal. This would consequently warrant the preservation of secrecy.

62        The tenor of the Act also seems to point towards Parliament’s understanding of the need to preserve the secrecy of a taxpayer’s taxation matters. In particular, s 83(3) of the Act supports this view. It provides that:

Where in the opinion of the Board or the High Court any proceedings heard in camera ought to be reported, the Board or the Court may publish or authorise the publication of the facts of the case, the arguments and the decision relating to these proceedings without disclosing the name of the taxpayer concerned. [emphasis added]

63        Therefore, many judgments of tax cases in Singapore have, prior to being reported, undergone amendments pursuant to s 83(3) of the Act so that the taxpayer cannot be identified. A most recent example is T Ltd v Comptroller of Income Tax ([33] supra), where Andrew Ang J at [2] noted that the details of the judgment had been duly amended “so as not to reveal the identity of the appellant”. It is obvious that the spirit of the Act is to enable the taxpayer to maintain the privacy of his taxation affairs.

64        Further, there does not seem to be any rationale for excluding the Court of Appeal from the purview of s 83(1) of the Act in relation to cases which were heard before the Board of Review at first instance. No relevant explanation was given during the Second Reading of the Income Tax (Amendment) Bill of 29 June 1977, by which s 83(1) was amended: see Singapore Parliamentary Debates, Official Report (29 June 1977) vol 37 at col 73. The amendment only related to the vesting in the Comptroller of the reciprocal right to apply to the Board of Review or the High Court for appeal cases to be heard in public instead of in camera, whereas prior to this, only taxpayers could make such an application.

65        In any case, we think that it would defeat the purpose of compulsorily requiring proceedings before the Board and the High Court to be held in camera if the same rule is not applied to all proceedings before the Court of Appeal regardless of where the case originated. There is little justification for protecting the taxpayer’s identity or confidential tax matters only in appeals before the Board or the High Court, but not before the Court of Appeal.

66        Even if we are wrong, we note that the Court of Appeal always has the discretion to hear an appeal in camera if this court is satisfied that it is “expedient in the interests of justice, public safety, public security or propriety, or for other sufficient reason to do so” [emphasis added]: see s 8(2) of the SCJA, which applies to both the High Court and the Court of Appeal. We think that on the facts of this case, it is necessary to hold the hearing in camera so as to protect the secrecy of the appellant’s taxation affairs.

67        In the result, the appeal is dismissed with costs. The security for costs shall be paid out to the respondent to account of costs.