|
Citation: 2004TCC207
|
|
Date: 20040310
|
|
Docket: 2002-2257(IT)G
|
|
BETWEEN:
|
|
IMPERIAL OIL LIMITED,
|
|
Appellant,
|
|
and
|
|
|
|
HER MAJESTY THE QUEEN,
|
|
Respondent.
|
REASONS FOR DETERMINATION
Miller J.
[1] This is a Reference to the Tax
Court of Canada to have certain questions raised in the appeal
determined by the Court pursuant to subsection 173(1) of the
Income Tax Act (the Act). The questions involve the
proper application of paragraph 20(1)(f) and subsection
39(2) of the Act in relation to foreign exchange
losses.
[2] In 1989, Imperial Oil issued US
debentures for a total face amount of US$300,000,000, issued at a
discount of 1.199%. Ten years later, at the time of redemption of
part of the debentures with a face value of US$87,130,000, the
US dollar had appreciated from a rate of C$1.1766 to US$1 to
a rate of C$1.48192 to US$1 resulting in a loss on redemption of
C$27,831,712 (see Appendix "A"). The Appellant
argues this entire amount is deductible under paragraph
20(1)(f) in computing Imperial Oil's 1999 income (see
Appendix "B"). The Minister of National Revenue (the
Minister) says that only C$1,229,181 is deductible under
paragraph 20(1)(f), as only it reflects the original issue
discount, with the balance being a capital loss under subsection
39(2) (see Appendix "C").
[3] The Reference is put in terms of
three questions:
(a) What portion, if any, of the
said C$27,831,712 is deductible under subparagraph
20(1)(f)(i) of the Act?
(b) What portion, if any, of the said
C$27,831,712 is deductible under subparagraph 20(1(f)(ii)
of the Act?
(c) What portion, if any, of the said
$27,831,712 is a capital loss under subsection 39(2) of the
Act?
[4] Paragraph 20(1)(f) must be
interpreted in its totality, and, to rely on the oft-cited
approach of Professor Elmer A. Driedger, in context, harmoniously
with the scheme and objectives of the Act and the intent
of Parliament. Both Parties' positions are premised on what
they call the "Gaynor principle", which, they
both maintain, requires a conversion of each element in the
paragraph 20(1)(f) formula into Canadian dollars. Based on
that principle, they then limit their disagreement to which
exchange rate (1989 or 1999?) to apply to the term -
"principal amount of the obligation". I question this
reliance on Gaynor and review both Gaynor and UK
authorities to find that the Parties place too broad an
interpretation on that case.
[5] I illustrate that, by
intermingling exchange rates in the paragraph 20(1)(f)
formula, the Parties inject ambiguity into the provision. It then
becomes necessary to look beyond the ordinary and grammatical
wording to the object and scheme of the Act and the
intention of Parliament. Following this approach, I conclude the
provision was not intended to provide a deduction for foreign
exchange losses on capital account. The formula in the provision
should be calculated in US dollars with the result converted into
Canadian dollars, or, in the alternative, the formula should be
calculated entirely in 1999 Canadian dollars (see Appendix
"D"). Either of these approaches yields a finding that
paragraph 20(1)(f) applies to allow Imperial Oil a
deduction of the 1.199% discount in 1999 Canadian dollars, or
C$1,548,145, with the balance being a capital loss under
subsection 39(2).
Facts
[6] The facts the parties agreed to
for this Reference are not lengthy, so I will simply reproduce
the relevant portions:[1]
(1) At all relevant
times the Appellant was a company whose principal business was
the refining, marketing and transportation of petroleum and
petroleum products.
(2) On October 16,
1989 the Appellant issued three sinking fund debentures of
U.S. $100,000,000 each, for a total of U.S. $300,000,000 at
an interest rate of 8.75% per annum, repayable in 2019.
(3) The said
debentures were issued under an indenture dated October 15, 1989
between the Appellant and the Bank of Montreal Trust Company (the
"Indenture").
...
(5) The said
debentures were issued to the public at less than their face
amount, namely, at a discount of 1.199%, amounting to
U.S. $3,597,000 in the aggregate (the "Discount").
The amount received from the public on October 16, 1989,
expressed as a percentage of the face amount of the debentures,
was therefore 98.801% or U.S. $296,403,000 expressed as an
aggregate dollar amount (the "Discounted Amount").
(6) The yield from
the debentures was 8.86%
(7) After the
deduction of underwriters' fees of U.S. $2,625,000 from the
proceeds of the Discounted Amount, the Appellant received a net
amount of U.S. $293,778,000 from the issuance of the said
debentures.
...
(9) The Appellant
deposited the said U.S. $293,778,000 with the Royal Bank of
Canada at New York, N.Y. and used that amount as part of its
refinancing of the purchase of the shares of Texaco Canada
Inc.
(10) Under the terms of the said
debentures, interest was payable in U.S. dollars
semi-annually on each of April 15 and October 15, beginning on
April 15, 1990. The Appellant deducted the Canadian dollar
equivalents of those interest payments pursuant to paragraph
20(1)(c) of the Act. No issue arises between the
Appellant and the Respondent regarding the deductibility of that
interest.
(11) Under the terms of the said
debentures, all or part of the debentures were redeemable at the
election of the Appellant at any time on or after October 15,
1999, at predetermined redemption prices.
(12) On October 15, 1999, the
Appellant elected to redeem part of its U.S. $300,000,000
debentures with a face amount of U.S. $87,130,000 (the
"Redeemed Amount").
...
(14) The portion of the Discount
that pertained to the Redeemed Amount was U.S. $1,044,689
(U.S. $87,130,000 x .01199), and the portion of the Discounted
Amount that pertained to the Redeemed Amount was
U.S. $86,085,311.
(15) The Bank of Canada noon
rate for conversion from U.S. to Canadian dollars on October 16,
1989 was 1.17660. Accordingly, the Canadian dollar equivalent of
the Redeemed Amount on that day was Cdn. $102,517,158 (U.S.
$87,130,000 x 1.17660). The Canadian dollar equivalent of the
portion of the Discounted Amount that pertained to the Redeemed
Amount was Cdn. $101,287,977 (U.S. $86,085,311 x
1.17660).
(16) On October 15, 1999 the
Appellant purchased U.S. $87,130,000 at a cost of Cdn
$129,119,689 (U.S. $87,130,000 x 1.48192) in order to pay the
Redeemed Amount. No issue arises between the Appellant and the
Respondent regarding the correctness of the rate of exchange of
1.48192 between the U.S. and Canadian dollars on that day.
Analysis
Introduction
[7] I will start by reproducing
subparagraph 20(1)(f)(i):
20(1) Notwithstanding paragraphs
18(1)(a), (b) and (h), in computing a
taxpayer's income for a taxation year from a business or
property, there may be deducted such of the following amounts as
are wholly applicable to that source or such part of the
following amounts as may reasonably be regarded as applicable
thereto:
...
(f) an amount
paid in the year in satisfaction of the principal amount of any
bond, debenture, bill, note, mortgage, hypothecary claim or
similar obligation issued by the taxpayer after June 18, 1971 on
which interest was stipulated to be payable, to the extent that
the amount so paid does not exceed,
(i) in
any case where the obligation was issued for an amount not less
than 97% of its principal amount, and the yield from the
obligation, expressed in terms of an annual rate on the amount
for which the obligation was issued (which annual rate shall, if
the terms of the obligation or any agreement relating thereto
conferred on its holder a right to demand payment of the
principal amount of the obligation or the amount outstanding as
or on account of its principal amount, as the case may be, before
the maturity of the obligation, be calculated on the basis of the
yield that produces the highest annual rate obtainable either on
the maturity of the obligation or conditional on the exercise of
any such right) does not exceed 4/3 of the interest stipulated to
be payable on the obligation, expressed in terms of an annual
rate on
(A) the principal amount
of the obligation, if no amount is payable on account of the
principal amount before the maturity of the obligation, or
(B) the amount outstanding
from time to time as or on account of the principal amount of the
obligation, in any other case,
the amount by which the lesser of the principal amount of
the obligation and all amounts paid in the year or in
any preceding year in satisfaction of its principal amount
exceeds the amount for which the obligation was issued,
and
(ii) ...
[8] The Parties are agreed that the
threshold tests set forth in subparagraph 20(1)(f)(i)
have been met in this case, and it is therefore that subsection,
and not subsection 20(1)(f)(ii), which is at issue in this
Reference.
[9] While there are 22 references in
this section to the term "amount", only the following
"amounts" are of real concern in this Reference:
(A) "principal amount of the
obligation";
(B) "all amounts paid in the year in
satisfaction of the principal amount";
(C) "amount for which the debentures
were issued".
The formula contained in paragraph 20(1)(f) can be put
mathematically as follows:
X = (lesser of A or B) - C
X is the deductible amount we are trying to ascertain. There
is no dispute that amount must be measured in Canadian
dollars.
A is the principal amount
of the obligation. There is no dispute that amount was
US$87,130,000. There is a dispute, indeed the only dispute
according to the Parties, as to the Canadian dollar equivalent,
whether it is in 1989 Canadian dollars ($102,517,158) or in 1999
Canadian dollars ($129,119,689).
B is the amount paid in
the year in satisfaction of the principal amount. There is no
dispute that amount was US$87,130,000 and C$129,119,689.
C is the amount for which
the obligation was issued. There is no dispute that was
US$86,085,311. There is also no dispute between the Parties that
this was C$101,287,977, given the 1.17660 exchange rate at the
time of issuance. I will have more to say on that later.
[10] The problem arises because, although
the principal amount remained constant in US dollars, the
exchange rate shifted dramatically from the time of issuance to
the time of redemption.
The Parties' Positions
[11] Both Parties have attempted to justify
the appropriate exchange rate to be applied to A. The Appellant
says A and B are the same; that is, the principal amount of the
obligation and the amount paid in satisfaction of the principal
amount are both to be determined at the time of redemption, when
the exchange rate was 1.48192. The Minister contends that the
principal amount of the obligation is set at the time of
issuance, that is US$87,130,000 at the 1.17660 exchange rate, and
never changes. The results, of course, are drastically different,
as seen from Appendices "B" and "C".
[12] The Appellant contends his position
follows the basic principles of statutory interpretation
recognized by Professor Driedger and considers the context of the
phrase "principal amount of the obligation". Mr. Meghji
also applies a strong dose of logic. He further seeks support
from Steffen E. Waltz v. The Queen,[2] where it was
determined, that in the absence of a statutory provision to the
contrary, "principal amounts" must be computed at the
time of disposition, or in this case, the time of redemption. The
Appellant relies also on the definition of "principal
amount" in subsection 248(1) of the Act which refers
to the "maximum amount payable in respect of the
obligation". Finally, the Appellant points to the Government
of Canada's own administrative policies by reference to
advance rulings and technical interpretations issued by Canada
Customs and Revenue Agency (CCRA) on the application of paragraph
20(1)(f) to other obligations, exchangeable debentures,
commodity-based loans and convertible debentures. In all such
cases, "principal amount" was computed at the time of
repayment. For reasons that will become apparent, I am not going
into Mr. Meghji's arguments in great detail, other than
to indicate that I found his arguments persuasive for this
particular interpretation of the term "principal
amount".
[13] The Respondent follows a similar path
as the Appellant in that he too deems it necessary to argue what
rate must be applied to "principal amount"; the rate at
the time of issuance, or the rate at the time of redemption. At
the time of creation says the Respondent. He seeks support in
section 248(26) which reads:
248(26) For greater certainty, where at any time a person or
partnership (in this subsection referred to as the "debtor")
becomes liable to repay money borrowed by the debtor or becomes
liable to pay an amount (other than interest)
(a) as
consideration for any property acquired by the debtor or services
rendered to the debtor, or
(b) that is
deductible in computing the debtor's income,
for the purposes of applying the provisions of this Act
relating to the treatment of the debtor in respect of the
liability, the liability shall be considered to be an obligation,
issued at that time by the debtor, that has a principal amount at
that time equal to the amount of the liability at that time.
[14] This provision does not go as far as
the Respondent suggests; it only indicates what the principal
amount is at the time of issuance. It does not stipulate that the
principal amount shall forever be the principal amount determined
at the rate at the time of issuance of a foreign debt obligation.
Also, Technical Notes suggest that this section's purpose had
more to do with timing of the issuance of an obligation for
purposes of section 80, than for any potential application to
subsection 20(1)(f).[3] But, the Respondent argues, relying on his
approach leads to a logical result; that is, only the cost caused
by Imperial Oil obligating itself to pay more to its creditor on
the debt than it received, is the deductible obligation expense.
The loss from the foreign exchange rate change is just that, from
the foreign exchange, not from the contractual discount.
[15] The Respondent then shifts to a review
of subsection 39(2) which specifically deals with gains and
losses sustained by foreign rate fluctuations. It reads as
follows:
39(1) For the purposes of this
Act,
...
(2) Notwithstanding
subsection (1), where, by virtue of any fluctuation after 1971 in
the value of the currency or currencies of one or more countries
other than Canada relative to Canadian currency, a taxpayer has
made a gain or sustained a loss in a taxation year, the following
rules apply:
(a) the
amount, if any, by which
(i) the total
of all such gains made by the taxpayer in the year (to the extent
of the amounts thereof that would not, if section 3 were read in
the manner described in paragraph (1)(a) of this section, be
included in computing the taxpayer's income for the year or
any other taxation year)
exceeds
(ii) the total of
all such losses sustained by the taxpayer in the year (to the
extent of the amounts thereof that would not, if section 3 were
read in the manner described in paragraph (1)(a) of this section,
be deductible in computing the taxpayer's income for the year
or any other taxation year), and
(iii) if the taxpayer is
an individual, $200,
shall be deemed to be a capital gain of the taxpayer for the
year from the disposition of currency of a country other than
Canada, the amount of which capital gain is the amount determined
under this paragraph; and
(b) the
amount, if any, by which
(i) the total
determined under subparagraph (a)(ii),
exceeds
(ii) the total
determined under subparagraph (a)(i), and
(iii) if the taxpayer is
an individual, $200,
shall be deemed to be a capital loss of the taxpayer for the
year from the disposition of currency of a country other than
Canada, the amount of which capital loss is the amount determined
under this paragraph.
It is not difficult to follow the Respondent's logic in
concluding that a foreign exchange loss is the difference between
the Canadian dollar value of the redeemed amount (US$87,130,000)
at the time of redemption, being C$129,119,689 (at the 1.48192
rate) and the Canadian dollar value of the redeemed amount at the
time of issuance, being US$87,130,000 or C$102,517,158 (at the
1.17660 rate). The difference is C$26,602,531.
[16] That approach however puts the cart
before the horse. In the ordering of the Act, foreign
exchange losses under subsection 39(2) only follow after the
appropriate deduction under section 20 has been taken. So, I do
not accept that one can calculate what may seem logical under
subsection 39(2) and work back to limit a paragraph
20(1)(f) deduction. This is not a matter of specific
provisions overriding general provisions, but is more a matter of
how the Act provides a road map for the computation of
income, primarily in section 3. Deductions under paragraph
20(1)(f) are taken into account under paragraph
3(a). Capital gains and losses under subsection 39(2) are
taken into account under paragraph 3(b). In effect,
losses deductible on income account are specifically excluded
from the scope of subsection 39(2). All to say, I must first
determine the proper deductible amount pursuant to paragraph
20(1)(f). If the whole loss is not deductible, only then
may I turn to subsection 39(2).
[17] What is the correct interpretation of
paragraph 20(1)(f)? The familiar and well-used modern
approach espoused by Professor Driedger and adopted by the
Supreme Court of Canada guides me to consider the words in their
entire context, and in their grammatical and ordinary sense,
harmoniously with the scheme of the Act, the object of the
Act, and the intention of Parliament.
[18] What is apparent to me is that there is
no difficulty or ambiguity reading paragraph 20(1)(f)
in the grammatical and ordinary sense, and in the context of a
deduction for obligations issued at a discount, if the
calculation is simply done in US dollars, and only the result
converted into Canadian dollars at the rate at the time of
redemption. The calculation would then be as set forth in
Alternative 1 in Appendix "D".
[19] I presented this simpler, and what I
believed and still do believe to be, sensible reading of
paragraph 20(1)(f) to the Parties. I was met with a stone
wall from both counsel. They agreed that principles of statutory
interpretation precluded me from taking this ordinary and
grammatical reading of the words. I, therefore, asked them to
provide written submissions as to what were the principles that
would so readily thwart my purpose. Their responses have fallen
short of convincing me that there are any principles of such
universal application that would take an otherwise
straightforward provision and turn it into an interpretative
minefield.
[20] Before justifying my interpretation, I
wish to comment on the Appellant's and Respondent's
interpretation. They both calculate the paragraph 20(1)(f)
formula, that is the determination of A, B and C, by
intermingling exchange rates. A basic understanding of math will
quickly reveal that this necessarily ensures a foreign exchange
gain or loss becomes an integral part of the
paragraph 20(1)(f) calculation. It does not surprise
me that the Appellant takes this view, as Mr. Meghji's whole
point is that the foreign exchange loss is a cost of borrowing,
otherwise on capital account, which was intended to be deductible
under paragraph 20(1)(f), similar to other costs of
borrowing. It does surprise me, however that the Respondent takes
the view that paragraph 20(1)(f) is not intended to be
impacted by foreign exchange, yet the Respondent also goes on to
intermingle exchange rates in his interpretation of the formula.
The flaw in the Respondent's approach is apparent when the
facts are changed, such that the Canadian dollar increases rather
than decreases in value. Not surprisingly, the result, following
Mr. Meghji's interpretation, would be that the effect of the
foreign exchange gain would completely wipe out the availability
of any deductible discount pursuant to paragraph 20(1)(f).
I have attached Appendix "E" to illustrate this result.
But, applying the Respondent's interpretation to the same
situation yields the same result. A taxpayer gets no deduction
for a discount. (See Number 2 in Appendix "E".)
[21] Appendix "E" also illustrates
the effect of reading paragraph 20(1)(f), as I suggest, in
its ordinary and grammatical sense and simply converting the
result. (Number 3 - Alternative (i)). It is apparent the
taxpayer still gets a deductible discount, but then faces a
foreign exchange capital gain subject to subsection 39(2).
This appears to be the only approach where economic and legal
reality so neatly mesh.
[22] In the Appendix "E" example,
there was a cost of borrowing arising from the discount. It does
not make sense to eliminate that legitimate deductible cost by
applying interpretative principles that simply are not
appropriate for this provision. The legal reality, I suggest, is
that paragraph 20(1)(f) was never intended to, nor on an
ordinary meaning approach does it make sense to, include the
foreign exchange gains or losses in its calculation. I will have
more to say on Parliament's intention with respect to this
section shortly.
The GaynorPrinciple
[23] I now turn to the principles that the
Parties agree preclude me from calculating the formula in US
dollars and converting the result. The primary reason cited to me
was the principle of statutory interpretation both Parties
extracted from the case of Gaynor. The Respondent
explained the principle as follows:[4]
The Appellant and Respondent are in agreement ... . This
principle requires that under the Act all amounts must be
reported in Canadian dollars, because tax is payable in Canadian
dollars, and because that tax must be calculated in Canadian
dollars. Accordingly, every transaction involving foreign
currency that results in an asset, liability, revenue item or
expense must be converted into Canadian dollars at the rate of
exchange on the date on which it came into existence. The
Gaynor principle is therefore transaction-oriented, rather
than end-results oriented.
Mr. Chambers also cites Shell Canada v. The Queen[5] and comments
of Professor Brian Arnold in Timing and Income Taxation: The
Principles of Income Measurement for Tax Purposes[6] in support of this
broad principle. Mr. Meghji supports Mr. Chambers' position
in this regard, and goes on to argue that on a careful reading of
Gaynor, the principle is not limited to capital gains
calculations but extends to the computation of business income.
He cites the following passage of Pratte J.A. of the Federal
Court of Appeal:[7]
... Paragraph 40(1)(a) of the Income Tax Act
makes it clear that the capital gain realized by the appellant in
each case was the 'amount' by which the proceeds of the
disposition of her securities exceeded the adjusted cost base of
those securities. When that provision speaks of the
'amount' of the capital gain, it obviously refers to an
amount expressed in Canadian currency. As that amount is the
result of a comparison between two other amounts, namely, the
amount representing the cost of the securities and the amount
representing the value of the proceeds of disposition, it
necessarily follows that both the cost of the securities and the
value of the proceeds of disposition must be valued in Canadian
currency which is the only monetary standard of value known to
Canadian law. ...
I see nothing in Justice Pratte's comment that broadens
the approach beyond that of the determination of a capital gain.
Quite the opposite.
[24] Mr. Meghji further relies on the
Government's Technical Interpretation 9907935 dated June
7, 1999, wherein CCRA stated:[8]
REASONS: 1) Income as used in section 3 means Canadian dollars
and Gaynor v. The Queen, ... confirms that an amount is
expressed in Canadian currency ... The Act is a Canadian
statute and thus, in general terms, it is our view that a
taxpayer's taxable income should be expressed in Canadian
dollars. The case law supports our view that transactions should
be measured in Canadian currency either at the time of the
transaction, or at a time that produces substantially the same
dollar amounts that would have resulted had the underlying
transaction been converted into Canadian currency on the dates
that they incurred. The Federal Court of Appeal case of Hope
R. Gaynor v. The Queen ... confirmed that an
"amount" refers to an amount expressed in Canadian
currency.
[25] In effect, the Parties' position is
that Gaynor stands for the proposition that wherever the
word "amount" appears in the Act, it must refer
to a Canadian amount. I wish to explore this interpretation of
Gaynor in some further detail.
[26] Based on the decisions in the Tax Court
of Canada and both Federal Courts in Gaynor, and the line
of cases from the United Kingdom that deal with the same issue,
the Federal Court of Appeal's decision in Gaynor does
not set down the general principle of interpretation, as
suggested by the Parties. The Canadian and United Kingdom courts
have held that different considerations apply when converting
from a foreign currency depending on whether one is calculating a
capital gain or income amount. Deduction under paragraph
20(1)(f) of the Act is on account of income, not
capital.
i. Tax
Court of Canada decision in Gaynor[9]
[27] In Gaynor, the Appellant owned a
portfolio of US securities at the time she became resident in
Canada. When she eventually sold her US stock, she argued that
her capital gain should be calculated based on the proceeds of
disposition in US dollars less the US dollar cost of acquisition
(i.e. the US dollar cost on the date she became a resident), and
converting the US dollar gain into Canadian funds at the exchange
rate prevailing at the date of disposition. Sarchuk J. disagreed,
holding that the approach mandated by the legislation was to
convert the cost into Canadian funds at the date she became
resident, and to subtract that amount from the Canadian dollar
proceeds calculated at the time of disposition.
[28] Sarchuk J. also dealt with and
commented on the UK case of Pattison v. Midland
Marine Ltd.,[10] which had been cited for the Appellant, Ms. Gaynor.
In Pattison, a UK bank borrowed US$15 million from its
American parent in order to conduct its banking business in
Europe. The UK Bank had not made any profit other than interest
from re-lending the US funds, however, it had made a notional
(i.e. foreign exchange) profit on the loans since sterling had
declined in value against the US dollar over the period. The
Inland Revenue sought to tax the foreign exchange gain that the
UK Bank had earned on the US dollar loans.
[29] The House of Lords had found for the
taxpayer, Midland Marine, on the grounds that returning to the
debtor the specific subject matter of a loan can never give rise
to a profit to the borrower, notwithstanding the change in value
of the subject matter of the borrowing. Consequently, the foreign
exchange gain was not taxable as profit.[11]
[30] Sarchuk J. distinguished the House of
Lords reasoning in Pattison on the grounds that it did not
apply to the calculation of a capital gain. He stated:
... In the Pattison case monies were borrowed and
repayed(sic). These monies were the asset, this asset was
borrowed by the taxpayer corporation and in due course returned.
In the interim it had increased in value.
The conclusions in Pattison cannot in my view be
applied to the facts before me. Taxation in the case at bar is
dependent upon the determination of capital gains or losses which
are by statutory definition the difference between the adjusted
cost base and the proceeds of disposition. In the Pattison
case there were no dispositions, there were no sales or
purchases. The asset was simply borrowed and increased in value
while the borrower had the asset. That is not the case here where
assets were sold and gains and losses were incurred.[12]
According to Sarchuk J., the disposition made the
difference.
ii. The Federal
Court Trial Division decision in Gaynor[13]
[31] Gaynor proceeded as a trial
de novo before the Federal Court Trial Division, although the
parties agreed to proceed on the evidence that was adduced before
the Tax Court. The headnote to the case reads in part:
... when computing the gain or loss realized on the
disposition of property such as that of the taxpayer, money must
be the measure, and this is so at every stage or step that needs
to be considered under the Act in establishing such a gain
or loss. These gains or losses must be calculated in Canadian
dollars at the relevant time, i.e. at the average exchange rate
prevailing at the time such gain or loss occurs, and such cost or
price is encountered.
Like Sarchuk J. before him, Pinard J. recognized that the case
dealt with a capital gain. He noted:[14]
One must also keep in mind that capital gain should not be
computed according to the same rules as income from a business or
property; indeed, the Income Tax Act provides special
rules for the computation of capital gain (see The Queen v.
Geoffrey Sterling, 85 DTC 5199 at page 5200).
iii. The Federal Court of
Appeal decision in Gaynor[15]
[32] By the time that Gaynor was
decided by the Federal Court of Appeal, the importance of a
disposition to the distinction between capital gains and income
had become a little less obvious. Consider however the comments
cited earlier of Justice Pratte, found in paragraph 23 of these
Reasons.
[33] Contrary to Mr. Meghji's position,
I suggest a careful analysis of this case at all three levels
reveals that there is a difference to be observed between the
conversion of a foreign currency amount for the purpose of
calculating a capital gain, and a conversion for the purpose of
calculating an income amount.
iv. The UK
jurisprudence
[34] There is a line of English cases that
deal with the same issue that the Canadian courts wrestled with
in Gaynor. They all support the view that foreign currency
conversions should be treated differently depending on whether
they are made for the purpose of calculating a capital gain or
income.
(a) Bentley v.
Pike, 53 TC 590.
[35] In Bentley, the High Court of
Justice (Chancery Division) of England dealt with the same issue
that our Federal Court of Appeal addressed in Gaynor, with
the exact same result. Vinelott J. summarized the Appellant
Bentley's position:
Mr. Englard submitted that if the chargeable gain is the
difference between the value ... at the death of Mr. Rosenfeld
and the ... proceeds of sale of the property when it was sold,
the gain should none the less be ascertained by translating
both the value ... at Mr. Rosenfeld's death expressed in
Deutschmarks and the ... proceeds of sale into sterling at the
rate of exchange ruling on 6 July 1973.[16]
In other words, the Appellant argued that she should be
entitled to calculate the capital gain based on the exchange rate
prevailing at the time of disposition, instead of calculating the
cost in sterling at the time of acquisition and subtracting it
from the proceeds calculated in sterling at the time of
disposition.
[36] However, Vinelott J. went on to rule
against the Appellant, stating that:
While I feel some sympathy for Mrs. Bentley, who is in large
measure called on to pay capital gains tax upon a gain resulting
from the devaluation of the pound, I can see no possible
justification in the capital gains tax legislation for this
approach.[17]
This case is on all fours with Gaynor, and takes the
exact same approach to the calculation of a capital gain.
(b) Pattison v. Marine
Midland Ltd., [1984] 1 A.C. 362.
[37] As discussed earlier in
Pattison, a UK bank borrowed US$15 million from its
American parent in order to conduct its banking business in
Europe. The UK Bank had not made any profit other than interest
from re-lending the US funds, however, it had made a notional
(i.e. foreign exchange) profit on the loans since sterling had
declined in value against the US dollar over the period. The
Inland Revenue sought to tax the foreign exchange gain that the
UK Bank had 'earned', arguing that the US dollar
borrowing should be converted to sterling at the exchange rate
prevailing at the time of the borrowing, and subtracted from the
value in sterling of the US dollar borrowing at the time of
repayment.
[38] The Appellant (Marine Midland), on the
other hand, argued that it had not made any profit, since it had
simply returned exactly what it borrowed. It took the position
that the correct approach was to calculate any profit or loss on
the loan in the currency of the loan first, before converting it
to sterling for tax purposes. If that was done, there was no
foreign exchange gain or loss since US$15M was borrowed, and
US$15M was repaid.
[39] The decision of the House of Lords was
given by Lord Templeman, who found for the Appellant. He made
much of the fact that the Appellant had never actually converted
the loan proceeds to sterling, and that the Appellant had always
maintained enough US dollar assets to provide a natural hedge
throughout the term of the loan, which effectively neutralized
any currency fluctuations. In Pattison, the Law Lords took
a practical and realistic approach to the calculation of the
Appellant's profit on its foreign denominated debt.[18] It does not appear
that Bentley was even raised before them.
(c) Capcount Trading v.
Evans, 65 TC 545
[40] In Capcount Trading, the
Appellant took the position that the capital gain it realized on
the disposal of its Canadian company shares should be calculated
by deducting the Canadian dollar cost of the shares from the
Canadian dollar proceeds before converting the Canadian dollar
amount into sterling at the exchange rate prevailing on the date
of disposal. The Inland Revenue argued that the cost of
acquisition and the proceeds of disposition must each be
converted to sterling at the exchange rates prevailing on the
dates of acquisition and disposition, respectively. The Appellant
was arguing against the application of the principle set down by
the English courts in Bentley and our Federal Court of
Appeal in Gaynor, on the basis of the House of Lords
decision in Pattison. The taxpayer appeared to be trying
to exploit the differences between the approach taken by the
House of Lords in Pattison, and the approach taken by
Vinelott J. in Bentley. Lord Justice Nolan observed at
page 558:
... Mr. Park relies in particular upon the judgments given by
this court in a tax case, Pattison v. Marine Midland Ltd,
57 TC 219. This case, in which the decision of the Court of
Appeal was upheld by the House of Lords, was decided after
Bentley v. Pike and, in Mr Park's submission, is
inconsistent with it.
Lord Justice Nolan went on to deny the appeal, finding
that:
... I agree with the conclusion reached by Vinelott J. in
Bentley v. Pike that sterling is "the only
permissible unit of account" for capital gains tax
purposes."[19]
But continued:
Is this conclusion inconsistent with the decision of this
court in Pattison v. Marine Midland Ltd (1983) 57 T.C.
219?
... At first sight, the Revenue claim in the Marine
Midland case bears a resemblance to the claim made against
the Appellant in the present case, but it is to be borne in
mind that the former claim was for tax upon trading profits, that
is to say upon income, whereas the latter is a claim for tax
under the capital gains legislation.[20]
Lord Justice Nolan recognized a difference between income and
capital gains in this context.
[41] The cases suggest that there is a
difference, though none of them truly explain the conceptual
basis for the difference. The capital gains provisions must be
capable of measuring the value of an asset at two different
points in time. Obviously, a consistent unit of measurement will
be required, since, as Lord Justice Nolan aptly points out in
Capcount Trading:
... In applying (the relevant paragraph under UK law) to a
particular acquisition and disposal one cannot, so to speak,
subtract peas from beans.[21]
Since the ultimate result must be an amount measured in
Canadian dollars, and since the unit of measurement must be
consistent over time, it is easy to see why the courts have held
that the acquisition and disposition must be converted into
Canadian currency for the purposes of the computation.
Gaynor provides that for the purpose of calculating a
capital gain, the currency of the computation must be Canadian,
regardless of the currency of the transaction.
[42] However, there is a distinct difference
between the measurement of a capital gain and the measurement of
'ordinary' income. In Capcount Trading, Lord
Justice Nolan also notes that:
... the income tax legislation, unlike the capital gains tax
legislation, is not generally concerned with the measurement of a
gain or loss on a single disposal but with a balance at the year
end computed on accounting principles.[22]
And in finding for the Appellant in Pattison, Lord
Justice Nolan also made a great deal of the fact that there had
been no currency conversion transaction. It was, he held, the
exchange transaction that gave rise to the foreign exchange gain
or loss.
[43] As a result, the key to the different
treatment of capital gains and income in respect of foreign
exchange transactions may be that a capital gain relies on the
acquisition and disposition of the asset. The acquisition and
disposition force (at least conceptually) a currency conversion
transaction to occur.
[44] However, an expense on income amount
(such as the discount under paragraph 20(1)(f) in the
present case) requires a snapshot at the time of payment. There
is no accumulation of wealth over a period of time, requiring a
measurement over time. The formula does not address an increase
or decrease in value: it is only meant to work at a single point
in time to determine a cost, an expense.
[45] I conclude that each "amount"
in this particular expense-related formula need not be converted
to conform to principles of statutory interpretation. It is
sufficient in the application of this provision to convert the
result, the deductible amount, into Canadian dollars; indeed, it
gives clear, ordinary meaning to the provision to do so.
[46] However, if I am wrong in my
application of Gaynor, and I must convert each amount of
the formula into Canadian dollars, then, to achieve the object of
the section, the Canadian dollars must all be at the same rate.
The one time snapshot rate has to be the 1999 rate. When the
formula is calculated on that basis, the result is the same as if
only the result was converted. Further, if the rate at the time
of repayment is used in the example of a Canadian rate increase,
(Appendix "E") again the result is the same as
calculating in US dollars and converting the final figure - a
result, which I have already indicated, is the only result which
meshes legal and economic reality; that is, a deduction for a
legitimate discount, plus a gain on capital account for the
foreign exchange gain.
[47] To put 1999 Canadian dollar amounts to
the case before me, the calculation would be as set forth in
Alternative 2 of Appendix "D".
[48] Does attaching the 1999 exchange rate
to "the amount for which the obligation was issued" in
1989 fly in the face of the ordinary meaning of those words? Not,
I would suggest, if the provision is read in context and in its
totality, with a view to its objective. To simply pluck the
phrase out of the section and give to it a stand-alone status,
then, of course, it becomes difficult to justify attaching any
rate other than the rate at the time of issuance. And, indeed,
that is how the Parties agreed to interpret it; and that is why,
I would suggest, a relatively clear provision has been subjected
to extensive interpretative analysis, with resulting confusion
and disagreement.
[49] The plain and ordinary meaning approach
to paragraph 20(1)(f) of the Act simply does not
work if the foreign exchange conundrum is introduced into this
particular provision by intermingling rates. Only by removing
that confusion by either determining the calculation in the
currency of the obligation and converting the result, or by
determining the calculation in Canadian dollars all at the same
rate, does the provision become susceptible to a plain and
ordinary meaning.
Purpose of paragraph 20(1)(f)
[50] Are my alternative interpretations
supportable considering the provision harmoniously with the
scheme of the Act, the object of the Act, and the
intention of Parliament?
[51] Paragraph 20(1)(f) deals with
obligations issued at less than face value, and allows that
difference on payment to be deducted. Mr. Meghji rightfully
points out that nowhere, other than in the margin, is the word
"discount" used. But neither does this section refer to
foreign exchange losses directly. Yet, by parsing the different
elements of the formula and intermingling rates, a positive
result is achieved for Imperial Oil. I just do not buy it. I have
been provided with no material to suggest that
paragraph 20(1)(f) of the Act was intended by
Parliament to have such an expansive reading.
[52] Analyzing the provision in context and
with a view to object and intention, the question really becomes:
are foreign exchange losses incurred in the course of borrowing
money in a foreign currency a deductible cost of borrowing under
paragraph 20(1)(f) akin to other subsection 20(1)
deductible costs of borrowing? My answer is no. Mr. Meghji
argues, following reasoning in Antosko et al. v. The
Queen,[23] that where the words are clear and
unambiguous, they must be followed notwithstanding the result may
provide a result abhorrent to the Respondent. But Justice
Iacobucci goes on to say in Antosko:[24]
... Where the words of the section are not ambiguous, it is
not for this Court to find that the appellants should be
disentitled to a deduction because they do not deserve a
"windfall", as the respondent contends. In the
absence of a situation of ambiguity, such that the Court must
look to the results of a transaction to assist in ascertaining
the intent of Parliament, a normative assessment of the
consequences of the application of a given provision is within
the ambit of the legislature, not the courts. ...
[53] Mr. Meghji and Mr. Chambers spent the
vast majority of their arguments illustrating how the section can
be interpreted quite differently. To use
Justice Iacobucci's expression, there is a situation of
ambiguity, and therefore the Court must look to the results of
the transaction to ascertain intent. This is not a matter of the
words themselves yielding a straightforward result (unless, as I
have hoped to make clear, one extracts the foreign exchange
aspect from the formulaic calculation altogether). So what did
Parliament intend?
[54] Firstly, unlike the other costs of
borrowing found in subsection 20(1), foreign exchange losses are
not specifically identified as a capital item to be treated as a
current expense. That result is only achieved by adroit
interpretative footwork - it is not clear on its face. This is no
minor item; indeed, it can be seen from this one case alone that
the potential for deducting foreign exchange losses on current
account is extraordinary. It is too much of a stretch to find
Parliament intended to effectively bury such a significant
deduction in paragraph 20(1)(f) with no commentary,
no headings, no guidance. It is not plausible.
[55] Consider the scheme of the Act.
There is a provision dealing with foreign exchange gains and
losses, subsection 39(2). It is clear. For foreign exchange
losses to escape such a provision must require equally explicit
language. Changing the treatment of a capital item, specifically
addressed as such in the Act, to provide current expense
treatment, should not rest on implication.
[56] While I accept Mr. Meghji's
argument that headings are not part of the Act, they are
inserted to offer some guidance to the reader. Paragraph
20(1)(f) is headed "Discount on certain
obligations". Consider this in conjunction with the only
technical notes to which I was referred, published by the
Government in 1988 as part of its comments on amendments to
subparagraph 20(1)(f)(ii), wherein the Government
stated:
Paragraph 20(1)(f) sets out rules concerning the
deductibility of discounts paid by a taxpayer in satisfaction of
the principal amount of any bond, debenture, bill, note,
mortgage, hypothec by the taxpayer on which interest was
stipulated to be payable.[25]
Though limited, these factors certainly do not suggest any
Government intention to capture foreign exchange losses under
paragraph 20(1)(f).
[57] Another distinction between the costs
of borrowing covered by the surrounding sections of subsection
20(1) and the foreign exchange loss is that the other costs
derive from the contract of origin, and are known to be costs at
the time of the contract. The foreign exchange "cost"
does not derive from the obligation itself - the obligation was
in US dollars. The "cost" arises from the external
forces of currency fluctuations. Further, at the time of the
issuance of the obligation, Imperial Oil does not know whether it
will have a cost or a gain. This is not a type of cost of
borrowing that can be categorized with other
subsection 20(1) costs. I disagree with Mr. Meghji that
context supports a finding that foreign exchange losses are just
costs like other subsection 20(1) costs. They are not.
[58] I have not been convinced that
paragraph 20(1)(f) was ever intended to allow deductions
of the type of foreign exchange loss sought by Imperial Oil. It
was however clearly intended for discounts. It is unnecessary for
me to determine what other "cost" may be the subject of
a 20(1)(f) deduction.
[59] The approach I have taken does result
in the foreign exchange having some impact on the final figure.
This explains the difference between the Respondent's result
of $1,229,181 and my result of $1,548,325. The additional
approximate $300,000 deduction available under paragraph
20(1)(f) is the part of the foreign exchange loss that
pertains specifically to the deductible discounts; that is, the
part of the foreign exchange loss that is logically and sensibly
connected to the income, as opposed to capital, item - the
discount.
Conclusion
[60] By relying on an expansive view of
Gaynor, both the Appellant and the Respondent's
approach to interpreting paragraph 20(1)(f) of the
Act necessarily incorporates a foreign exchange element.
Their disagreement over the interpretation of paragraph
20(1)(f) is based on this intermingling of exchange rates.
This results in an ambiguity which leaves no ability to rely
solely on an "ordinary and grammatical meaning of the
words" approach; it requires that I discern Parliament's
intention. The Parties became embroiled in the meaning of
"principal amount of the obligation", not in terms of
the ordinary meaning of the words themselves, but in terms only
of the exchange rate to be applied to the expression. This narrow
interpretative approach led to a result contrary to the purpose
of the provision as a whole. Parliament did not intend foreign
exchange losses to be deductible under paragraph 20(1)(f)
as they pertained to the capital element of a borrowing.
[61] The expense to be determined under
paragraph 20(1)(f) is an expense determined by a
mathematical formula - a formula intended to effectively allow a
current expense of a capital item. That is all. Applying the
formula in the currency of the obligation and converting the
result into current Canadian dollars, not only makes the most
sense in the ordinary and grammatical sense, but also does not
offend any well-established principle of statutory
interpretation. Such an approach is in accordance with the object
and scheme of the Act and the intention of Parliament. In
the alternative, calculating all elements of the
paragraph 20(1)(f) formula in 1999 Canadian dollars
achieves the same result.
[62] The answers to the questions in this
Reference are:
(a) The amount of C$1,548,325 is
deductible under subparagraph 20(1)(f)(i).
(b) Nothing is deductible under
subparagraph 20(1)(f)(ii).
(c) The amount of C$26,283,387 is the
capital loss under subsection 39(2) of the Act.
[63] While these answers substantially
favour the Respondent, it is based on reasoning not supported by
the Respondent. I am therefore not prepared to award any costs in
this matter.
Signed at Ottawa, Canada, this 10th day of March, 2004.
Miller J.
Email this Content