Trinova Corp. v. Commissioner

MEMORANDUM OPINION

TANNENWALD, Judge: Respondent determined the following deficiencies in petitioner's Federal income taxes and additions to tax:*96

*97 After concessions by the parties, the issue for decision is how various deductions should be allocated and apportioned between the subpart F income and non-subpart F income of petitioner's wholly owned subsidiary for the purposes of determining petitioner's net subpart F income includable in petitioner's gross income for the taxable year l986. ¹

This case was submitted fully stipulated under Rule 122. ² The stipulation of facts and accompanying exhibits are incorporated herein by this reference and found accordingly.

Petitioner, an accrual basis taxpayer, is a corporation having its principal offices in Maumee, Ohio, at the time it filed the petition herein. Petitioner changed its name to Trinova from the Thibbey-Owens-Ford Company on July 31, 1986. It timely*98 filed a consolidated Federal income tax return with certain of its subsidiaries for each of the years at issue either with the Internal Revenue Service Center, Cincinnati, Ohio, or the Internal Revenue Service office in Toledo, Ohio.

Throughout 1986, petitioner owned 100 percent of Aeroquip International, Inc. (AII), a Delaware corporation, which in turned owned 100 percent of Aeroquip, AG (AG), a Swiss corporation. During 1986, AG was a "controlled foreign corporation" (CFC) for purposes of section 957(a), and AII was the "United States Shareholder" of AG for purposes of section 951(b). Under subpart F of the Code, the net subpart F income derived by AG during the taxable year was currently includable in petitioner's gross income. Sec. 951 (a).

In the ordinary course of AG's business, AG derived income, incurred expenses, acquired assets, and recorded liabilities denominated in currencies other than the U.S. dollar, which was its functional currency for Federal income tax purposes. For 1986, AG's assets included loans to affiliated companies denominated in British pounds sterling ( £ ), German Deutsche marks (DM), and Italian lira. Foreign currency denominated assets and liabilities*99 resulted in AG's realization of foreign currency gains and losses in the ordinary course of its business, due to fluctuations in the value of foreign currencies with respect to the dollar, which gains and losses cannot be ascertained in advance.

For 1986, AG's assets consisted of the following:

The above asset values for all assets reflect averaging the beginning of the year and the end of the year values stated on the balance sheets of AG's financial statements.

During 1986, some of AG's loan assets were repaid in part, as follows:

*100 For 1986, AG had the following items of gross income:

By stipulation of the parties, $ 7,856,230 constituted gross subpart F income, and the exchange gain of $ 12,371,700 did not constitute gross subpart F income for purposes of section 952(a).

The parties disagree as to the allocation and apportionment of the following expenses incurred by AG:

The capital tax was imposed under Swiss law and was calculated by AG based on the amount of AG's assets less liabilities.

The swap losses consist of the following items:

The swap losses arose out of currency and interest rate exchange agreements entered into by AG with third parties. Under these agreements, AG agreed to pay interest in one currency at a certain rate*101 on a specific principal amount in that currency. In return, AG would receive interest payments in another currency based on a certain interest rate on a specific principal amount in that currency.

The total currency exchange loss consisted of the following items:

Both petitioner and respondent allocated the deductions for interest expense, Swiss capital tax, and swap losses using the asset method set forth in section 1.861-8(e)(2)(v), Income Tax Regs. Petitioner allocated these three deductions solely against subpart F income. Petitioner treated the items constituting exchange losses individually, allocating exchange losses attributable to accrual of payments under the swap agreements and to accrual of interest payments solely against subpart F income, and allocating "accrual of income tax" and other" items ratably across all gross income.

Respondent allocated and apportioned the deductions for interest expense, swap losses, and Swiss capital tax*102 using the asset method to both subpart F and non-subpart F income, based on the value of AG's assets, divided between the two categories based on the gross income in each category produced by the loan assets. Respondent allocated and apportioned all exchange losses ratably across all gross income.

There is no dispute as to whether any of the expenses at issue may be deducted, but rather merely as to how they should be allocated and apportioned.

The statutory authority governing deductions from subpart F income is section 954(b)(5), which provides that gross subpart F income "shall be reduced, under regulations prescribed by the Secretary, so as to take into account deductions (including taxes) properly allocable to such income." The regulations under section 954 offer little guidance, other than the admonition that "no expense, tax, or other deduction shall be allocated to an item or category of income to which it clearly does not apply." Sec. 1.954-1(c), Income Tax Regs.

The "regulations prescribed by the Secretary" in force for the taxable year at issue are found in section 1.861-8, Income Tax Regs., governing the allocation of deductions*103 between income from sources within and without the United States. Section 1.861-8(f)(1)(v), Income Tax Regs., specifies that "This section provides rules for identifying which deductions are properly allocable to foreign base company income" under the authority of section 954(b)(5).

A taxpayer first allocates deductions to the appropriate class of gross income and then apportions those deductions within the class of gross income (defined as "the gross income to which a specific deduction is definitely related") between the statutory grouping (in this case, subpart F income) and the residual grouping (in this case, non-subpart F income) Sec. 1.861-8(a)(2), (3), and (4), Income Tax Regs. If a deduction is not related to a specific class of gross income, it will be treated as relating to all gross income and will be allocated ratably by gross income. Sec. 1.861-8(b)(1), (5), Income Tax Regs. Once a particular deduction has been allocated to a class of gross income, it must be apportioned within that class between the "statutory" and "residual" groupings of income. The regulations stress that allocation and apportionment reflect the "factual relationship" between the deductions and the*104 gross income. Sec. 1.861-8(a)(2), (b)(1), and (c)(1), Income Tax Regs.

In addition, there are special rules for interest deductions. Because money is fungible, and a taxpayer has a great deal of flexibility in the use of borrowed funds, interest expense is normally allocated "to all the gross income which the income producing activities and properties of the taxpayer generate, have generated, or could reasonably have been expected to generate." Sec. 1.861-8(e)(2)(ii), Income Tax Regs.³ Interest deductions are apportioned on the basis of values of assets which produce income in such groupings, according to the "asset method" set out in section 1.861-8(e)(2)(v), Income Tax Regs.

The use of the asset method requires the taxpayer to place its assets in either the statutory*105 or residual grouping, according to the type of income the assets produce. The asset values are determined by an average of the asset values at the beginning and the end of the year. Sec. 1.861-8(e)(2)(v), Income Tax Regs. On the basis of the ratio of asset values, the taxpayer apportions the deductions to those two groupings. Id., see sec. 1.861-8(g), Examples (1) and (2), Income Tax Regs.⁴

*106 Thus, the regulations effectively combine the allocation and apportionment steps for interest expense.

Finally, there are special rules for losses incurred in the sale, exchange, or disposition of property. Normally, the deduction for such a loss is allocated "to the class of gross income to which such asset or property ordinarily gives rise in the hands of the taxpayer." Sec. 1.861-8(e)(7)(i), Income Tax Regs. In "the unusual circumstances" where apportionment is necessary, it is done "in the same proportion that the amount of gross income within such statutory grouping * * * and such residual grouping bear, respectively, to the total amount of gross income within the class of gross income." Sec. 1.861-8 (e)(7)(ii), Income Tax Regs.

To summarize, deductions are first allocated then apportioned based on a factual relationship between the income and the deduction. If the deduction is one for interest, it is allocated and apportioned among the groupings of income according to asset values of the income producing activities and properties. If the deduction is one for loss on the disposition of property, it is allocated based on the income ordinarily produced by that property, and further*107 apportioned, if necessary, according to gross income. Finally, if the deduction relates to no specific item of gross income, it is allocated ratably across all income.

Following the instruction of section 1.861-8(e)(2)(ii), Income Tax Regs., the parties applied the asset method to apportion the deduction between the statutory and residual groupings according to asset values. ⁵ Petitioner apportioned all of AG's assets to the statutory grouping, because they all normally produce subpart F income. It apportioned none of AG's loan assets to the residual grouping, because, although they sometimes produce non-subpart F income (in this case, foreign exchange gain), it is impossible to predict whether there will be gain or loss, based on the nature of foreign exchange rates. According to petitioner, the loan assets do not normally produce such gain. petitioner thus allocated the entire interest deduction, in the amount of $ 1,498,562, to gross subpart F income.

*108 Respondent disagrees with petitioner's application of the asset method. While she agrees that the non-loan assets should all be apportioned to the subpart F grouping, she reasons that the loan assets should be apportioned in some manner between the statutory and residual groupings, because they actually produced both subpart F and non-subpart F income. Respondent applies the asset method by: (1) apportioning the non-loan assets all to the subpart F grouping; (2) then adding together (a) all the assets and the portions of loan assets that produced subpart F income according to the income they produced, and (b) all the assets and portions of loan assets that produced non-subpart F income according to the income they produced; (3) using these subtotals to calculate a percentage of the total asset values in each grouping; (4) and then apportioning the interest deduction between the two groupings, based on these percentages.

The divergent approaches of the parties to the asset method are explained by the fact that section 1.861-8(e)(2)(v), Income Tax Regs., assumes that any given asset will produce income belonging to only one grouping--statutory or residual. ⁶ This assumption is reflected*109 in the examples found in section 1.861-8(g), Income Tax Regs. Because the regulations simply do not contemplate a situation where a single asset produces income in more than one grouping, we are left with the task of determining the apportionment of the loan assets "in a manner which reflects to a reasonably close extent the factual relationship between the deduction and the grouping of gross income." Sec. 1.861-8(c)(1), Income Tax Regs.

In making our determination, we note that in Occidental Petroleum Corp. v. Commissioner,55 T.C. 115, 124 (1970),*110 the regulations at issue there, section 1.613-4(a), Income Tax Regs., provided that expenses not directly attributable to specific property were to be "fairly apportioned", and we held for the party whose approach produced a "fairer apportionment". We see little difference between an approach that is "fairer" and one that more closely reflects the "factual relationship" between the income and the expenses. For the reasons hereinafter set forth, we find that respondent's approach to the asset method better reflects the factual relationship involved in this case.

Preliminarily, we reject petitioner's argument that the predictability of income is the critical element in the allocation of deductions against it. Even though petitioner cannot always predict AG's foreign exchange income from year to year, AG nonetheless had exchange income in 1986. This is no different from the case of many businesses that do not know, based on the vagaries of the market, whether they will have losses or gains in a given year. In the case of those businesses, they allocate deductions against their income, and if they have no income, they carry over the deductions.

Interest expense is allocated under the *111 regulations to all gross income that the property or activity "could reasonably have been expected to generate." Sec. 1.861-8(e)(2)(ii), Income Tax Regs. Even if not predictable with precision, exchange gains were certainly a reasonable possibility. Furthermore, the regulations specifically contemplate a situation where there will be a deduction even with no corresponding income. ⁷

Next, although exchange gain has traditionally been treated separately from the underlying income from the transaction, Philip Morris Inc. v. Commissioner,104 T.C. 61, 66 (1995), affd. 71 F.3d 1040 (2d Cir. 1995), we do not see how such separate treatment undermines the validity of respondent's approach. In prorating the assets across two groupings of income, respondent is not challenging the separateness of the income, but is merely attempting a reasonable allocation of interest expense to that income.

Furthermore, while the regulations are silent on this particular*112 point, they do provide a foundation for respondent's approach. When a single item of deduction is attributable to two different sources, it must be prorated between them. See supra note 4. When a piece of property produces income attributable to two different sources, the loss on the sale of that property can be prorated between them. See sec. 1.861-8(e)(7)(ii), Income Tax Regs., supra p. 11.

Finally, this is not a case where respondent's determination conflicts with the regulations, see, e.g., Woods Investment Co. v. Commissioner,85 T.C. 274 (1985), nor is it a case where the regulations are found to conflict with the statute, see, e.g., Jackson Family Foundation v. Commissioner,97 T.C. 534 (1991), affd. 15 F.3d 917 (9th Cir. 1994). This is a case where the Secretary's regulations do not adequately deal with a problem. In such a situation, it is our task to construct the best solution we can. Cf. First Chicago Corp. v. Commissioner,88 T.C. 663, 676 (1987), affd. 842 F.2d 180 (7th Cir. 1988).

In sum, respondent takes into account*113 the fact that the loan assets produced income in both groupings, whereas petitioner does not. Such being the case, we approve of respondent's method of allocating and apportioning interest expenses because it better fits the facts and constitutes the more reasonable interpretation of the statute and the regulations Cf. Occidental Petroleum Corp. v. Commissioner, supra. At least, petitioner, who has the burden of proof, Rule 142(a), ⁸ has not convinced us that its allocation method is more reasonable than that of respondent.

We are of the opinion, however, that respondent's approach must be modified in one respect. That approach is based on the notion that assets which produce income in two groupings must be prorated between the two groupings. If an asset produces income in only one grouping, it may only be apportioned to that grouping. Thus, the U.S. dollar loan asset, which could not have produced any exchange gain, can only be apportioned to subpart*114 F income, as respondent concedes on brief. In addition, some of the foreign currency loans were not repaid to any extent in 1986, could thus yield no exchange gain, and may therefore only be counted toward subpart F income since they only produced interest income. Other loans were repaid only in part, and only the amount of repayment yielded exchange gain. These loan assets should only be apportioned between the two groupings to the extent of their repayment. The portion of the loan not repaid should only be apportioned to the subpart F grouping. We agree with petitioner that the proper formula to accomplish this is to prorate the average outstanding principal balance of the foreign currency denominated loans repaid during 1986 between the statutory and residual groupings, and to apportion the average outstanding principal balance of the foreign currency denominated loans not repaid during 1986, as well as the average outstanding principal balance of the U.S. dollar denominated loans during 1986, solely to the subpart F grouping. ⁹

Both parties have allocated and*115 apportioned the deductions for swap losses using the asset method. Given the fact that the regulations provide for the use of the asset method only in respect of interest expense, section 1.861-8(e)(2)(v), Income Tax Regs., we interpret the use of the asset method by both parties as an agreement to treat swap losses in the same manner as interest expense. In this connection, we note that, based on the record herein, the swap losses involved streams of interest payments, so that the treatment of the losses as interest expense is appropriate. We therefore hold that the deduction for swap losses will be allocated and apportioned using respondent's approach to the asset method, discussed and upheld above.

Similarly, both petitioner and respondent have allocated and apportioned the deduction for the Swiss capital tax using the asset method. ¹⁰ Again, we interpret this use of the asset method as an agreement to treat this deduction in the same fashion as interest expense. We have some doubt whether the Swiss capital tax would generally qualify as interest for purposes of section 1.861-8(e)(2), Income Tax Regs. However, the factual foundation for determination may *116 be a critical element in resolving such issue. Such being the case, we accept the agreement of the parties for purposes of the case. Accordingly, this item should be allocated in the same manner as we have concluded is proper for interest expense.

The parties treat the deductions for exchange loss differently. Respondent would put the deductions in a catch-all category, and would apportion these losses all at once, ratably across all income. Petitioner would deal with each item individually, based on the nature of the underlying expense. We think petitioner has the better approach. The most accurate way to treat the deductions in question, in keeping with the mandate that allocations be made "on the basis of the factual relationship of deductions to gross income", is to deal with each item individually. *117 Sec. 1.861-8(a)(2), Income Tax Regs. Because petitioner agrees with respondent that the "accrual of income tax" and "other" items are properly allocated against all income, we need only discuss the proper treatment of the "accrual of interest payments" and "accrual of swap payments" items.

Respondent argues for the application of section 1.861-8(e)(7), Income Tax Regs., which provides that loss from the sale, exchange, or disposition of property be allocated to "the class of gross income to which such asset or property ordinarily gives rise in the hands of the taxpayer." Respondent further argues that the facts of this case represent "unusual circumstances" and that apportionment by gross income is necessary. Sec. 1.861-8(e)(7)(ii), Income Tax Regs. Respondent contends that foreign currency is "property" for purposes of the Code, and that the losses at issue resulted from the exchange of this property.

We are not convinced. The cases cited by respondent do not support the application of section 1.861-8(e)(7), Income Tax Regs., to the facts of this case. It is true that this Court has held that foreign currency is property. Federal National Mortgage Association v. Commissioner,100 T.C. 541, 582 (1993);*118 National-Standard Co. v. Commissioner,80 T.C. 551 (1983), affd. 749 F.2d 369 (6th Cir. 1984). Nevertheless, this Court has also held that a foreign exchange transaction does not involve an exchange of property. National-Standard Co. v. Commissioner, supra.Furthermore, the many cases interpreting foreign currency as property have, for the mose part, involved the issue characterization of gain as ordinary or capital, and consequently are of little help in the present case.

Respondent relies heavily on Black & Decker Corp. v. Commissioner,986 F.2d 60 (4th Cir. 1993), affg. T.C. Memo. 1991-557, which we find distinguishable on its facts. In Black & Decker Corp., the taxpayer liquidated the stock of a Japanese subsidiary at a loss, and argued that the loss should be apportioned under section 1.861-8(e)(7), Income Tax Regs., to its worldwide income, because the taxpayer's purpose in setting up the subsidiary was to augment its worldwide economic position. The Tax Court and the Court of Appeals for the Fourth Circuit both rejected the*119 taxpayer's argument, and held that the loss had to be apportioned to foreign source income, because the stock of a wholly owned subsidiary, viewed objectively, "ordinarily gives rise" to dividend income, which in that case would have been foreign source.

In the instant case, the foreign currency, while qualifying as property, would not ordinarily produce a type of income itself, independent of the assets that were denominated in that currency. We are not prepared to say that foreign currency, which produced the losses involved here, "ordinarily gives rise" to a gain from the sale, exchange or disposition of property, for purposes of section 1.861-8(e)(7), Income Tax Regs., within the meaning of National-Standard Co. v. Commissioner, supra, and Black & Decker Corp. v. Commissioner, supra.

We move then to the proper treatment of the items in question under the general rules for allocation and apportionment. First, exchange losses attributable to the accrual of interest payments should be treated in the same fashion as the underlying interest expense. Interest income cannot properly be measured without taking into account *120 interest expense. Interest expense, in turn, cannot properly be measured without the costs related to measuring it. In this case, the related accrual of exchange losses was a cost related to measuring the underlying expense. That item should be apportioned in accordance with our disposition of interest expense. See supra pp. 12-18.

For the same reasons, exchange losses attributable to the accrual of swap payments must be treated in the same fashion as the underlying swap payments. The gain or loss relating to the currency swap agreements cannot accurately be measured if one does not take into account the cost attributable to the payments. Exchange losses were an intrinsic cost of these transactions, and must be taken into account. We have already held that the swap losses are to be treated as interest expense under section 1.861-8(e)(2), Income Tax Regs. Consequently, the exchange losses related to swap losses should also be apportioned as interest expense in accordance with our treatment of that item. See supra pp. 12-18.

To take into account the concessions of the parties and our disposition of the investment tax credit issue, see supra note 1,

Decision will be entered*121 under Rule 155.