Eli Lilly & Company and Subsidiaries, Cross-Appellees v. Commissioner of Internal Revenue, Cross-Appellant

856 F.2d 855

• Court: Court of Appeals for the Seventh Circuit
• Decided: October 27, 1988
• Docket: 86-2911, 86-3116
• Status: Published

Opinion

CUDAHY, Circuit Judge.

In 1976, the Commissioner of Internal Revenue (“the Commissioner”) served a notice of deficiency on Eli Lilly and Company (“Lilly”) assessing tax deficiencies against Lilly for the years 1971 through 1973. The bulk of the disputed total amount, roughly $34 million under the Commissioner’s revised calculations, derives from the Commissioner’s reallocation of income to Lilly from Lilly’s subsidiary Eli Lilly and Company, Inc. (“Lilly P.R.”). Lilly P.R. manufactured prescription drugs for Lilly in Puerto Rico taking advantage of tax incentives provided by federal and Puerto Rican law. Lilly purchased these drugs from its subsidiary and resold them to American wholesalers. The Commissioner argued that Lilly’s transfer of patents and manufacturing know-how to Lilly P.R. in exchange for stock in the subsidiary did not entitle Lilly to allocate the income derived from those assets to Lilly P.R.

Lilly filed a petition with the Tax Court seeking redetermination of the alleged deficiencies. The Tax Court held that Lilly’s transfer of intangible assets for stock was valid in principle and that the income from these assets could properly accrue to Lilly P.R. The Tax Court also held, however, that Lilly had transferred the assets5 to Lilly P.R. for less than arm’s length consideration and shifted income to Lilly P.R. through various distortions in the pricing of Lilly P.R.’s output. Making its own income allocations under section 482 of the Internal Revenue Code, the Tax Court increased Lilly’s share of the earnings on the drugs manufactured by Lilly P.R. (Darvon and Darvon-N products) by adjusting the prices that Lilly P.R. charged to Lilly. The Tax Court’s decision effectively upheld roughly half of the Commissioner’s deficiency determination.

Lilly appeals from the Tax Court’s adjustment of the transfer prices; the Commissioner cross-appeals from the Tax Court’s determination that returns on the patents and manufacturing know-how are allocable to Lilly P.R. We affirm in part, reverse in part and remand.

I. Background

The Tax Court devoted the first 112 pages of its 196-page opinion to a summary of the facts in this case. See Eli Lilly & Co. v. Commissioner, 84 T.C. 996 (1985). Neither party alleges error in this portion of the Tax Court opinion. We confine our summary to the bare essentials and refer readers seeking greater detail to the Tax Court’s able efforts.

Lilly obtained a patent for propoxyphene hydrochloride, the active ingredient of Darvon, in December 1955, and began marketing the product in 1957. In late 1962, the company obtained a patent on a related substance, propoxyphene napsylate, which it sold in the United States as the key ingredient of Darvon-N products beginning in 1971. Because these products filled a need for non-addictive pain relievers acting on the central nervous system, they proved extremely profitable to Lilly, generating $30.2 million in pre-tax net income in 1965.

In 1965 Lilly created Lilly P.R. to manufacture Darvon and Darvon-N products (hereinafter referred to simply as “Darvon products”) in Puerto Rico. The Tax Court found that this move was motivated by a variety of factors, important among which were tax incentives provided under Puerto Rican law and section 931 of the Internal Revenue Code, 1 and Lilly’s desire to expand and diversify its manufacturing base. 84 T.C. at 1018-19. In late 1966, Lilly transferred ownership of its two propoxyphene patents along with proprietary information on manufacturing processes to Lilly P.R. in exchange for stock in the subsidiary. Proceeding under the terms of a private letter ruling from the Internal Revenue Service, Lilly claimed nonrecognition treatment for the transfer under section 351. 2 From the time it commenced manufacturing operations, Lilly P.R. was the only manufacturer of propoxyphene products in the United States or Puerto Rico.

During the tax years at issue, 1971 through 1973, Lilly P.R. manufactured finished Darvon products using raw ingredients purchased mainly from outside sources. Lilly provided certain specialized services, including engineering, quality control and research and development aimed at developing new propoxyphene products, for which Lilly P.R. paid set fees. Lilly P.R. sold its entire output of Darvon products to Lilly at prices calculated to achieve a proper division of earnings (net of costs) between parent and subsidiary. Lilly’s approach to this problem begins, in essence, with four major categories of Darvon earnings: Lilly’s return on its marketing activities, Lilly P.R.’s return on its manufacturing activities (including an adjustment for cost savings achieved by operating in Puer-to Rico), Lilly’s return on marketing intangibles (such as its trademark and good will) and Lilly P.R.’s return on manufacturing intangibles. Lilly first identified “normal” rates of return on marketing and manufacturing activities and computed appropriate returns for Lilly and Lilly P.R. based on these rates; remaining Darvon earnings were then allocated between manufacturing and marketing intangibles. To determine the proper division for the years 1971 through 1973, Lilly relied on a formula that Lilly and the IRS had used to settle a dispute over Lilly’s taxes for 1968. 3 Under the formula, Lilly P.R.’s return with re spect to its manufacturing activities was set at 100% of its manufacturing costs (cost of goods sold), minus its operating expenses (research and development specific to Darvon products, general administration and sample expenses), plus its location savings (cost savings of operating in Puerto Rico). Lilly’s normal return with respect to marketing activities was set at 25% of Darvon-related distribution expenses. Remaining earnings were treated as returns on intangibles. For 1971 and 1972, Lilly P.R. claimed 60% of the returns to intangibles and Lilly 40%; in 1973, Lilly P.R.’s share of returns to intangible assets was reduced to 30% to reflect the drop in the value of the manufacturing intangibles caused by the expiration of the Darvon patent.

The Notice of Deficiency contended that the Commissioner could ignore Lilly P.R.’s nominal ownership of the manufacturing intangibles for purposes of allocating income under section 482 of the Internal Revenue Code. The Tax Court rejected the view that Lilly P.R.’s ownership could be disregarded in its entirety, but held that the transfer of the manufacturing intangibles was not an arm’s length transaction and that section 482 could be used to achieve a division of Darvon earnings more nearly approximating the division that an arm’s length transaction would have produced. Specifically, the Tax Court objected to the transfer of the manufacturing intangibles on the grounds that in an arm’s length exchange Lilly would have insisted on some form of lump sum or periodic cash payments to help meet the costs of its general research and development activities. In addition, the Tax Court reviewed the Darvon transfer prices themselves and found independent grounds for invoking section 482. The Tax Court corrected the perceived distortion of Lilly’s income by reducing Lilly P.R.’s prices to Lilly, thereby increasing Lilly’s share of the Darvon profits.

Both Lilly’s appeal and the Commissioner’s cross-appeal dispute the Tax Court’s application of section 482 to the Darvon profits. The Commissioner contends that the Tax Court adopted an overly restrictive view of the Commissioner’s authority under section 482 by holding that the transfer of intangible rights to an offshore subsidiary in exchange for stock must be accepted, in principle, as a valid arm’s length transaction. Lilly contends that the Tax Court arrogated too much authority to itself by reallocating earnings between parent and subsidiary based on an unsupported rejection of the taxpayer’s own efforts to establish arm’s length transfer prices for the Darvon products. We consider the Commissioner’s challenge first.

II. The Applicability of Section 482

Section 482 confers broad authority on the Commissioner to allocate income or deductions among commonly controlled entities where “such ... allocation is necessary in order to prevent evasion of taxes or clearly to reflect the income of any of such organizations, trades, or businesses.” 4 Citing section 1.482-l(c) of the regulations applicable during the period in question, the Commissioner claims a general mandate to adjust incomes wherever the income of a controlled taxpayer “either by inadvertence or design ... is other than it would have been had the taxpayer in the conduct of his affairs been an uncontrolled taxpayer dealing at arm’s length with another uncontrolled taxpayer.” Treas.Reg. § 1.482-1(c) (1986). More specifically, the Commissioner relies on the regulations’ assertion of authority to reallocate income between related parties even where such reallocations effectively nullify nonrecognition transactions authorized by other sections of the Internal Revenue Code, such as section 351. See 26 Treas.Reg. § 1.482-1(d)(5) (1986). 5 Examining Lilly’s transfer of patents and manufacturing know-how to Lilly P.R. under these authorities, the Commissioner concluded that an arm’s length deal between Lilly and Lilly P.R. would have limited Lilly P.R. to a reasonable return on its manufacturing activity. In the Commissioner’s view, if Lilly had been negotiating with Lilly P.R. at arm’s length, it would not have exchanged the intangible assets necessary to the manufacture of Darvon for stock, even stock that provided a controlling interest in a highly profitable enterprise.

The Commissioner exercises broad authority under section 482. The courts will overrule a section 482 reallocation by the Commissioner only if it is found to be arbitrary, capricious or unreasonable. Foster v. Commissioner, 756 F.2d 1430, 1432 (9th Cir.1985), cert. denied, 474 U.S. 1055, 106 S.Ct. 793, 88 L.Ed.2d 770 (1986); Powers v. Commissioner, 724 F.2d 64, 66 (7th Cir.1983); Baldwin-Lima-Hamilton Corp. v. Commissioner, 435 F.2d 182, 185 (7th Cir.1970). When a taxpayer rebuts the presumption of reasonableness afforded the Commissioner’s determinations, see, e.g., United States Steel Corp. v. Commissioner, 617 F.2d 942, 947 (2d Cir.1980); Woodward Governor Co. v. Commissioner, 55 T.C. 56, 65-66 (1970), the Tax Court (or district court) must determine whether the taxpayers own allocations conform to the arm’s length requirement. This court, for example, has remanded a section 482 case to the district court to determine a “partial re-allocation” where both the Commissioner’s reallocation of a subsidiary’s entire profit and the taxpayer’s opposition to any reallocation were found to be erroneous. Baldwin-Lima-Hamilton Corp., 435 F.2d at 186-87. The Tax Court has employed this power to reallocate in several cases in which a taxpayer has rebutted the Commissioner’s reallocations without providing alternative allocations that satisfy the arm’s length standard, establishing its own allocations to ensure that income, deductions and credits accurately reflect income. See American Terrazzo Strip Co. v. Commissioner, 56 T.C. 961, 973 (1971); Nat Harrison Assocs., Inc. v. Commissioner, 42 T.C. 601, 617-18 (1964); Ach v. Commissioner, 42 T.C. 114, 126-27 (1964), aff’d, 358 F.2d 342 (6th Cir.), cert. denied, 385 U.S. 899, 87 S.Ct. 207, 17 L.Ed.2d 131 (1966).

We think that the Tax Court’s approach in these cases, an approach consistent with this court’s remand in Baldwin-Lima-Hamilton, is a sensible one. Both the Commissioner and taxpayer have the opportunity to present evidence in support of their allocations. Where the evidence shows that neither side is correct, we think it would be unreasonable to restrict the court to acceptance or rejection of the Commissioner’s position in its entirety, rather than allowing the court to reallocate “in a manner the evidence ... demonstrates to be correct.” Nat Harrison Assocs., 42 T.C. at 618.

We are unaware of any decision discussing the standard that governs appellate review of a Tax Court’s reallocation. One might argue that appellate review of the Tax Court’s allocations should parallel the Tax Court’s review of the Commissioner’s allocation — that the allocation should be upheld unless a de novo review of the facts shows it to be arbitrary, capricious or unreasonable. However, we lack the fact-finding capabilities that are essential to resolving factual discrepancies under that standard — especially conflicts involving disparate expert testimony. Our usual standards of review must therefore apply. The Tax Court’s legal conclusions will receive plenary, review. Factual findings will be reversed only if “clearly erroneous.” Anderson v. City of Bessemer City, 470 U.S. 564, 573, 105 S.Ct. 1504, 1511, 84 L.Ed.2d 518 (1985). “Mixed questions of law and fact,” entailing the application of a legal standard to a given factual pattern, are reviewed under the clearly erroneous standard. Standard Office Bldg. Corp. v. United States, 819 F.2d 1371, 1374 (7th Cir.1987).

A. The Validity of the Intangibles Transfer in Principle

The parties’ dispute over the transfer of the manufacturing intangibles implicates the conflict between section 482 and the Code’s nonrecognition provisions. The issue is one of priority: To what extent is the Commissioner’s authority under section 482 constrained by the Code’s nonrecognition provisions, and, conversely, to what extent are taxpayers’ efforts to invoke nonrecognition treatment constrained by the proviso that nonrecognition treatment may be nullified if it results in the evasion of taxes or the distortion of income? In view of the open-ended language of section 482, and the inherent problems with either affording the Commissioner plenary power to nullify statutory provisions on the one hand, or allowing taxpayers to take full advantage of tax schemes permitted by literal terms of the Code on the other, it is unsurprising that the answer provided by the case law is that the applicability of section 482 depends on the circumstances of the nonrecognition transactions in question.

One view of the matter is that the Commissioner’s power to override nonrecognition treatment depends upon the extent to which the nonrecognition transaction served business aims other than the reduction of taxes. In this view, the Commissioner exercises greater authority under the “tax evasion prong” of section 482, applicable to tax-motivated transactions, than under the “distortion of income prong,” applicable to business-motivated transactions with favorable tax consequences. In Ruddick Corp. v. United States, 643 F.2d 747, 226 Ct.Cl. 426 (1981) (“Ruddick I”), for example, the Court of Claims held that section 482 could take precedence over nonrecognition treatment for the distribution of property from subsidiary to parent under section 311 of the Code only if “a significant element of tax avoidance or evasion” had tainted the distribution. Id. at 751-52. The court assumed, in assessing the government’s motion for summary judgment, that the transaction in question — the transfer of greatly appreciated shares of stock (along with their basis) from a subsidiary to a parent that had operating losses to offset these gains — was motivated entirely by nontax considerations. It then rejected the Commissioner’s attempt to negate nonrecognition treatment, though it limited its analysis to the particular transaction and nonrecognition provision at issue, warning that “the presence of a business purpose or the absence of any plan to lessen taxes does not, in and of itself, exclude the operation of [section 482].” Id. at 750. 6

Other decisions, including the Tax Court opinion in this case, suggest that nonreeog-nition provisions can at least limit the scope of section 482 even where tax considerations formed part of the motivation for the transaction in question. See G.D. Searle & Co. v. Commissioner, 88 T.C. 252, 363-67 (1987); Eli Lilly & Co., 84 T.C. at 1119-21; Bank of America v. United States, 79-1 U.S.Tax Cas. (CCH) ¶ 9170 (N.D.Cal.1979). Both the decision below and G.D. Searle, which also involved the transfer of intangible assets to a manufacturing subsidiary operating in Puerto Rico, found that tax considerations influenced the decision to transfer intangibles. However, both decisions found that taxpayers could not be attacked under the tax avoidance prong of section 482 for having engaged in conduct that Congress sought to encourage and that the Commissioner lacked authority under the income distortion prong to negate an otherwise valid section 351 transfer in its entirety. G.D. Searle, 88 T.C. at 364-65, 367; Eli Lilly, 84 T.C. at 1120-21, 1127-29; see also Commissioner v. First Sec. Bank, 405 U.S. 394, 398 n. 4, 92 S.Ct. 1085, 1089 n. 4, 31 L.Ed.2d 318 (1972) (“Taxpayers are, of course, generally free to structure their business affairs as they consider to be in their best interests, including lawful structuring (which may include holding companies) to minimize taxes.”). Bank of America rejected the Commissioner’s attempt to disregard a nonrecognition transfer of a subsidiary’s foreign branches to its parent under section 311. The district court acknowledged that the transaction would defer payment of taxes on the increased value of the branches, but refused to allow reallocation absent any showing that the parent could forever avoid taxes on the gain. The court held that section 482 could not, in the circumstances of that case, override the “Congressional policy decision that such a transfer would not be an appropriate occasion to impose tax liability on the distributing corporation.” Id. at 86,253.

One problem for the view that the Commissioner’s section 482 authority depends on the taxpayer’s motive and on the relationship between the taxpayer’s conduct and the conduct Congress has sought to encourage is to account for a number of decisions that have upheld section 482 allocations that effectively nullified nonrecognition transfers. The Tax Court divided these cases into two distinct and narrow lines. The first line, which the Tax Court viewed primarily as tax avoidance cases in which there was “no valid business purpose for the transfer,” began with National Securities Corp. v. Commissioner, 137 F.2d 600 (3d Cir.), cert. denied, 320 U.S. 794, 64 S.Ct. 262, 88 L.Ed. 479 (1943). See 84 T.C. at 1119. In National Securities, a nonrecognition stock transfer from parent to subsidiary was followed by the subsidiary’s sale of the stock and realization of a loss that had substantially more favorable tax consequences for the subsidiary than for the parent. 7 The Tax Court noted that similar transactions received similar treatment in Northwestern National Bank v. United States, 556 F.2d 889, 892 n. 5 (8th Cir.1977) (nonrecognition transfer of appreciated property followed by donation to charity under “preconceived plan to make the contribution through the parent” for tax reasons), and Southern Bancorporation v. Commissioner, 67 T.C. 1022 (1977) (distribution of Treasury bonds as dividends to parents in effort to obtain more favorable tax treatment on gain). Cases sustaining reallocations since the Tax Court’s decision below may also be grouped under the tax evasion rubric. 8

The second line of cases, headed by Central Cuba Sugar Co. v. Commissioner, 198 F.2d 214 (2d Cir.), cert. denied, 344 U.S. 874, 73 S.Ct. 167, 97 L.Ed. 677 (1952), involved nonrecognition transfers of mature crops which effectively decoupled the expenses and revenues of producing these crops. See also Rooney v. United States, 305 F.2d 681 (9th Cir.1962). Although there was no finding in either case that the transfers were motivated by tax considerations, the courts sustained the Commissioner’s position that the transferee should be treated as having incurred the expenses of planting and cultivating the crops that produced their income in order to prevent a mismatching of expenses and income for a single taxable year. The Tax Court properly concluded that the proscription against mismatching income and expenses incurred in a single year has no bearing on Lilly’s transfer of intangibles to Lilly P.R. 84 T.C. at 1121-22, 1124; accord Ruddick I, 643 F.2d at 751; G.D. Searle, 88 T.C. at 365.

The Tax Court found Lilly’s transfer of intangibles to Lilly P.R. closer to the transfer considered in Bank of America, in which the district court rejected the Commissioner’s attempt to override nonrecognition treatment and force the subsidiary to realize a capital gain on a transfer of assets to its parent. The Tax Court noted that the transferee in Bank of America, like Lilly P.R. but unlike the transferees in the National Securities line, used the transferred assets rather than disposing of them, thus eliminating any presumption that the transfer was designed merely to obtain more favorable tax treatment of a planned disposal of the assets by the controlled group. 84 T.C. at 1123-24.

The Commissioner attempts to distinguish Bank of America on the grounds that Lilly’s transfer of intangibles was not an arm’s length transaction. 9 But the Commissioner has not identified the defining characteristics of an arm’s length transfer of intangibles or specified which of these characteristics was missing from the transfer at issue here. The Commissioner insists that the transfer of the Darvon patents and processes for stock and distribution rights (along with the prospect of modest fees for providing technical assistance) would have been “inconceivable” if Lilly P.R. had been an unrelated party. 10 But such a transaction seems perfectly conceivable if the unrelated firm had been able to offer Lilly an equity interest of comparable value, at a comparable level of risk, together with equally valuable technical assistance contracts and exclusive distribution rights. The form of the consideration that Lilly received does not strike us as decisive in this case. To be sure, Lilly P.R. occupied a preferred position due to section 351 nonrecognition treatment and section 931 tax incentives. 11 These provisions insured that the Darvon intangibles could “purchase” a more valuable interest in Lilly P.R. than they could have purchased in an uncontrolled company on the mainland. The Commissioner, by insisting that an arm’s length transferee would have been confined to the role of a contract manufacturer, implies that Lilly P.R.’s only legitimate role was the role of a firm with which Lilly might have contracted for the production of Darvon in the absence of sections 351 and 931. It is difficult to see, however, how Lilly’s claiming these advantages in any way distorts the incentives that Congress sought to establish when it enacted sections 351 and 931. Cf. Ruddick I, 643 F.2d at 751-52.

Our view is reinforced by the Service’s own published position on the tax treatment of intangibles transferred to Puerto Rican affiliates. In Revenue Procedure 63-10, 1963-1 C.B. 490, the Service specifically considered whether section 482 empowered it to disregard transfers of intangible assets between Puerto Rican manufacturers and mainland affiliates and concluded that its authority did not extend this far. This Revenue Procedure provided specific guidelines for the application of section 482 to transfers of raw materials and finished products between United States companies and their manufacturing affiliates in Puerto Rico. 12 Section 4 of this document addresses pricing issues that arise when intangible assets contribute materially to combined earnings. This section provides that a determination of ownership must precede the determination of arm’s length transfer prices and that these prices should allow the island affiliate an appropriate return on the intangible assets it owns. The Revenue Procedure does not indicate how ownership of intangibles should be established, but accompanying guidelines called for an examination of documents and other evidence to determine whether “intangibles, such as patents, trademarks, trade names, etc., originally developed or owned by the mainland affiliate ... have been transferred to the island affiliate.” 13 The Commissioner does not contend that Lilly failed to provide adequate documentation of the section 351 transfer that had been approved in the Service’s 1966 private letter ruling. Instead, the Commissioner asserts, despite the obvious implication of the Revenue Procedure and the plain language of the guidance, that Revenue Procedure 63-10 applied only to intangible assets developed by Puerto Rican affiliates. The Tax Court correctly found the Commissioner’s construction of Revenue Procedure 63-10 implausible on its face (even without reference to the guidance), since “except in unusual cases, any significant intangibles related to products manufactured by the Puerto Rican affiliate are first developed or acquired by the mainland affiliate.” 84 T.C. at 1188. 14 We think it is beyond ques tion that Revenue Procedure 63-10 took a position with respect to the transfer of intangibles to Puerto Rican affiliates fundamentally inconsistent with the Commissioner’s position in this case.

A strong argument can be made that Revenue Procedure 63-10 should be given binding effect. While the Commissioner may have policy reasons, in retrospect, for opposing transfer of intangibles to affiliates manufacturing in Puerto Rico, when the Service issued Revenue Procedure 63-10, it adopted a contrary view. The Revenue Procedure, the validity of which was reaffirmed in 1968 when the Service issued regulations governing transfer prices, represented the only clear indication of the Service’s view of the interaction between the Commissioner’s amorphous section 482 powers and the incentives provided by sections 351 and 931 applicable to Lilly’s establishment of Lilly P.R. and purchase of the Darvon manufactured by the subsidiary. Concededly, some cases have characterized particular revenue procedures as “directory not mandatory.” See Cleveland Trust Co. v. United States, 421 F.2d 475, 481-82 (6th Cir.), cert. denied, 400 U.S. 819, 91 S.Ct. 35, 27 L.Ed.2d 46 (1970); see also Geurkink v. United States, 354 F.2d 629, 632 (7th Cir.1965) (dictum). However, these cases did not purport to announce a rule applicable to all revenue procedures; and the content of Revenue Procedure 63-10, as it affects this case, seems distinguishable from the distinctly procedural revenue procedures at issue in Cleveland Trust and Geurkink. It is also true that the Commissioner possesses unusually broad power to enforce changes in the Service’s legal position retroactively. See Dixon v. United States, 381 U.S. 68, 70-71, 85 S.Ct. 1301, 1303-04, 14 L.Ed.2d 223 (1965); Manocchio v. Commissioner, 78 T.C. 989, 1000-02, aff’d, 710 F.2d 1400, 1402-03 (9th Cir.1982). But this authority is not a license for the IRS to disregard all expectations, no matter how firmly rooted in the Service’s public representations about the tax consequences of particular transactions. Cf. LeCroy Research Sys., Inc. v. Commissioner, 751 F.2d 123, 127-28 (2d Cir.1984). But cf. CWT Farms, Inc. v. Commissioner, 755 F.2d 790, 804 (11th Cir.1985), cert. denied, 477 U.S. 903, 106 S.Ct. 3271, 91 L.Ed.2d 562 (1986). More important, the Commissioner has not argued that the power of retroactive enforcement allows him to take a position in litigation that repudiates the substantive content of a narrowly tailored revenue procedure that has never been revoked. See United States v. Nixon, 418 U.S. 683, 694-96, 94 S.Ct. 3090, 3100-02, 41 L.Ed.2d 1039 (1974); Lansons, Inc. v. Commissioner, 622 F.2d 774, 776 (5th Cir.1980). If Revenue Procedure 63-10 is properly characterized as a substantive statement rather than a procedural directive, then neither a belated reassessment of the scope of the Commissioner’s power under section 482 nor belated alarm at the cost to the Treasury of recognizing transfers of intangibles to Puerto Rican manufacturing affiliates should allow the Commissioner to evade its effect here.

Ultimately, however, we rely on the legal position reflected in Revenue Procedure 63-10, rather than on any binding effect that the Revenue Procedure might have. Revenue Procedure 63-10 was plainly predicated on the view that transferring intangibles, and the associated income streams, to Puerto Rican affiliates would not necessarily result in the evasion of taxes or the distortion of income. The Commissioner’s arguments have not convinced us that this view was incorrect.

B. The Effect of Lilly’s Failure to Provide for a Steady Stream of Income from the Manufacturing Intangibles

While the Tax Court rejected the Commissioner’s view that the exchange of manufacturing intangibles for stock was entirely voidable for purposes of allocating income, it nevertheless found that Lilly’s exchange of these intangibles for stock in Lilly P.R. distorted Lilly’s income. The Tax Court found it “inconceivable that [Lilly], negotiating at arm’s length, would have transferred valuable income-producing intangibles without a royalty, lump-sum payment, or other agreement” to generate cash for Lilly’s general research and development efforts. 84 T.C. at 1130. (Lilly’s general research and development efforts, aimed at developing entirely new products, are distinct from product-specific investments in new ways of packaging and using existing products; the costs of research specific to Darvon products were paid by Lilly P.R.) Having determined that the initial transfer of the Darvon intangibles distorted Lilly’s income, the Tax Court undertook a detailed analysis of the prices that Lilly P.R. charged Lilly for Darvon products, under the regulations governing the determination of arm’s length prices.

Lilly’s appeal of these adjustments focuses primarily on the Tax Court’s authority to apply section 482. In Lilly’s view, the Tax Court rightly rejected the Commissioner’s sweeping application of section 482, but erroneously substituted its own flawed theory as to how the sale of the Darvon intangibles deviated from the requirements of an arm's length transaction and therefore distorted Lilly’s income. Lilly contends that the Tax Court’s erroneous characterization of the stock-for-intangibles exchange constituted the only intelligible basis for invoking section 482.

There is considerable force to Lilly’s criticism of the Tax Court’s characterization of the exchange of stock for manufacturing intangibles. The Tax Court’s argument that Lilly did not receive arm’s length consideration appears to be no sounder logically than the Commissioner’s argument that the transfer should be disregarded entirely in the allocation of returns on intangible assets. After rejecting the Commissioner’s argument that Lilly P.R.’s ownership of the intangibles should be disregarded because the very idea of Lilly’s transferring the Darvon intangibles for stock would have been unthinkable in an arm’s length transaction, the Tax Court found instead that research and development expenses should be reallocated because the very idea of transferring the Darvon intangibles for stock “without a royalty, lump-sum payment, or other agreement that would enable [Lilly] to continue its general research and development activities” would have been unthinkable in an arm’s length transaction. 84 T.C. at 1130; accord G.D. Searle, 88 T.C. at 370.

The Tax Court’s objection to the terms of the intangibles transfer, unlike the Commissioner’s objection to the very idea of such an exchange, can at least be reconciled with Revenue Procedure 63-10. However, the Tax Court’s view of the requisites of an arm’s length transaction rests on questionable legal and factual footings. The Tax Court’s position conflicts with holdings that the stock of a corporation is by definition equal in value to that corporation’s assets and cannot be worth less than arm’s length consideration. Nash v. United States, 398 U.S. 1, 90 S.Ct. 1530, 26 L.Ed.2d 1 (1970), rejected the Commissioner’s position that partnerships recovered bad debt reserves when they transferred the underlying debts to new corporations in exchange for stock. The Court dismissed the Commissioner’s objection&emdash;which con- cededly rested on different grounds from the arm’s length transfer argument in the case before us&emdash;reasoning that the re- serves were not recovered, but rather were reflected in the value of the stock, since of the stock, since “[a]ll that [the taxpayers] received from the corporation were securities equal in value to the net worth of the accounts transferred.” Id. at 4, 90 S.Ct. at 1551; see also Livingston v. Commissioner, 18 B.T.A. 1184, 1191 (1930) (for purposes of determining gain when real estate is transferred for stock, “fair market value of the stock was equal to the fair market value of the assets at the date of conveyance”); Zeigler v. Commissioner, 1 B.T.A. 186, 193 (1924) (same). These decisions, rendered in contexts distinguishable from the case at hand, do not entirely foreclose the Tax Court’s position merely by equating the value of stock with the value of assets transferred in exchange. The broad language of the regulation defining arm’s length consideration for transfers of intangibles, however, further undermines the Tax Court’s restrictive view of adequate compensation. It provides that “arm’s length consideration” may take any of several listed forms, or “any other form ... which might reasonably have been adopted by unrelated parties under the circumstances.” Treas.Reg. § 1.482-2(d)(2)(i) (1986).

The argument upon which the Tax Court and the Commissioner (as appellee) rely to limit earlier holdings on the equivalence of stock and asset values and to disregard the broad language of the regulations depends on an exceedingly strict and, so far as appears, entirely unsupported view of the constraints that Lilly’s research and development program places on its financial planning. The Tax Court insists that Darvon contribute a “proportionate share,” set at eighty percent (the ratio of Lilly’s sales of Darvon products to its total sales), to Lilly’s research and development expenses. 15 The Tax Court found that if Lilly had been dealing with an uncontrolled transferee, if Lilly had not been influenced by the prospect of accumulating untaxed profits in an offshore subsidiary, the company would have insisted on obtaining a stream of income to defray the costs of developing new products. Commensurate, or perhaps even greater, increases in corporate wealth, held in the form of stock in a profitable subsidiary, would be unacceptable as a substitute for cash. 84 T.C. at 1130; accord G.D. Searle, 88 T.C. at 370-71.

We assume, without deciding, that the Tax Court was correct in its view that drug companies ordinarily expect different product lines to generate pro rata contributions to company-wide research and development efforts. 16 Even granting this, however, the Tax Court’s own figures show that Lilly did not need to be concerned about Darvon products’ sustaining their assigned burden. Lilly’s gross profit (sales less marketing costs and payments to Lilly P.R.) derived from purchasing Darvon products at prices established under the 1968 formula and reselling to wholesale distributors, more than covered both the operating expenses and the research and development expenses that management, according to the Tax Court, ordinarily would have insisted on drawing from Darvon profits. 17 Lilly’s failure to secure a steady cash flow from the manufacturing intangibles themselves did not, so far as the record shows, induce the parent company to cut back on its research and development efforts; it only required Lilly to dip more deeply into its substantial profits from marketing Darvon in order to maintain the “Darvon contribution” to research and development expenses. 18 In other words, money is money, from whatever source derived; and the record does not disclose that Lilly scaled back any of its normal activities, including research and development aimed at producing new drugs, due to the loss of the income stream generated by the Darvon intangibles. The financial facts therefore undermine the Tax Court’s finding that the transfer of the Darvon intangibles for stock alone distorted Lilly’s income. The Tax Court’s subsequent examination of the Darvon transfer prices, however, revealed various independent grounds for concluding that those prices deviated from arm’s length prices and distorted Lilly’s income. We turn to these aspects of the Tax Court’s opinion next.

III. The Tax Court’s Adjustments of Darvon Transfer Prices

Lilly contends that the Tax Court’s adjustments of the Darvon transfer prices, undertaken to reallocate income to Lilly, stand or fall on the validity of its conclusion that the terms of the initial exchange of Darvon intangibles for stock distorted Lilly’s income. This is wrong, or at least only partially right. Some of the Tax Court’s reallocations of Darvon revenues were predicated on Lilly’s supposed need to derive a continuous stream of income for general research and development expenses from the Darvon manufacturing intangibles, an argument we have already rejected. These reallocations must fall. However, the Tax Court also undertook a detailed analysis of other aspects of the Darvon transfer prices and found additional reasons for concluding that they were not arm’s length prices. Adjustments flowing from these analyses are correct and we affirm them.

The Tax Court evaluated the Darvon transfer prices in accordance with the published standards for assessing whether transactions conform to arm’s length practices. In general, an arm’s length price in the sale of tangible property is the price that would prevail if the transaction involved unrelated parties. See Treas.Reg. § 1.482-2(3)(l)(i) (1986); see also Rev.Proc. 63-10 § 3.01, 1963-1 C.B. 490, 492. The Service’s specific guidelines for applying the arm’s length standard describe and rank four methods for determining arm’s length prices in sales of tangible property. Under these guidelines, taxpayers must, if possible, base prices in transactions between commonly controlled entities on “comparable uncontrolled prices,” prices in sales between unrelated parties that involve “identical” or “nearly identical” prod ucts and conditions. The second option, the “resale price method,” subtracts “an appropriate markup” from the price for which goods that the taxpayer purchases from an affiliate are resold to outside parties. The “cost plus method,” third in line, starts with the transferor’s production costs and adds an “appropriate gross profit percentage” to arrive at an arm’s length price to the transferee. Finally, in the event that none of these methods applies, arm’s length prices may be determined by “some appropriate method of pricing other than [the listed methods], or variations on such methods.” 19

The Tax Court conducted separate analy-ses of the Darvon transfer prices for the period from 1971 to 1972, the period before the primary Darvon patent lapsed, and for 1973, the year in which generic manufacturers entered the market and established comparable uncontrolled prices for propox-yphene products.

A. Reallocations for 1971 and 1972

To justify its 1971 and 1972 prices Lilly presented an “opportunity cost” analysis. Lilly’s experts sought to demonstrate that the prices Lilly paid for Darvon were arm’s length prices by showing that Lilly’s profits from reselling Darvon exceeded profits that it could have expected to earn from alternative uses of its marketing resources. These experts testified that Lilly’s Darvon marketing operations were more profitable — measuring profitability as the ratio of operating income to operating expenses — than its combined manufacturing and marketing operations for its nine next leading products.

Before the Tax Court, Lilly argued that its opportunity cost approach conformed to the essential elements of the resale price approach. But the Tax Court found that the resale price method required Lilly to identify markups used in similar uncontrolled transactions, that is, purchases of finished drugs from unrelated manufacturers for resale to independent distributors. See Treas.Reg. § 1.482-2(e)(3)(vi) (1986). The Tax Court therefore concluded that Lilly’s profits from manufacturing and marketing its other leading products did not provide an appropriate standard for assessing profits on marketing Darvon. 84 T.C. at 1143-45.

Lilly accuses the Tax Court of applying the resale price guidelines mechanically, with insufficient attention to the purposes of comparing transactions between controlled and uncontrolled entities. Lilly is specifically critical of the Tax Court’s reliance on Lufkin Foundry & Machine Co. v. Commissioner, 468 F.2d 805 (5th Cir.1972), which dealt with a clear-cut instance of a taxpayer’s failing to provide the information necessary to support its claim that transactions among affiliates took place at arm’s length prices. In Lufkin Foundry, the taxpayer made no attempt to apply the specifically enumerated methods of justifying prices, relying instead on an expert’s opinion of a reasonable division of revenues. Lilly’s opportunity cost analysis, in contrast, tracked the resale price method set forth in the guidelines insofar as Lilly’s operations allowed. Nevertheless, because there were no products that Lilly purchased from unrelated manufacturers and resold to wholesale distributors, Lilly, like the taxpayer in Lufkin Foundry, lacked “evidence of transactions between uncontrolled corporations” parallel to its Darvon dealings. Lufkin Foundry, 468 F.2d at 808. 20 Moreover, even if Lilly had been able to compare Darvon purchases and resales to uncontrolled purchases and resales of other drugs, we are not convinced that the Tax Court would have been obliged to accept that these other transactions were comparable. Propoxyphene was the first synthetic pain reliever that acted on receptors in the brain without posing a significant addiction risk; and Lilly’s Darvon trademark completely dominated the American market from its inception. Darvon was the most heavily prescribed drug in the United States during the period from 1960 through 1973; it placed in the top ten in total sales every year during that period. These factors suggest that the Darvon trademark and other product-specific marketing intangibles may have been unique among Lilly products. 21 In short, we cannot say that the Tax Court erred in rejecting Lilly’s claim that the ratio of operating income to operating expenses (excluding the cost of goods sold) for the manufacture and sale of Lilly’s next nine leading products provided an appropriate yardstick for assessing the ratio of operating income to operating expense for the purchase and resale of Darvon under the resale price method. 22

On appeal, Lilly maintains that even if its opportunity cost approach need not be accepted as a variant on the resale price method, it must be accepted as an “appropriate method of pricing” other than those specifically enumerated, under the fourth option provided by the guidelines. See Treas.Reg. § 1.482 — 2(e)(l)(iii) (1986). But the problems that led the Tax Court to reject Lilly’s approach as a valid application of the resale method also undermine its validity as an alternative “appropriate method.” The comparison of Lilly’s rate of return on the marketing of Darvon to its rates of return on manufacturing and marketing of its other products is unreliable as an alternative methodology for the same reasons that it is unreliable as a variant of the resale method. Again, the Tax Court did not err in rejecting Lilly’s method. 23

Having determined that none of the specifically enumerated approaches to determining arm’s length prices applied and having rejected Lilly’s opportunity cost approach as an alternative appropriate method, the Tax Court elected to apply a “profit split” methodology. The profit split approach divides combined revenues based on an ad hoc assessment of the contributions of the assets and activities of the commonly controlled enterprises. See, e.g., PPG Indus., Inc. v. Commissioner, 55 T.C. 928, 997 (1970). The 1968 settlement formula under which Lilly determined its prices (as distinct from the opportunity cost approach under which Lilly later sought to justify those prices) employed the profit split approach. Under this approach, Lilly and Lilly P.R. received “normal” rates of return on manufacturing and marketing expenses. Remaining profits, viewed as returns on intangibles, were divided according to an estimate of the relative contributions of marketing and manufacturing intangibles to the remaining Darvon profits. Lilly urged the Tax Court, in the event that it rejected the opportunity cost justification, to affirm the profit split embodied in the 1968 formula.

The Tax Court declined to adopt the 1968 formula in its entirety for 1971 and 1972 or to accept Lilly’s assignments of costs on faith. Based on its thorough review of the facts, the Tax Court shifted some costs from Lilly to Lilly P.R., thereby reallocating some of the Darvon profits from Lilly P.R. to Lilly. The Tax Court increased Lilly P.R.’s costs for line items including issuance of production tickets (detailed instructions for production runs), management of finished product inventories and general administration. 84 T.C. at 1153-63. The only cost increase for Lilly P.R. to which Lilly specifically objects on appeal is a charge to Lilly P.R. for general research and development expenses. Because we reject the Tax Court’s basis for allocating general research and development expenses to Lilly P.R., we uphold this objection. In all other respects we affirm the Tax Court’s cost adjustments. 24

The Tax Court’s revision of the fractions used to allocate profits under the 1968 formula are inherently more controversial. The Tax Court accepted two elements of the revenues guaranteed to Lilly P.R. under the 1968 pricing formula: locational savings and a 100% return on manufacturing costs. But it increased Lilly’s return on marketing costs from 25% to 100% and Lilly’s share of the residual return to intangibles (i.e., Darvon profits remaining after the deduction of locational savings, manufacturing profits and marketing profits) from 40% to 45%. 84 T.C. at 1152, 1163-67. In essence, the Tax Court ordered that Lilly take a larger cut of the Darvon earnings as returns on its marketing activities, and that the ratios used to divide the residual returns to Darvon intangibles between Lilly and Lilly P.R. change from 40-60 to 45-55.

Different factors supported these two adjustments. The Tax Court noted that a 1965 study conducted by Lilly had concluded that the parent company’s return on the marketing of Darvon products purchased from a manufacturing subsidiary in Puerto Rico should be 90% to 100% of marketing costs. The Tax Court concluded, based on its exposure to the economics of the pharmaceuticals industry and to the details of Lilly’s Darvon operations, that a return of only 25% of marketing costs would be “unreasonably low considering [Lilly’s] extensive marketing operations and net sales of Darvon and Darvon-N products in excess of $55 million and $73 million during 1971 and 1972.” 84 T.C. at 1151-52. With respect to the division of the residual profits between marketing and manufacturing intangibles, the Tax Court carefully assessed Lilly’s expert testimony concerning the relative importance of Lilly’s trademark and goodwill on the one hand and Lilly P.R.’s manufacturing intangibles on the other. The court agreed that the patents and manufacturing know-how contributed more to Darvon earnings than the trademark and other marketing intangibles, but found that a 45-55 split more accurately reflected the relative contributions than the 40-60 split that the 1968 formula incorporated.

Allocations of combined revenues under the profit split method are inherently imprecise. No unassailably precise methodology exists for determining normal profit rates on marketing expenses or the relative contributions of manufacturing and marketing intangibles. These judgments must rely largely on intuitions informed by an understanding of the business in which the affiliated companies are engaged. Employees of the industry in question and academics who study it — groups which were both represented among Lilly’s experts— bring considerable expertise to bear on this task. But absolute deference to expert assessments that cannot be objectively validated would essentially nullify section 482’s mandate to reallocate where “necessary in order to prevent evasion of taxes or clearly to reflect the income of any ... organizations, trades, or businesses.” For this reason, the courts have afforded a rebuttable presumption of correctness to the Commissioner’s section 482 determinations. See, e.g., Baldwin-Lima-Hamilton, 435 F.2d at 185; Spicer Theater, Inc. v. Commissioner, 346 F.2d 704, 706 (6th Cir.1965).

Lilly rebutted the presumption of correctness and showed that the Commissioner’s notice of deficiency was wrong, but it made no definitive showing that its allocations of Darvon income mimicked arm’s length transactions. Unable to justify the Darvon prices under any of the specifically enumerated methods set forth in the regulations and in Revenue Procedure 63-10, Lilly was forced to defend its prices as the product of some “other appropriate method." When the Tax Court found Lilly’s alternative wanting in certain respects, the Tax Court was obliged to establish the proper allocations, based on its extensive exposure to the testimony at trial and the voluminous documentary record in this case. See Baldwin-Lima-Hamilton, 435 F.2d at 187; Nat Harrison Assocs., 42 T.C. at 616-17; Ach, 42 T.C. at 126-27; see also American Terrazzo Strip Co., 56 T.C. at 973. The Tax Court was no more bound to approve or disapprove the Commissioner’s allocations in their entirety than we are bound to view the Tax Court’s conclusions as a package deal; and Lilly has not persuaded us that the Tax Court’s application of the profit split method was unreasonable. We therefore affirm the Tax Court’s increase in Lilly’s returns on marketing expenditures and on marketing intangibles for 1971 and 1972.

B. Reallocations for 1973

The Darvon market changed dramatically in 1973 due to the expiration of the principal propoxyphene patent in late 1972 and the prompt entry of other manufacturers into the market. Lilly relied on sales from one such manufacturer, Milan Pharmaceuticals, Inc., to an unaffiliated purchaser, Smith, Kline & French Laboratories, to justify the 1973 transfer prices for Darvon products under the “comparable uncontrolled price method” of determining arm’s length prices. The benchmark product, “SK-65” capsules in 500-eapsule bottles, traded between Milan and Smith Kline at $7.55 per bottle, while Lilly P.R.’s price for a similar product was $12.17. Lilly sought to reconcile this difference, arguing that Lilly P.R.’s higher price reflected Lilly P.R.’s more generous credit terms, stricter quality control, continued ownership of the napsylate patent and other factors. The Tax Court concurred with Lilly’s use of the Milan product as a benchmark. The Tax Court concluded, however, that about half of Lilly’s suggested per bottle adjustment was unwarranted. After examining these adjustments m considerable detail, see 84 T.C. at 1176-86, the Tax Court found that Lilly should have purchased the benchmark propoxyphene capsules from Lilly P.R. for $9.80 instead of $12.17 per bottle. Generalized across the entire line of Darvon products, this meant that Lilly’s 1973 profits should have been calculated as if it had purchased Lilly P.R.’s output at 66 percent below the resale prices to wholesale drug distributors; rather than for 56 percent below, the discount dictated by the pricing formula that Lilly and Lilly P.R. actually used in 1973.

Our review of the Tax Court’s analysis of the 1973 prices parallels our review of the 1971 and 1972 prices. For reasons given earlier, we reject the Tax Court’s attempt to enforce a contribution from Lilly P.R. to Lilly’s general research and development expenses. (The amount of this specific adjustment is unclear from the Tax Court’s opinion because some categories of costs are aggregated. 84 T.C. at 1185-86.) In all other respects, we affirm the Tax Court’s detailed category-by-category adjustments, which appear to be reasonable on their face and to which Lilly has not, at any rate, raised any specific objection. 25

IV. Conclusion

For the foregoing reasons, the Tax Court’s decision is affirmed in part and reversed in part. The case will be remanded for recalculation of the Tax Court’s real-locations of income to Lilly from Lilly P.R. The revised reallocations shall not require contributions from Lilly P.R.’s income stream to Lilly’s general research and development expenses.

AFFIRMED IN PART, REVERSED IN PART AND REMANDED.

1 . See Revenue Act of 1921, Pub.L. No. 67-98, § 262, 42 Stat. 227, 271 (1921). The current version of this investment incentive, which now takes the form of a tax credit instead of an exemption, appears at 26 U.S.C.A. § 936 (West Supp.1988). Since 1982, this section has included a provision, subsection 936(h), which permits possessions corporations to earn income from intangibles only if they compute income under one of two accounting methods (the “cost share” method or the "profit split” method). Under either approved method, the mainland affiliate must receive a share of the income generated from intangible assets, not less than a specified proportion of research and development expenses that the affiliated group incurs in the relevant "product area.” See Tax Equity and Fiscal Responsibility Act of 1982, § 213(a), Pub. L. No. 97-248, 96 Stat. 452 (codified as further amended at 26 U.S.C.A. § 936(h)(5)(C) (West Supp.1988)). Section 936(h) was added to the Code in response to this case. But both the Senate and House reports expressly warned against any inference that the amendment espoused Congress’ views on the meaning of the preexisting statutory language. See S.Rep. No. 494, 97th Cong., 2d Sess. 158 & n. 4, reprinted in 1982 U.S.Code Cong. & Admin.News 781, 921 & n. 4; H.Conf.Rep. No. 760, 97th Cong., 2d Sess. 504 & n.*, reprinted in 1982 U.S.Code Cong. & Admin.News 1190, 1281 & n*.

2 . Section 351(a), which has not changed materially since the period in question, provides

No gain or loss shall be recognized if property is transferred to a corporation by one or more persons solely in exchange for stock or securities in such corporation and immediately after the exchange such person or persons are in control (as defined in section 368(c)) of the corporation.

26 U.S.C. § 351 (1982).

3 . Prior to the adoption of the settlement formula, Lilly had set transfer prices at levels calculated to allow Lilly to recover the expenses of selling and distributing Darvon products and to earn a profit equal to 90% to 100% of those expenses. The settlement formula increased Darvon earnings allocated to Lilly.

4 . At all times relevant to this proceeding, Section 482 stated:

In any case of two or more organizations, trades, or businesses (whether or not incorporated, whether or not organized in the United States, and whether or not affiliated) owned or controlled directly or indirectly by the same interests, the Secretary or his delegate may distribute, apportion or allocate gross income, deductions, credits, or allowances between or among such organizations, trades, or businesses, if he determines that such distribution, apportionment, or allocation is necessary in order to prevent evasion of taxes or clearly to reflect the income of any of such organizations, trades, or businesses. Internal Revenue Code of 1954 § 482, Pub.L. No. 83-591, § 482, 68A Stat. 162 (codified as amended at 26 U.S.C.A. § 482 (West Supp.1988)). Section 1231(e)(1) of the Tax Reform Act of 1986 amended section 482 to provide that for taxable years beginning after 1986 "[i]n the case of any transfer (or license) of intangible property ..., the income with respect to such transfer or license shall be commensurate with the income attributable to the intangible.” Pub.L. No. 99-514, Title XII, § 1231(e)(1), 100 Stat. 2562. The accompanying House Report warned that “[n]o inference is intended as to whether the same result could nevertheless be reached under present law for transfers prior to these effective dates.” H.R.Rep. No. 426, 99th Cong., 2d Sess. 426, reprinted in 1986-3 C.B. Vol. 2 at 1426.

5 . Citations co Treasury Regulations refer to the regulations effective under the 1954 Code. Treas.Reg. § 1.482-1(d)(5) (1986) provided:

Section 482 may, when necessary to prevent the avoidance of taxes or to clearly reflect income, be applied in circumstances described in sections of the Code (such as section 351) providing for nonrecognition of gain or loss. See, for example, National Securities v. Commissioner of Internal Revenue....

6 . When this case was tried on remand, the claims court determined that the stock distribution was driven primarily by tax considerations and entered judgment for the government. Ruddick Corp. v. United States, 3 Cl.Ct. 61 (1983), aff’d, 732 F.2d 168 (Fed.Cir.1984) (mem. op.).

7 . The Commissioner argues that National Securities eliminated any conflict between sections 351 and 482 by establishing that section 351 controls only the timing of the recognition of a gain or loss, while section 482 determines which taxpayer may claim the gain or loss at the appropriate time. 137 F.2d at 602. Whatever merit this approach may have for harmonizing nonrecognition transfers and section 482 reallo-cations where a transfer is followed promptly by the transferee’s disposal of the transferred asset — and it is noteworthy that the National Securities’ decision did not rely entirely on this rationale, see id. at 603 — the "no conflict” position seems untenable where the recognition transaction initiated by the transferee occurs years after the initial transfer. See Ruddick I, 648 F.2d at 752. This seems especially true where, as here, the transferred asset is used rather than retransferred intact. See generally Note, Section 482 and the Nonrecognition Provisions: Resolving the Conflict, 77 Nw.U.L.Rev. 670, 685-87 (1982).

8 . See Dolese v. Commissioner, 811 F.2d 543 (10th Cir.1987) (distribution of two tracts of partnership land to partners prior to donation and sale); Foster v. Commissioner, 756 F.2d 1430 (9th Cir.1985) (transfer of appreciated property prior to sale by transferee), cert. denied, 474 U.S. 1055, 106 S.Ct. 793, 88 L.Ed.2d 770 (1986).

9 . The Commissioner's position ... is that parties dealing at arm’s length in negotiating the amount of a royalty for use of the manufacturing intangibles would have struck a deal allowing the manufacturer to realize a profit only on its manufacturing activity. They would not have agreed that the manufacturer was entitled to derive any income attributable to the intangibles.

Brief for the Commissioner 34.

10 . The Commissioner refers us to financial data that purportedly reveal abnormally high profits for Lilly P.R., placing particular emphasis on Lilly P.R.’s high ratios of earnings to tangible capital and earnings to sales. But these ratios are only relevant if one accepts the legal conclusion that the Commissioner’s argument must prove — that the transfer of the Darvon intangibles, a key input in the production of Darvon products, gave Lilly P.R. no claim to the income from those intangibles under the tax code. Moreover, the Commissioner’s financial comparisons between Lilly P.R. and Lilly also ignore differences between pharmaceuticals and Lilly’s other lines of business and the tax and labor cost advantages of operating in Puerto Rico. The Tax Court correctly disregarded these data in its assessment of the applicability of section 482.

11 . See, e.g., Reply Brief for the Commissioner at 8 ("By virtue of its special tax status as a possessions corporation, Lilly P.R. exploited the intangibles in a manner and to a degree taxpayer could not have achieved.”).

12 . Section 1.482-2 of the Treasury Regulations was promulgated five years after the issuance of Rev.Proc. 63-10, but did not supersede it. Rev. Proc. 68-22, issued contemporaneously with the 1968 regulations, gave taxpayers the option of applying either Rev.Proc. 63-10 or the regulations. Rev.Proc. 68-22, 1968-1 C.B. 819.

13 . Manual Supplement 42G-86, Allocation of Income and Expenses Between United States Companies and Their Manufacturing Affiliates in Puerto Rico at 9 (Jan. 15, 1963), reprinted in Joint Appendix 264, 272.

14 . The Commissioner’s argument that Rev.Proc. 63-10 applies only to intangibles developed in Puerto Rico relies principally on the following excerpt from that document:

The guidelines concern what may be considered as the standard type of allocation problem that has arisen in these cases. They do not deal with other problems that may be involved in particular cases, including those which may be present in cases involving the transfer of income-producing intangibles from the United States to an affiliate located in Puerto Rico.

Rev.Proc. 63-10 § 1, 1963-1 C.B. at 490.

Read in isolation, this language is susceptible to the Commissioner’s interpretation. We agree with the Tax Court, however, that in light of the Service’s contemporaneous guidance and the practical insignificance of cases in which Puerto Rican manufacturing affiliates of U.S. companies derive appreciable income from intangibles developed in Puerto Rico, the quoted passage must be construed as a caution that Rev.Proc. 63-10 provides no guidelines for problems other than those relating to the application of section 482 to the transfer of intangibles. 84 T.C. at 1188-89.

The Commissioner also seeks support in various internal communications by Treasury Department officials who advocated that the Internal Revenue Service bar Puerto Rican affiliates from claiming income derived from transferred intangibles. See, e.g., Letter from Acting Treasury Secretary Henry Fowler to the Governor of Puerto Rico, Jan. 3, 1963 (Exhibit 166 at 2) (referring to Treasury’s "understanding ... that in fact intangibles normally are not transferred to affiliates in Puerto Rico”); Report on Interviews with Puerto Rican Tax Lawyers and Accountants, Nov. 28, 1962 (Memo to file by Paul Guttentag, Dec. 1, 1962) (Exhibit GC). Lilly, in response, points us to Treasury Department documents supporting its view that income from transferred intangibles belonged to Puerto Ri-can manufacturing affiliates. Lilly asserts that an internal study completed prior to the issuance of Rev.Proc. 63-10 suggested that the then prevailing interpretations of sections 351 and 931 allowed Puerto Rican manufacturing affiliates to claim income derived from transferred intangibles. See The Allocation of Income Between Affiliated Puerto Rican and American Corporations at 58-64 (Report of Dr. Charles Berry to Assistant Secretary Stanley Surrey, Aug. 1, 1961).

The reports and memoranda upon which the Commissioner relies establish at most that some officials within the Treasury Department argued against the approach that was ultimately taken in Rev.Proc. 63-10 and the accompanying guidance and that the Service was unaware of the magnitude of these intangible transfers for several years after the Revenue Procedure was issued. They do nothing to counter the logical force of the Tax Court’s view that the Revenue Procedure applied to intangibles that were transferred from mainland affiliates or the unequivocal language to this effect in the Service’s guidelines.

15 . The Tax Court relied on the testimony of Dr. James Wheeler, an expert for the Commissioner, that Lilly’s pharmaceuticals division would have operated at a loss if it had transferred all the patents and manufacturing know-how relating to its nine next most profitable drugs to Lilly P.R. and purchased these drugs at prices that allowed Lilly P.R. to realize the same markup it realized on Darvon products. 84 T.C. at 1130, 1161; see Tr. 1798-1801, 1855-56 (Sept. 28, 1981); Respondent’s Exhibit FP. The testimony does not justify the use of the Lilly P.R. Darvon markup, as opposed to the Darvon profit split, across the board. We therefore fail to see the relevance of Dr. Wheeler’s scenario to the legitimacy of the Darvon transfer prices, which were set under a formula that did not guarantee Lilly P.R. a particular percentage markup.

16 . This assumption underscores that uncertainty regarding the characteristics of the benchmark arm’s length transaction in any particular context introduces enormous ambiguity into the application of section 482. Lilly argues that even if the intangibles transfer had depressed the parent company's cash flow below the level needed to sustain “normal” research and development efforts, Lilly was under no obligation to continue to develop new pharmaceuticals; it could, in the normal course of events, have decided to move into other areas. In a slightly different vein, it might have decided, applying its normal approach to corporate financial planning, that the unusually high rate of growth in the value of Lilly P.R. stock justified liquidating other long-term investments or even borrowing the needed cash. We do not need to decide whether a reduction in research and development spending or a sharp departure from established methods of funding these activities would have defeated Lilly’s claim that the transfer met the arm’s length standard; these developments would certainly have presented a closer case than the one now before us. We only note, in passing, that the wide range of financial options that a business like Lilly might reasonably exercise in transactions unaffected by tax considerations severely complicates arguments that the form of a transaction alone causes it to fail the arm’s length test.

17 . The Tax Court’s opinion states that Lilly’s operating income from marketing Darvon products before subtracting research and development expenses was $17.0 million in 1971 and $16.6 million in 1972. General research and development expenses allocable to Darvon — to be distinguished from expenses related to research and development specific to the Darvon line, which were paid by Lilly P.R. — were $7.1 million and $7.8 million for 1971 and 1972, respectively. See 84 T.C. at 1161 & n. 99, 1163 n. 100. (The Tax Court did not perform these computations for 1973, but the Commissioner does not challenge Lilly’s assertion that the results would be similar.)

18 . The Tax Court essentially acknowledges this point in its statement that by exchanging manufacturing tangibles for stock Lilly "deprived itself of the means to carry on a portion of its general research and development activities, or at least, was forced to fund those activities with income that would otherwise have gone elsewhere (e.g., shareholders’ dividends, capital investments).’’ 84 T.C. at 1161.

19 . See Treas.Reg. § 1.482-2(e)(l)(iii) (1986). See generally id. § 1.482-2(e)(2) to -2(e)(4) (1986). Rev.Proc. 63-10 also set forth specific guidelines for applying the arm's length standard. See Rev.Proc. 63-10 §§ 3.021-.023, 4.03-.04, 1963-1 C.B. 490, 493-97. Rev.Proc. 68-10, 1968-1 C.B. 819, which was issued contemporaneously with the section 482 regulations, afforded the taxpayer the option of continuing to determine arm’s length prices under Rev.Proc. 63-10 or relying on the new regulations. Although Rev.Proc. 63-10 does not impose the same precise ranking of methods, the two sources of guidance are in no way inconsistent and Lilly does not argue that any differences between the Revenue Procedure and the regulations are relevant here.

20 . The Court of Claims has stressed the need for precision in applying the resale price regulations:

It is quite plain from the text of the regulation itself that the evident purpose for the use of the particular resale price method, as set forth in the regulation, is to proffer a relatively precise mechanism for determining a realistically comparable, controlled, arm’s-length resale price — not to leave the taxpayer, the Service, or the courts to grope at large for some figure drawn out of overly general in-dices or statistics.

E.I. Du Pont de Nemours v. United States, 608 F.2d 445, 451, 221 Ct.Cl. 333 (1979).

21 . By focusing on ratios of operating income to operating expenses, Lilly's experts discounted the possibility that the Darvon trademark (and other product-specific marketing intangibles) accounted for a higher proportion of Darvon earnings than the product-specific marketing intangibles associated with Lilly’s next nine leading products contributed to earnings for those products. The experts’ report defends this view on the grounds that the Darvon trademark would have been worthless (during 1971 and 1972) without access to products manufactured under the patents owned by Lilly P.R. Report of Petitioner’s Economic Experts at 10 (Sept. 1, 1981) (Exhibit 238). Thus, the experts hypothesized that in arm’s length bargaining over prices the patent owner would capture virtually all the rents associated with the trademark. Given the extensive evidence on the value of the Darvon trademark, and the allocation of 40% of intangible profits to the trademark under the 1968 formula, we do not think that the Tax Court erred in declining to rely upon an analysis that discounted the value of the trademark so heavily.

22 . Lilly, citing Akers v. Commissioner, 798 F.2d 894, 897 (6th Cir.1986), and Capitol-Barg Dry Cleaning Co. v. Commissioner, 131 F.2d 712, 715 (6th Cir.1942), asserts that the Tax Court committed reversible error by ignoring the uncon-troverted testimony of its experts. But these cases are inapposite. In Akers, the Tax Court had sharply reduced the value that expert appraisals assigned to certain equipment without offering any basis for the reduction. In Capitol-Barg, the court imposed its own unsupported intuitions about reasonable executive compensation. Here, in contrast, the Tax Court assessed the experts’ methodology against the Commissioner’s published guidelines and identified sound legal reasons for not accepting them.

23 . Lilly cites United States Steel Corp. v. Commissioner, 617 F.2d 942 (2d Cir.1980), for the proposition that the regulations are to be interpreted flexibly when the Commissioner and the courts review a taxpayer's proof that its prices were arm’s length. In United States Steel, the Second Circuit overturned the Tax Court’s reallocation of profits from a shipping subsidiary to a parent steel manufacturer. The Tax Court refused to accept the shipping company’s rates to unrelated firms as “comparable uncontrolled prices” for purposes of defending rates charged the parent company for higher volume hauls. The Second Circuit upheld the taxpayer’s use of the lower volume rates as a benchmark, cautioning against ”engraft[ing] a crippling degree of economic sophistication onto a broadly drawn statute,” by equating "comparable” with "identical” for purposes of applying the comparable uncontrolled price approach. Id. at 951.

We think that Lilly’s reliance on United States Steel is misplaced. The Tax Court’s rejection of Lilly’s justification under the resale price method rested not on some hypertechnical and overly restrictive interpretation of the resale price guidelines, but on Lilly’s clear failure to identify similar purchase and resale transactions to serve as a reliable example of arm's length marketing profits.

24 . Some fraction of Lilly’s general research and development expenses may in fact have been Darvon-related research and development that should have been billed to Lilly P.R. under the companies’ joint research and development agreement. The Tax Court declined to resolve this issue in light of its substantial reallocation of general research and development costs to the subsidiary. 84 T.C. at 1162-63. On remand, the Tax Court will have an opportunity to reexamine this question and to correct any leakage from the Darvon-related research and development account.

25 . Lilly refers obliquely to "several errors” that the Tax Court allegedly committed in its analysis of the Milan price. Reply Brief for Appellants and Answering Brief for Cross-Appellees at 20. However, the only clue as to the nature of these errors is a reference to a Motion for Reconsideration filed with the Tax Court. Id at 20 n. 42. Incorporation by reference is insufficient to place these arguments before this court. See Hunter v. Allis-Chalmers Corp., 797 F.2d 1417, 1430 (7th Cir.1986). We therefore decline to pursue these alleged errors further.

Lilly also suggests that the Tax Court’s detailed comparison of the Milan and Lilly P.R. prices conceals the real impetus for the Tax Court’s 1973 reallocations, which was "to find allocations of income to correct for the misper-ceived distortion of income” that the Tax Court attributed to the 1966 exchange of Darvon intangibles for stock. Brief for Appellants at 43. Taking this characterization of the Tax Court’s opinion one step farther, Lilly suggests that the Tax Court’s entire analysis was aimed at imposing Congress’ solution to the problem posed by intangibles ownership by possessions manufacturers {see note 1 supra) to a period that ended nine years before the effective date of the relevant legislation. Id. at 37, 38, 47-48; see Brief for Appellants and Answering Brief for Cross-Appellees at 20 n. 44 (citing 26 U.S.C. § 936(h) (1982)). The Tax Court, however, expressly recognized that this element of section 936(h) was strictly prospective and that “no inference can be drawn from those provisions and the fact of their enactment can have no effect on our decision herein.” 84 T.C. at 1112 n. 44. It is perhaps possible that the judge disingenuously relied on considerations that the decision itself expressly disavows. But to urge reversal based merely on speculation to that effect is a course most unlikely to be rewarded.