Table of Contents
III. Backstopping the Debt Dividend Rule
This report will focus on a key decision inherent in the April 4 proposal of regulations under section 385: the abolishment of debt dividends between related parties. At its core, prop. reg. section 1.385-31 prevents a corporation from issuing debt for tax purposes to its parent company in a distribution. Ever since the Second Circuit's 1956 decision in Kraft,2 the received wisdom has been that debt issued as a dividend (that is, debt issued to a shareholder for no consideration) should be respected if the usual indicia of debt are established. The proposed regulations upend that wisdom. They do not, however, upend a different article of faith -- namely, that debt between related parties will generally be respected if the usual indicia of debt are established. This report examines from a policy and technical point of view Treasury and the IRS's decision to go after debt dividends while leaving in place the tax law's general acceptance of related-party debt.
Taxpayers and tax advisers have responded to the proposed regulations with an unusual degree of concern over their far-reaching -- and in some cases, counterintuitive and potentially unintended -- consequences.3 Although much of the articulated concern centers on a funding rule and a 72-month per se rule,4 the reaction can also be attributed to the proposed regulations' approach of targeting debt dividends in the first place. Tax lawmakers are well-advised to avoid cliff effects. Accordingly, the proposed regulations cover not only debt dividends but also transactions that are similar in some sense to debt dividends -- and transactions that are similar to transactions that are similar to debt dividends, and so on.
The likely true target is over-leverage of a domestic borrower. In the proposed regulations, debt dividends serve as a proxy for over-leverage. However, a debt dividend does not always lead to over-leverage, nor is over-leverage limited to situations involving debt dividends. An approach that directly addressed related party over-leverage would lead to fewer counterintuitive results.
The centerpiece of the proposed regulations is the equity treatment of related-party debt issued in a distribution. Prop. reg. section 1.385-3(b)(2)(i) provides that with some exceptions, "a debt instrument is treated as stock to the extent the debt instrument is issued by a corporation to a member of the corporation's expanded group . . . [i]n a distribution" (the debt dividend rule). One of the exceptions permits those distributions up to the amount of the corporate issuer's current-year earnings and profits (the E&P exception).5 There is also a $50 million threshold exception.6
The other elements of prop. reg. section 1.385-3 are all intended to backstop the debt dividend rule, although in some cases, they sweep farther. Prop. reg. section 1.385-3(b)(2)(ii) provides that debt issued by a corporation to a member of the corporation's expanded group "in exchange for expanded group stock" is treated as stock (the group member stock acquisition rule).7 Likewise, debt issued by a corporation to a member of the corporation's expanded group "in exchange for property in an asset reorganization" is also treated as stock, but only to the extent that under the plan, a shareholder that is a member of the expanded group receives the debt instrument for its stock in the transferor in the reorganization (the asset reorganization with boot rule).8
The proposed regulations would overrule Kraft. The Second Circuit's majority and dissenting opinions in that case articulate key arguments that remain relevant in considering the appropriate tax treatment of debt dividends.
Kraft involved two domestic corporations: a parent and a subsidiary. The parent was an obligor on third-party debt, and the subsidiary was an operating company. For the years in question, the filing of consolidated returns was not permitted. Thus, the operating company would have had taxable income without any offset for the interest expense incurred by the parent. The subsidiary, engaging in self-help, paid a dividend in the form of a debt instrument to the parent. The debt instrument was plain vanilla -- an unconditional obligation to pay principal and interest -- and it had no equity-like features. In the taxpayer's view, interest on the debt owed to the parent was deductible by the subsidiary, thereby reducing the subsidiary's taxable income.
The Second Circuit agreed with the taxpayer. First, the court observed that the parent-subsidiary relationship alone did not preclude the debt from being respected. Acknowledging that the parent corporation could deal as it wished with the subsidiary, the court nonetheless found the debtor-creditor relationship to be genuine, not a sham. It acknowledged that in a "broad economic sense" (emphasis in the original), the form of relationship between a sole stockholder and a wholly owned corporation has limited significance, but it found that the law in general and tax law in particular "affords significance to and honors the type of investment chosen."9
The court also found that Congress's decision to abolish consolidated returns reflected a view that the parent and subsidiary should be respected as separate taxable entities, which supported the argument that the debt should be respected. Congress chose to tax those entities separately, even though it recognized that "in a realistic economic sense affiliated corporations were but departments or branches of single enterprises," the court said. It concluded that a "necessary corollary" of that choice was that "proper legal transactions among affiliates must be accorded tax recognition."10 The court also reasoned that parent-subsidiary debt is generally respected for bankruptcy purposes. It disagreed with the Tax Court's view that the subsidiary was "a department" of the parent.11
Second, the court rejected the commissioner's argument that the intercompany debt should be disregarded because no new capital was infused into the subsidiary. The commissioner asserted that because the debt was not issued for cash or issued against accumulated earnings,12 it represented the initial equity capital, which could not be transmuted into debt. The court, however, believed that the statute contemplated the creation of debt through a distribution or recapitalization, and it viewed the transaction as if cash had been distributed by the subsidiary to the parent and then reinvested by the parent in the form of debt.
Third, the commissioner did not convince the court that a disproportionate ratio of debt to capital had been created.
Fourth, the court did not believe that the fact that the debt dividend was tax motivated meant that it should be disregarded. It distinguished Gregory13 and other substance-over-form cases on the grounds that the transactions in Kraft were genuine. The court saw Gregory and the other cases as permitting tax minimization purposes but not a "sham or masquerade."14
The dissenting judge in Kraft characterized the result as "startling."15 He was troubled by the tax avoidance flavor of the transaction: "Surely the process whereby the taxpayer has supplemented its product -- cheese -- by a marvelous by-product -- the pure gold of tax avoidance -- will stimulate imitators."16 The dissent rejected the majority's view that Congress brought the result upon itself by abolishing consolidated returns. To the dissent, it was not a corollary of abolishing consolidated returns that intercompany transactions must be respected, since the result of doing so was inconsistent with congressional intent: "Congress in attempting to stop a comparatively small leak in revenues through the operation of consolidated returns from intertwined corporations now finds itself with the intent by repeal of this provision to open up a large leak, indeed, operable merely by denominating an intercorporate allocation of surplus a debt, not a dividend." The dissent was skeptical of the intercompany debt, stating that repayment is "obviously" not of current concern and possibly not of "potential concern."17 It endorsed the Tax Court's approach of testing the genuineness of intercorporate debt objectively and viewed each of the factors rejected by the majority opinion as persuasive.
In targeting debt dividends, Treasury and the IRS decided not to go after all related-party debt.18 A long-standing premise of tax law is that related parties can transact with one another and that those transactions are generally meant to be respected for tax purposes. In 1943 the Supreme Court held in Moline Properties19 that a taxpayer could not disavow the existence of a corporation organized for business reasons. The taxpayer had to take the good with the bad: The business advantages afforded by the corporation brought along with them the tax disadvantages. The Supreme Court concluded that the corporation "had a tax identity distinct from its stockholder,"20 and it rejected the Board of Tax Appeals' view that the corporation was "a mere figmentary agent."21 The Court's decision in that case has come to stand for the proposition that in the absence of a sham transaction, a corporation's separate existence will be respected for tax purposes. Consistent with Moline Properties, the code and regulations accept intercompany debt as a general matter. The premise of sections 163(e)(3), 267(a)(3), for example, is that debt can exist between related parties.22
But some have disagreed with Moline Properties, at least in the intercompany debt situation, arguing that deductions should not be allowed because intercompany debt is not real debt.23 For example, H. David Rosenbloom has argued that because corporations are only "pieces of paper," it is not inevitable that a tax system must give credence to putative debt between commonly owned corporations.24 He claims that by "most standards of economics, substance, or common sense, [related-party debt] is not debt" but rather merely a transfer of funds from one pocket to another -- not a true cost of doing business.25 From his perspective, "Two pieces of paper owned ultimately by a single economic interest cannot do business together."26 He contends that a putative debt instrument cannot represent an "unconditional promise when the parties can, with a wave of the pen, waive their prior handiwork. There is no sum certain when the sum identified can be decreased or increased through purely formal (and easily accomplished) means."27
Although the argument has force, related-party debt is not purely formal. The question of what claims, including debt claims, one entity in a group has against another can become significant when any entities in the group are bankrupt. The debt instrument's effect in bankruptcy is important, because bankruptcy is a key setting in which debt or equity status should make a difference. Different entities may have different creditor, shareholder, regulatory, and other constituencies all vying for a limited set of assets. Identifying the full panoply of claimants and assets of a particular entity, as distinguished from the group, becomes critical. In a bankruptcy, intercompany debt may not be freely rewritten, and the common ultimate equity ownership of the group becomes insignificant because equity is often wiped out. Tax law follows the lead of bankruptcy law in generally respecting the separate existence of separate entities. That said, when intercompany debt is subordinated, its effect may converge toward that of equity.
An implication of the argument that related-party debt should not be respected would be that an entity in a group must borrow from third parties to obtain interest deductions. But lenders to a multinational corporate group often require guarantees from numerous entities in the group. If, for example, a domestic corporation is a subsidiary in a foreign-parented group, a third-party lender to the domestic corporation would in many cases require guarantees from the foreign parent and foreign sister corporations of the domestic issuer, as well as from domestic subsidiaries of the domestic issuer. Economically, that borrowing by the domestic subsidiary may be similar to a borrowing by the foreign parent with a guarantee from all the same entities plus a guarantee from the domestic subsidiary. Thus, one could question why interest deductions should be denied on related-party debt owed from a domestic subsidiary to a foreign parent when the foreign parent has issued third-party debt with a guarantee from the domestic subsidiary.
In contrast with the position that all related-party debt should be rejected is the belief that related-party debt should somehow be limited by reference to third-party debt owed by other members of the corporate group. Under this approach, a group member's interest deductions for intercompany debt could be limited by reference to an allocation to that member of third-party debt owed by group members.
J. Clifton Fleming Jr., Robert J. Peroni, and Stephen E. Shay have followed Rosenbloom's lead in taking a skeptical view of cross-border related-party debt, but they appear to believe that some portion of third-party debt owed by group members should be able to be allocated to a domestic member of the group through related-party debt.28 They argue that under the "no deduction for costless payments principle," deductions should not be permitted on intercompany debt because the interest expense does not represent a cost to the group but rather a reallocation of income from one entity (the borrower) to the other (the lender).29 Those authors also believe that an entity in a group should be entitled to earn a market return on costs it incurs (but no greater than an arm's-length amount) and that the payer member of the group should be entitled to deduct that amount -- what they refer to as the "investment location primacy principle."30 Fleming, Peroni, and Shay do not think that either principle generally should result in an interest deduction on related-party debt. However, they maintain that a theoretically correct system would distinguish between interest on loans purely from group funds (which should be nondeductible, in their view) and interest on loans from funds supplied by a third party (which should be deductible to the extent not in excess of an arm's-length benchmark). As a practical matter, those authors recommend a proportional allocation system under which a U.S. subsidiary's share of group interest expense on debt owed to third parties would be allocated by reference to the subsidiary's assets relative to worldwide assets and would be deductible (but not in excess of actual interest payments by the subsidiary).
Their approach resembles one of the Obama administration's budget proposals for 2017.31 The administration calls for legislative change on the view that current law, which does not account for the group's overall third-party debt, allows foreign-parented groups to over-leverage U.S. operations. It proposes a proportionate allocation system based on the U.S. subsidiary's financial statement earnings relative to those of the group. The proposal would limit interest deductions of a U.S. entity that is a member of a group that prepares consolidated financial statements by calculating the ratio of the U.S. subsidiary's stand-alone net interest expense for financial reporting purposes with its proportionate share of the group's net interest expense for financial reporting purposes. The U.S. subsidiary's proportionate share would be based on financial statement earnings computed by adding back net interest expense, taxes, depreciation, and amortization. If the stand-alone net interest expense for financial reporting purposes exceeds the subsidiary's proportionate share of the group's net interest expense for financial reporting purposes, interest deductions for tax purposes would be disallowed in the same ratio that the excess net interest expense for financial reporting purposes bears to the U.S. subsidiary's total stand-alone net interest expense for financial reporting purposes.32
Thus, both the budget proposal and the proposal by Fleming, Peroni, and Shay provide a framework for accepting intercompany debt when the group overall is leveraged with debt owed to third parties. In that circumstance, they support a portion of that group leverage being pushed down to a U.S. subsidiary in the group and a portion of the interest expense being deducted by the U.S. subsidiary. That is, the U.S. subsidiary's interest deductions on related-party debt would be limited by reference to the subsidiary's allocable portion of the debt owed by the group to third parties. Fleming, Peroni, and Shay note, however, that under that framework, if the group is not leveraged, a U.S. subsidiary should not be leveraged with related-party debt because that leverage would merely reallocate the U.S. subsidiary's income, not reflect a cost to the group of doing business.33
A rationale for permitting interest deductions on related-party debt when the group has third-party debt is that the tax system should not require the subsidiary to have its own third-party debt. Many groups prefer for administrative reasons to have their third-party debt issued by one entity, often the parent, and to provide guarantees from subsidiaries.
As noted, it may make little substantive difference for a subsidiary to borrow from a third party and obtain guarantees from the other members in the group versus the subsidiary borrowing from the parent that in turn borrows from third parties with subsidiary guarantees. Status as the borrower rather than as guarantor on third-party debt is largely formalistic when the guarantors have the same responsibilities and exposure under a debt instrument as a borrower. In those cases, debt is economically lent to the group, not to a single member of the group, and it would seem appropriate for the tax system to permit deductions on related-party debt that represents an allocation of the group debt.
One argument against the allocation approach is that the appropriate tax base for the U.S. subsidiary should be determined by reference only to that entity's characteristics, not the characteristics of the group as a whole. Indeed, this has been the approach of current law. After determining under Plantation Patterns34 that a guarantor is not the issuer of a debt instrument for tax purposes, traditional debt-equity analysis generally looks only at the debt-equity ratio, the wherewithal, and the other characteristics of the issuer to determine whether an instrument is debt or equity. Referring to the characteristics of the group as a whole could cut in favor of the government or the taxpayer, depending on the facts. For instance, companies in the same business as the U.S. subsidiary might or might not typically have the amount of leverage that the group as a whole has. Arguably, the U.S. subsidiary should be permitted to deduct interest on the amount of debt appropriate to it regardless of the amount of leverage on the group. That approach would generally be premised on the view that the related-party debt is real and arm's-length and would have consequences in a bankruptcy.
The proposed regulations reject the theory that interest on related-party debt should inevitably be nondeductible, but they also reject an approach that would limit deductions on related-party debt by reference to the issuer's proportionate allocation of group debt. Instead, the proposed regulations treat debt issued in the form of a dividend as stock.
From one perspective, debt dividends do seem distinguishable. Debt is meant to arise when the borrower receives something (usually cash) and owes it back. So debt created not in exchange for anything raises questions.
But from another perspective, that debt does not seem strange at all. Dividends are a common and familiar type of transaction. At their essence is a transfer not in exchange for anything. If a corporation can pay a dividend in the form of $100 cash, it would seem to follow that a company can pay a dividend in the form of a promise to pay $100 in cash.
By the same token, a shareholder should be able to adjust the capital structure of its subsidiary over time. If a corporation is incorporated by the shareholder transferring $100 worth of assets to the corporation in exchange for its stock and debt, the debt is issued in exchange for something -- a portion of the assets transferred to the corporation at the time of incorporation. No debt dividend occurs in form or substance. Even for a person inclined to be suspicious of debt dividends, the issuance of debt by the corporation on formation in exchange for assets seems unassailable (unless one is of a mind to reject related-party debt altogether). Suppose the corporation's assets then double in value after formation. If the shareholder at that point sought to readjust the corporation's capital structure by replacing debt with equity, the simplest way to do so would be for the subsidiary to pay a debt dividend.35 Further -- although one would not do this because it would not be tax efficient -- all the assets could be distributed from the corporation to the shareholder, and the shareholder could then recontribute them to the corporation in exchange for stock and an amount of debt appropriate to the corporation now that it owns assets worth $200. Thus, formally, debt created through a dividend could be created another way. One could argue that the capital structure of the corporation should not be required to be established upon formation of the corporation and should be permitted to be adjusted over time as the value of the corporation changes. Moreover, conceptually, a debt dividend is the same as a liquidation of the corporation followed by a reincorporation of the corporation with a new ratio of debt to equity. It would seem that the debt dividend is not itself the concern.
Consistent with that view, the code has long acknowledged that debt can be issued as a dividend. The 1954 code provided that E&P would be reduced by the principal amount of an obligation distributed by a corporation on its stock.36 In 1984 section 312 was amended to clarify that for a distribution of an obligation having original issue discount, E&P would be reduced by the aggregate issue price of that obligation, not the principal amount of the obligation. Section 1275(a)(4), enacted in 1984, also acknowledges the possibility of debt dividends. It provides that debt distributed by a corporation on its stock is treated as if it had been issued for property. Thus, rather than eradicating debt dividends, Congress made them subject to the OID rules.37 By the same token, reg. section 1.301-1(l) contemplates a substitution of equity of a corporation for debt.
Among the Kraft arguments in favor of allowing debt dividends was the claim that Congress decided not to allow the filing of consolidated returns between two related domestic entities and hence had to accept the consequence that genuine transactions between two domestic entities must be respected. Under current law, consolidated returns are permitted. The relevance of debt dividends arises instead in other situations, including when the parent is foreign and the subsidiary is domestic. Applied in that setting, the Kraft argument would be that Congress, viewing the foreign parent and domestic subsidiary as representing a single economic group, could have permitted a portion of the foreign parent's debt to be deducted against domestic earnings. Having chosen not to do so, the argument continues, Congress acknowledged that the foreign parent is separate and therefore the government should take the good with the bad and respect genuine transactions between the parent and the subsidiary.
That argument is debatable, however. Truly recognizing that the group is a single enterprise would mean that the United States should be entitled to tax an allocable portion of the group's income, not merely permit an allocable portion of the group's interest expense to be deducted in the United States.38 That apportionment system may or may not be within international norms of comity insofar as it would likely involve taxing non-U.S.-source income. So it may be too quick to say that Congress had the chance to tax the parent and subsidiary and by virtue of treating them as separate for some purposes should be required to respect intercompany transactions.
Moreover, one could oppose respecting debt dividends with the same arguments as those that are made generally against related-party debt.
A rejoinder would be that debt dividends are of a piece with related-party debt and that if, for the reasons discussed in Part II.B above, one is prepared to accept related-party debt, then debt dividends should be respected as well. One could argue that if Treasury and the IRS are concerned about the over-leverage of foreign-parented domestic subsidiaries, that concern should be addressed more directly and in a more straightforward fashion than by tackling debt dividends as a proxy for over-leverage.
As mentioned, the group member stock acquisition rule serves as a backstop to the debt dividend rule. The group member stock acquisition rule treats debt as stock if the debt is issued to acquire stock of any member (domestic, foreign, parent, subsidiary, or sister) of the expanded group. The rationale for covering acquisitions of group member stock varies, however, depending on the facts.
The group member stock acquisition rule covers acquisitions of parent stock by a subsidiary in exchange for subsidiary debt. That acquisition would create hook stock -- stock in the parent owned by a subsidiary. If the proposed regulations did not cover the acquisition of hook stock for debt, taxpayers would likely substitute this transaction for debt dividends, although the economic consequences are not identical. Hook stock gives the subsidiary an economic interest in the parent's other assets, whereas a debt dividend provides no additional asset to the subsidiary.39 However, hook stock can be economically neutralized in part because the subsidiary can often waive dividends.40
It also makes sense that in covering debt dividends, the proposed regulations also cover acquisitions of a subsidiary's own stock for debt in section 302 redemptions.
Covering an acquisition of sister stock for debt is a less obvious decision because the transaction is a value-for-value exchange. An acquisition of sister stock does not have the signal feature of a dividend, which is that nothing is received by the debt issuer in exchange for the debt. Here, the debt issuer acquires something of value -- namely, stock in the sister company. But as with hook stock, if the proposed regulations did not cover this case, taxpayers would substitute acquisitions of sister company stock for debt dividends. In some cases, covering acquisitions of sister stock would be necessary to backstop the debt dividend rule, but in other cases, it is less apparent whether those acquisitions should be covered.
In the preamble,41 Treasury and the IRS provide two rationales for applying the proposed regulations to debt issued for sister stock. First, they argue that an acquisition of affiliate stock for debt is similar to a distribution of a debt instrument. They observe that Congress recognized that similarity in enacting section 304 and that taxpayers would substitute an acquisition of sister stock for debt for a debt dividend if acquisitions of sister stock were not covered by the proposed regulations. Second, they argue that issuances of related-party debt in exchange for sister stock have "limited non-tax significance" because those transactions do not change "ultimate ownership" and "introduce no new operating capital."42 Treasury and the IRS do acknowledge that the change in direct ownership of the sister can have nontax significance in some circumstances, such as "harmonization of a group's corporate structure following an acquisition."43 However, the analysis is more complex than the preamble suggests.
1. Relevance of section 304. First, although an acquisition of sister stock for a note is usually a section 304 transaction, the policies behind section 304 differ from those behind efforts to curb earnings stripping, making the relevance of section 304 questionable. The issuance of a note in exchange for sister stock would generally be assimilated to a distribution under section 301 by operation of sections 304, 302, and 301. But that treatment of the issuance of the note as a section 301 distribution is not a reason to apply the regulations to acquisitions of sister stock for a note, since the problem in the first place is not an E&P bailout -- the type of transaction section 304 is meant to police.
2. Expansive concept of expanded group. Second, the government's argument that ultimate ownership remains unchanged does not resonate with the expansive concept of expanded group provided in the proposed regulations. The definition of expanded group takes as its starting point the concept of an affiliated group as defined in section 1504(a),44 that group generally consisting of a common parent corporation and chains of 80-percent-owned corporate subsidiaries. The proposed regulations broaden the group concept, however, by layering on a notion of indirect ownership determined by applying section 304(c)(3). Section 304(c)(3) applies section 318, expanded for upstream and downstream attribution from and to corporations. The rules of section 318 notoriously include downstream attribution to partnerships. With that expansive view of indirect ownership, the expanded group definition in the proposed regulations abandons the section 1504(a) constraint of having a common parent corporation or the notion of having any common ownership at all. In some cases, an expanded group can consist of parties that are "related" only as a result of common ownership of a partnership or corporation.
Thus, in those cases, an acquisition of sister stock does in fact change ultimate ownership. Consider, for example, two separate public companies whose only existing relationship is that they are both partners (no matter how small) in a partnership somewhere in the world. Suppose each public company has many corporate subsidiaries and that a corporate subsidiary of one of the public companies is meant to sell a corporate subsidiary to a corporate subsidiary of the other public company. Lo and behold, all the corporate subsidiaries of the first public company and all the corporate subsidiaries of the second public company are members of one single expanded group. The magic of downstream attribution to partnerships means that the first-tier subsidiaries of each of the public companies are attributed to the jointly owned partnership, and from there, under the magic of downstream attribution to corporations, each of those first-tier subsidiaries can be downstream attributed to any other of the first-tier subsidiaries. Thus, any of the first-tier subsidiaries can serve as the common parent of the fictionalized group. The contemplated sale of the subsidiary turns out to be a cross-chain sale within an expanded group even though ultimate ownership of the sold subsidiary absolutely changes in the real world from the shareholders of the seller public company group to the shareholders of the buyer public company group. If the buyer pays a note (or, more likely, borrows from its affiliate and pays cash), the transaction is caught by the proposed regulations.
Treasury speakers have stated informally that they did not intend that result and that they intend to fix it in the final regulations. One hopes that the final regulations will take a more tailored approach to the concept of ultimate ownership not changing. Downstream attribution to partnerships is notorious for leading to unintuitive results because of the lack of a minimum threshold for that attribution in section 318(a)(3)(A). It is probably inadvisable to exacerbate that problem by adopting low thresholds for downstream attribution to corporations, which the proposed regulations do by invoking section 304(c)(3)(A) in their definition of expanded group.
3. Coming and going. Third, a debt dividend's resemblance to an acquisition of sister stock for a note does not seem to ring true when an affiliate is joining or leaving an expanded group. For example, ultimate ownership does change when the sister has recently been acquired by the expanded group if the overall transaction, including that acquisition, is taken into account. Suppose that Parent and Sub wish to acquire unrelated Target for cash. Parent could lend funds to Sub, and Sub could acquire Target from Target's shareholders for cash, a transaction not covered by the proposed regulations. Note that Parent may have borrowed the funds that it lent to Sub from third parties, in which case the intercompany debt between Parent and Sub could be viewed as a push-down of Parent's third-party debt. Alternatively, if Parent buys Target from Target's shareholders and Sub then buys Target from Parent for a note, the note is caught by the proposed regulations. It is unclear why that transaction should be caught while the economically equivalent transaction should not. The two transactions seem to differ only in the order of the steps. The ultimate ownership of Target changes when the whole series of steps is considered.
By the same token, intercompany debt can be created when a member is on the way out of an expanded group. For example, suppose Parent owns Target 1 and Target 2. Unrelated Acquirer wishes to acquire Target 1 and Target 2. If Acquirer bought Target 1 and Target 2 for cash and a note, the note would not be recharacterized under the proposed regulations. If instead Target 1 acquires Target 2 from Parent for a note and then Parent sells Target 1 to Acquirer,45 it appears that the note is recharacterized as equity in the first instance (only to be later recharacterized when Target 1 leaves Parent's expanded group). The two cases differ in that Acquirer is the issuer of the note in the first case while Target 1 is the issuer of the note in the second case, but it is unclear why that distinction should matter vis-à-vis the concerns of the proposed regulations, because when the dust settles, the lender and borrower are not members of the same expanded group. The ultimate ownership of Target 1 and Target 2 does change in the transaction.
Moreover, in an even more classic case, suppose Subsidiary issues debt to Parent and Parent then sells Subsidiary to Acquirer.46 Although one could debate whether the debt distribution should be governed by section 301, the concerns of the proposed regulations would not appear to be implicated since this is a situation in which relatedness between the lender and the borrower ceases as part of the transaction. Acquirer could have issued debt to Parent or formed a subsidiary to issue debt to Parent and acquire Target -- transactions that would not have been covered by the proposed regulations because the debt in those cases is never between related parties.
4. Similarity to a debt dividend when something is acquired. Next consider scenarios in which all the affiliates are "old and cold" members of the expanded group and remain so. An acquisition of sister stock does in fact preserve ultimate ownership, but whether the transaction is similar to a debt dividend requires further analysis. The key feature of a debt dividend would seem to be that the debt is not lent in exchange for any property; no new capital is provided to the issuer. But when the issuer does in fact acquire something -- namely, stock in its affiliate -- the resemblance to a debt dividend depends on the facts.
Suppose Foreign Parent owns US Sub and Foreign Sub, and US Sub buys Foreign Sub from Foreign Parent for a note. Thus, Foreign Sub was outside the U.S. tax net before the transaction but is brought into it by virtue of the transaction because Foreign Sub is now owned by US Sub. In fact, Foreign Sub has become a controlled foreign corporation. The transaction does not seem similar to a debt dividend by US Sub because US Sub has acquired an asset, the stock in Foreign Sub. The transaction should not be troubling because the increase in U.S. interest deductions arising from the related-party debt owed by US Sub should be offset by U.S. taxable income generated by Foreign Sub (for example, dividends or subpart F income). However, Foreign Sub might not generate income that is taxable in the United States any time soon, since the timing of dividends from Foreign Sub can be managed and Foreign Sub might have no subpart F income. Subpart F income is, of course, taken off the table if the facts are changed and not enough shares of Foreign Sub are acquired for it to become a CFC. So absent the proposed regulations, the transaction may be a concern from an earnings stripping point of view, even though Foreign Sub is brought within the U.S. tax net.
One could argue that the above transaction is similar to a debt dividend because section 304 would treat the note dividend as a section 301 distribution. But that argument puts the cart before the horse, as mentioned above. Section 304 may apply for the same reasons that the proposed regulations arguably should apply, but the application of section 304 is not itself a reason to apply the proposed regulations.
Indeed, the mechanic of section 304 to bifurcate a single exchange into two other transactions47 may suggest that other types of exchanges can be bifurcated. A bifurcation mindset could be invoked to find a debt dividend when any related-party debt is involved, even related-party debt issued for cash. Consider the simplest related-party debt for money borrowed: Parent owns Sub 1 and Sub 2, and Sub 1 lends $100 cash to Sub 2. That transaction could be deconstructed into a pair of distributions and contributions -- Sub 1 distributing $100 cash to Parent and Sub 2 distributing a $100 note to Parent, and then Parent contributing $100 cash to Sub 2 and contributing the $100 Sub 2 note to Sub 1. This is an unnatural way to view the transaction, but if one is of a mind to bifurcate exchanges, one could follow the lead of section 304 and bifurcate a simple related-party borrowing into transactions that include a debt dividend.
With the group member stock acquisition rule, the proposed regulations abandon testing for the defining economic feature of a debt dividend, namely, that the debt is issued not in exchange for anything. The similarity between debt dividends and acquisitions of sister stock for debt is not that no new capital is introduced; when a corporation buys sister stock for a note, the corporation acquires capital in the form of the sister stock. The similarity instead would appear to derive from the fact that ultimate ownership does not change.
Although the transaction may have some economic consequence, the government appears to have concluded that the rearrangement lacks enough economic consequence to justify the tax benefits associated with creating the intercompany debt. As the preamble says, those transactions "frequently have limited non-tax significance particularly in relation to the significant federal tax benefits that are generated in the transaction."48 Another example shows a simple end run around the debt dividend rule if the proposed regulations did not cover acquisitions of affiliate stock: Suppose Foreign Parent owns US Sub 1 and US Sub 2, and US Sub 1 buys US Sub 2 from Foreign Parent for a note. If the proposed regulations did not cover this transaction, it would be simple to issue related-party debt in the United States without in form paying a debt dividend. And if US Sub 1 were a newly formed entity that held no assets, the transaction would strongly resemble a debt dividend.
But once the proposed regulations sweep beyond debt dividends to cover value-for-value exchanges, it can be difficult to define where they should stop. They generally do not cover the issuance of debt to a related party in exchange for an asset, as distinguished from group member stock. It has been argued that they should.49 Treasury and the IRS may have decided against covering acquisitions of assets for a note because acquisitions of assets would be impeded in any event by various frictions and taxpayers would therefore unlikely substitute acquisitions of assets for debt dividends. For example, an acquisition of an asset from a domestic group member would be a taxable transaction,50 and an acquisition of an asset by a domestic group member from a foreign group member would bring the asset into the U.S. tax base.
Moreover, if acquisitions of assets were covered, a further principle would be needed to prevent the proposed regulations from applying to all related-party debt. Cash is a type of asset. Thus, a typical borrowing for cash is an issuance of debt in exchange for the acquisition of an asset: cash.
A key scenario in which the group member stock acquisition rule applies is when Foreign Parent acquires Domestic Target for Foreign Parent stock (or part stock and part cash) and Foreign Parent creates related-party debt owing from the domestic group to Foreign Parent at the time of the transaction using a mechanic in which the related-party debt is issued in exchange for the Foreign Parent stock. Example 2 in the proposed regulations illustrates that situation. There, a foreign parent issued its own stock to its domestic subsidiary in exchange for a note. The domestic subsidiary uses the foreign parent stock to acquire a publicly traded domestic target. The example treats the note as equity under prop. reg. section 1.385-3(b)(2)(ii) because the domestic subsidiary bought stock of a member of the group (the parent) in exchange for the note. That the hook stock (the foreign parent stock owned transitorily by the domestic subsidiary) is transferred to the public shareholders of the domestic target as part of the transaction does not take it outside the application of the group member stock acquisition rule.
Example 2 could well be a reorganization under section 368(a)(1)(B). Before the promulgation of the proposed regulations, Treasury and the IRS took steps to impede leveraging up a domestic target (or the domestic acquirer of a domestic target) with debt owed to a foreign parent in a so-called Killer B transaction. In Notice 2014-32,51 Treasury and the IRS announced that in a triangular reorganization, domestic acquirer debt issued to a foreign parent in exchange for foreign parent stock used to acquire the domestic target would generally be treated as distributed in a section 301 distribution out of E&P of both the domestic acquirer and the domestic target.52 Treaties would generally not reduce the withholding tax imposed on the section 301 distribution because they generally require a one-year holding period.53 Thus, the notice could lead to a substantial dividend under section 301(c)(1), subject to 30 percent withholding tax, a major deterrent to that transaction. Since the issuance of Notice 2014-32, taxpayers have generally not sought to place such intercompany debt in the context of a reorganization. But the notice did not address intercompany debt in an acquisition that does not qualify as a reorganization. Moreover, in some cases, reorganization treatment would not be a priority for the taxpayer because target shareholders would be required to recognize gain in any event under section 367(a).54 Thus, taxpayers who were prepared to bust reorganization treatment could put in place debt owing from the domestic group to the foreign parent at the time of the transaction without the withholding tax resulting from the application of Notice 2014-32.
Example 2 evokes the facts of NA General Partnership.55 In that case, ScottishPower PLC, a U.K. public company, sought to acquire PacifiCorp, a domestic public company, for ScottishPower stock. To accomplish the transaction, ScottishPower formed NA General Partnership and Subsidiaries (NAGP), a Nevada general partnership, and elected to treat NAGP as a corporation for federal income tax purposes. NAGP formed a merger subsidiary, which merged into PacifiCorp. NAGP issued debt to ScottishPower at closing in an amount equal to 75 percent of the value of the Scottish Power shares being transferred to the PacifiCorp shareholders. The debt was issued by NAGP "in consideration for ScottishPower transferring on behalf of NAGP" its shares to PacifiCorp shareholders. Thus, the taxpayer's theory was that NAGP, the domestic acquirer, issued debt to its foreign parent, ScottishPower, in exchange for ScottishPower shares, which were used as 75 percent of the consideration to buy the domestic target, PacifiCorp. The remaining 25 percent of the consideration was also ScottishPower stock but was not acquired by NAGP in exchange for debt of NAGP. Thus, the equity of the domestic target appears to have shrunk by 75 percent in the transaction. The domestic consolidated group sought to deduct interest on the newly issued related-party debt. The commissioner challenged those interest deductions.
The court held in favor of the taxpayer, finding that the alleged debt was debt for tax purposes. The court applied a traditional debt-equity analysis based on 11 factors. That the debt was issued in exchange for ScottishPower stock did not enter the analysis. Of the factors considered, the court examined the purported debtor's ability to obtain loans from third-party lenders, stating that this factor was relevant in determining the economic reality of the transaction. The court said that the test is not whether debt on the same terms and at the same price could have been borrowed from third parties: "The requirement of precise matching misses the mark." Rather, the test is whether the terms reflect a "patent distortion" of what the purported debtor could have borrowed. That third parties would have demanded a "slightly higher" interest rate did not prevent the debt from being considered debt.56
The court also examined whether the purported borrower was inadequately capitalized, noting that the appropriate leverage ratio for a company depends on the industry it is in. More cyclical or volatile industries can bear a lower debt-equity ratio than industries with lower business risk, such as utilities. The court concluded that NAGP was adequately capitalized even though its debt-equity ratio was significantly higher than that of PacifiCorp before the transaction. Citing Kraft, the court also rejected the commissioner's argument that the debt should not be respected because the taxpayer's motive in creating the debt was to obtain interest deductions. Consistent with current law, the court did not analyze whether the new debt of the domestic group reflected an allocation of existing group debt owed to third parties.
It is interesting to compare one's intuition about Example 2 in the proposed regulations with the first example discussed in Section III.B.3. There, the argument was that when related-party debt is established in an acquisition of a target, ultimate ownership of the target changed and the preamble's stated rationale for the group member stock acquisition rule therefore did not apply. The acquisitive all-cash transaction in the first paragraph in Section III.B.3 could be done in a way that did not cause the group member stock acquisition rule to apply (by having the parent lend to its subsidiary, which acquires the target for cash from the target's shareholders). The argument was thus that the rule should not apply if the steps are taken in a different order. The group member stock acquisition rule would not appear to be needed as a backstop to the debt dividend rule in that example.
Because the ultimate ownership of the target also changes in Example 2, it should arguably lead to the same conclusion that the group member stock acquisition rule does not apply. However, the member whose stock is acquired for debt in Example 2 is the foreign parent, not the target. Thus, one could argue that the preamble's rationale holds -- ultimate ownership of the member whose stock is acquired (the foreign parent) does not change. The foreign parent is a member of the same expanded group both before and after the transaction. Yet, this also seems too quick. The foreign parent's ultimate ownership does change by virtue of its issuance of shares that wind up in the hands of the domestic target's shareholders. The transaction represents more than a debt dividend from an old and cold subsidiary or a reshuffling of existing sister companies. One could argue that an acquisition of a target company is an appropriate time to reexamine the capital structure of the target. It may be that anticipated synergies arising from the acquisition justify a greater debt-to-equity ratio for a target than was appropriate for the target as a public company.
But the resemblance of Example 2 to a debt dividend remains. A transaction like Example 2 could be accomplished in other ways. For instance, the foreign parent could acquire the domestic target for foreign parent stock and then cause the domestic target to pay a debt dividend to the foreign parent. The end result of that structure differs from the structure in Example 2 only in that no new domestic holding company has been interposed between the foreign parent and the domestic target -- likely not a major substantive economic difference. Alternatively, the foreign parent could acquire the domestic target for foreign parent stock and then contribute the domestic target to a new domestic holding company in exchange for debt of the domestic holding company -- a section 304 transaction that would be treated as a section 301 distribution. These alternatives involve a section 301 distribution from the United States to the foreign parent, something that the Example 2 structure does not involve as long as the issuance of the debt in exchange for the foreign parent stock is respected. Yet they demonstrate the resemblance of Example 2 to an acquisition followed by a debt dividend by the domestic group.
The real difference between Example 2 and a scenario in which a domestic target is acquired by a foreign parent in an all-cash transaction in which related-party debt is created concerns the use of foreign parent stock rather than cash to acquire the domestic target. In an all-cash acquisition, one would expect that some portion of group debt is likely being pushed down to the domestic target. The transaction may well be a leveraged acquisition from the expanded group's perspective, not solely from the domestic group's perspective. The allocation theories discussed in Section II.C may support interest deductions on the intercompany debt in the all-cash example. Also, those theories may support the interest deductions in the all-stock example if the group as a whole is leveraged.
The analysis of Example 2 in the proposed regulations hinges on the existence of hook stock, and one might therefore think that the problem is the existence of hook stock (albeit transitory). But unlike the hook stock discussed in Section III.A, which is held indefinitely by a subsidiary, the hook stock in Example 2 is immediately used to buy a valuable asset -- namely, the stock of the domestic target. The value of this hook stock is amply demonstrated. The hook stock itself would not seem to be the rationale for the proposed regulations to cover Example 2.
The proposed regulations prevent debt dividends for tax purposes, but debt dividends are not the true target. Other principles are at play, which lead to the proposed regulations covering transactions that are similar to, or easily substitutable for, debt dividends. Targeting debt dividends leads to unexpected results.
1 Prop. reg. section 1.385-3 was issued as part of a package (REG-108060-15) along with prop. reg. section 1.385-1, which would allow the IRS to bifurcate purported debt instruments into part debt and part equity; prop. reg. section 1.385-2, which would require additional documentation for related-party debt; and prop. reg. section 1.385-4, governing debt among consolidated group members.
2Kraft Foods Co. v. Commissioner, 232 F.2d 118 (2d Cir. 1956).
3E.g., Andrew Velarde and Amanda Athanasiou, "Debt-Equity Reclassification Regs Usher In a Brave New World," Tax Notes, Apr. 11, 2016, p. 159 (practitioner stating that proposed section 385 regulations would "upset the way corporate business has been done in America for well over half a century"); Lee A. Sheppard, "Treasury Goes After Earnings Stripping, Hits Cash Management," Tax Notes, Apr. 18, 2016, p. 263 (reporting practitioners' concern about treatment of cash pooling); and letter of James Peaslee to the IRS (May 18, 2016) (characterizing the proposed section 385 regulations as "too broad" and likely to have "significant, immediate, and seriously adverse effects on conventional intercompany loans of cash," disrupting "routine group treasury functions").
4See infra note 8.
5 Prop. reg. section 1.385-3(c)(1).
6 Prop. reg. section 1.385-3(c)(2).
7 Prop. reg. section 1.385-3(b)(2)(ii). An exception applies for an exempt exchange, which covers some types of asset reorganizations. Prop. reg. section 1.385-3(f)(5).
8 Prop. reg. section 1.385-3(b)(2)(iii). The group member stock acquisition rule and the asset reorganization with boot rule backstop the debt dividend rule. The proposed regulations contain a further rule also intended as an ultimate backstop to the debt dividend rule. Under that further rule, with some exceptions, a debt instrument issued by a corporation to a member of the corporation's expanded group in exchange for property with a principal purpose of funding specific distributions to members of the expanded group or acquisitions from members of the expanded group is treated as stock. Prop. reg. section 1.385-3(b)(3). A debt instrument is generally treated as having been issued with such a principal purpose if it is issued during the 72-month period that begins 36 months before the distribution or acquisition. Prop. reg. section 1.385-3(b)(3)(iv)(B)(1).
9Kraft, 232 F.2d at 124.
11Id. at 125.
12 The E&P exception echoes the commissioner's willingness in Kraft to permit debt dividends up to the amount of accumulated earnings. The E&P exception exempts debt dividends only up to the amount of current-year E&P, however.
13Gregory v. Helvering, 293 U.S. 465 (1935).
14Kraft, 232 F.2d at 128.
15Id. (Clark, C.J., dissenting).
16Id. at 129 (Clark, C.J., dissenting).
17Id. (Clark, C.J., dissenting).
18 It has been argued that Treasury would have had authority to do so. Ronald E. Creamer and Isaac J. Wheeler, "A Framework for Testing Regulatory Authority," at 38 (Apr. 25, 2016) (working paper presented to the Tax Club).
19Moline Properties Inc. v. Commissioner, 319 U.S. 436 (1943).
20Id. at 440.
21Id. at 438, citing Moline Properties, 45 B.T.A. 647 (1941).
22See reg. section 1.482-2(a)(1) (district director may make appropriate allocations to reflect an arm's-length rate of interest on related-party debt); and reg. section 1.1502-13(g).
23See H. David Rosenbloom, "Banes of an Income Tax: Legal Fictions, Elections, Hypothetical Determinations, and Related-Party Debt," Tax Notes Int'l, Dec. 15, 2003, p. 989, at 997 (related-party debt is not "compensation for money lent"). See also Sheppard, supra note 3 (intercompany debt is "essentially unreal").
24 Rosenbloom, supra note 23, at 991.
25Id. at 997 (emphasis omitted).
26Id. at 999-1000.
27Id. at 1000.
28 Fleming, Peroni, and Shay, "Getting Serious About Cross-Border Earnings Stripping: Establishing an Analytical Framework," 93 N.C. L. Rev. 673, 695-696 (2015).
29Id. at 695.
30Id. at 709-711.
31 Treasury, "General Explanations of the Administration's Fiscal Year 2017 Revenue Proposals," at 2 (Feb. 1, 2016) (hereinafter "budget proposal"). See also Treasury, "General Explanations of the Administration's Fiscal Year 2016 Revenue Proposals," at 10 (Feb. 2015) (same proposal); Treasury, "General Explanations of the Administration's Fiscal Year 2015 Revenue Proposals," at 49 (Mar. 2014) (similar proposal). The Fleming, Peroni, and Shay proposal also resembles the method for allocating a foreign corporation's interest expense to income that is effectively connected with the conduct of a trade or business in the United States or to business profits of a permanent establishment under reg. section 1.882-5. See OECD Model Tax Convention on Income and Capital (2014), art. 7, cl. 2 (attributing profits to a PE as if it were "a separate and independent enterprise"); and U.S. Model Income Tax Convention (2016), art. 7, cl. 2 (same).
32 Alternatively, if the subsidiary's proportionate share is not substantiated, or if the subsidiary so elects, the U.S. subsidiary's deductions would be limited to interest income plus 10 percent of the subsidiary's adjusted taxable income as defined in section 163(j).
33 One could argue that a subsidiary should be entitled to its allocable share of interest deductions on debt owed by any member of the group to a third party regardless of whether the subsidiary itself is an obligor on intercompany debt. That approach would eliminate the administrative burden of creating intercompany debt and arguably lead to a fair sharing of the tax benefits associated with the interest expense. However, the approach of permitting interest deductions to an entity that is not a borrower at all is beyond the scope of this report. That approach is a step toward the "cost of capital allowance" system advocated by Edward D. Kleinbard, under which a corporation would be entitled to a deduction reflecting its cost of capital, regardless of whether its capital is denominated as debt or equity. Kleinbard, "Beyond Good and Evil Debt (and Debt Hedges): A Cost of Capital Allowance System," 67 Taxes 943, 947 (1989).
34Plantation Patterns Inc. v. Commissioner, 462 F.2d 712 (5th Cir. 1972). See Murphy Logging Co. v. United States, 378 F.2d 222 (9th Cir. 1967) (respecting the corporation, rather than the guarantors, as the issuer of debt).
35See Creamer and Wheeler, supra note 18, at 40 (debt versus equity status depends on the relationship between the parties, not whether debt is issued at formation or at a later stage).
36 Internal Revenue Code of 1954 section 312(a)(2).
37See H.R. Rep. No. 98-432, pt. 2, at 1203 (1984).
38 Eric Toder and Alan D. Viard, "Major Surgery Needed: A Call for Structural Reform of the U.S. Corporate Income Tax," Urban-Brookings Tax Policy Center working paper, at 23-24 (Apr. 3, 2014) (describing formulary apportionment as a method to allocate profits of a multinational group among jurisdictions).
39See Stephen B. Land, "Strange Loops and Tangled Hierarchies," 49 Tax L. Rev. 53, 58-59 (1993) (hook stock adds nothing to the parent's net worth but gives the subsidiary an interest in the parent's other assets).
40 Rev. Proc. 67-14, 1967-1 C.B. 591.
41 81 F.R. 20912 (Apr. 8, 2016).
42Id. at 20917.
44 Prop. reg. section 1.385-1(b)(3)(i).
45See Merrill Lynch & Co. Inc. v. Commissioner, 386 F.3d 464 (2d Cir. 2004).
46See Waterman Steamship Corp. v. Commissioner, 430 F.2d 1185 (5th Cir. 1970).
47 When it applies, section 304 generally treats an exchange of issuer stock for property as a transfer of the issuer stock to the acquirer in a section 351 transaction separate from the transfer of the property and deemed redemption of stock of the acquirer.
48 81 F.R. at 20917.
49 Michael L. Schler, "Altera and the Proposed Debt-Equity Regulations," Tax Notes, May 2, 2016, p. 680 (authority for the proposed regulations is weakened by not applying them to leveraged asset purchases).
50 If the acquisition were part of a reorganization, the asset reorganization with boot rule would cover it.
51 2014-20 IRB 1006.
52See Deborah L. Paul and Michael B. Love, "A Perfect Storm for Corporate Inversions: Causes and Responses," 43 Tax Mgmt. Int'l J. 671 (2014).
53See, e.g., Convention between the Government of the United States of America and the Government of the United Kingdom of Great Britain and Northern Ireland for the Avoidance of Double Taxation and the Prevention of Fiscal Evasion with Respect to Taxes on Income and Capital Gains, art. 10, cl. 3.a.
54 If specified requirements are met, reg. section 1.367(a)-3(c) turns off the usual gain recognition requirement of section 367(a) when a U.S. person transfers stock of a domestic corporation to a foreign corporation. Among other requirements, a substantiality test must be satisfied to the effect that the fair market value of the foreign corporation must be at least equal to the FMV of the domestic corporation. Reg. section 1.367(a)-3(c)(3)(iii)(A). Thus, if a foreign corporation acquires a smaller domestic corporation, shareholders of the domestic corporation will generally recognize gain on the transaction even if the transaction would otherwise qualify as a reorganization.
55NA General Partnership v. Commissioner, T.C. Memo. 2012-172.
56Id. at *14.
END OF FOOTNOTES