SPECIAL REPORT. tax notes. Untangling the BEPS Hybrid Mismatch Rules, Part 2. By Robert Cassanos

Transcription

1 Untangling the BEPS Hybrid Mismatch Rules, Part 2 By Robert Cassanos Robert Cassanos is a partner with Fried, Frank, Harris, Shriver & Jacobson LLP. Prior versions of this report were presented to the Tax Forum and the Tax Review. Cassanos is very grateful to Daniel Jacobson and Katherine Erbeznik for their tireless assistance in the preparation of this report. He benefited enormously from conversations with Danielle Rolfes, Douglas Poms, Michael Schler, Kimberly Blanchard, and William Cavanaugh about the base erosion and profit-shifting project in general and the hybrid mismatch rules in particular. All errors, omissions, and opinions are solely the author s. In this report, the second of three parts, Cassanos analyzes the OECD s final report on action 2 of its BEPS project from a technical and policy perspective, and he suggests ways to make the hybrid mismatch rules more effective and easier to comply with. (For Part 1, see Tax Notes, Aug. 3, 2015, p. 539.) Copyright 2015 Robert Cassanos. All rights reserved. Table of Contents IV. Selected Issues and Suggestions A. Making the Punishment Fit the Crime B. Investment Vehicles C. Timing Differences IV. Selected Issues and Suggestions A. Making the Punishment Fit the Crime Many of the difficulties involved in applying recommendation 1 arise because the primary response (the denial of deductions) produces results that appear neither rational nor administrable in many situations. As a result, application of the hybrid mismatch rules becomes much more complex than would be the case if there were a rule that simply disallowed preferences for payments made on hybrid instruments. In fact, this is exactly what recommendation 2, the defensive rule, and the exception for investment vehicles do. However, the primary response (when it is not trumped by another rule) is to deny the deduction to the payer. SPECIAL REPORT tax notes 1. Proportional or dollar for dollar? Because the hybrid mismatch rules strive to be neutral between competing tax authorities, base erosion can be defined only by comparison to what each jurisdiction s normal or ordinary rules of taxation are. It cannot be defined by reference to the rate differences between the two countries, by whether one of the two countries is a tax haven, or by whether one country s tax characterization or treatment of the payment is right or wrong. For example, if the payee jurisdiction grants a preference or exemption for a specified type of payment, base erosion must be measured by the degree of that preference in the payee jurisdiction. Of course, that degree can be measured two different ways: by proportion or by dollars. As long as the punishment is visited on the payee jurisdiction, these two measures should be identical. If the payee jurisdiction (Country A) taxes ordinary income at 40 percent and either exempts 60 percent of dividends or taxes them at 40 percent of the ordinary income rate, the difference here (ignoring graduated rates) will always be $24 of preference per $100 of dividend income, regardless of the rate at which income is taxed in the payer jurisdiction (Country B). Similarly, a rule (like the defensive rule and recommendation 2) that requires the payee in Country A to forgo its capital gain preference or exemption will always impose a cost of $24 per $100 on the payee of the hybrid payment. Assume that the measure of Country A s preference or exemption is defined in percentage or proportional terms, not in dollar terms. Again, as long as the rule being applied is the defensive rule or recommendation 2, application of the rule exactly offsets the reduction caused by the preference or exemption. However, suppose that the rule applied is the primary response, in which case, rather than disallowing the preference or exemption to the payee, the payer is denied a deduction. The cost imposed on the payer (Company B) will equal the cost imposed in the example above if, and only if, countries A and B have identical rate schedules. Yet, it is a principle of the hybrid mismatch rules that such symmetry is neither expected nor even necessarily desirable. Inevitably, at least some of the time, the ordinary rates in Country A and Country B will differ. When they do, application of the primary response to a deduction/no inclusion (D/NI) mismatch will always produce a different economic TAX NOTES, August 10,

2 outcome than would application of the defensive rule (or recommendation 2). In general, application of the primary response will result in a positive arbitrage for Company A and Company B (in the aggregate) if the payer jurisdiction imposes a lower tax rate on ordinary income than the payee jurisdiction, and a negative arbitrage if the payee jurisdiction has the lower rate. Suppose, for example, that the payer (Company B) is a U.S. corporation (ordinary rate of 35 percent) and the payee (Company A) is a resident of Country A, a country with an ordinary rate of 20 percent and a 60 percent dividends exclusion. In that case, application of the defensive rule will result in Company A paying $12 more in taxes than it would otherwise have paid. However, application of the primary rule would raise Company B s tax bill by $21 (60 percent of $35). By contrast, if the primary response had been applied on a dollar-for-dollar basis, the asymmetry would disappear and Company B s tax increase would be only $12 no matter which rule applied. The shift from the primary response to the defensive rule would change the identity of the party that bore the burden of the hybrid mismatch rules, but not the amount of the burden. 36 Using a proportional approach changes both the identity of the burdened party and the amount of the burden in a way that appears to have no relationship to the benefit of the hybrid mismatch to either party. However, even a dollar-for-dollar approach will not always produce consistent results if the payee jurisdiction has a higher tax rate than the payer jurisdiction. Suppose instead that Country B (the payer country) has a 10 percent ordinary rate, whereas Country A (the payee jurisdiction) has a 40 percent ordinary rate and a 60 percent exclusion for dividends. Imposing the defensive rule in that case results in a tax cost of $24 to Company A (60 percent of 40 percent), while the maximum penalty one could inflict on Company B a complete disallowance would result in a penalty of only $10. Therefore, unless Company B s unused deduction disallowances could be carried forward or spread to other deductible payments, the results are not comparable. So here is a case in which the dollar-for-dollar approach cannot (without resorting to complicated expansions of the hybrid mismatch rules) produce an equivalent penalty to that imposed by the defensive rule. (Note, however, that the proportional response creates an even greater discrepancy in the cost of the remedy in this example.) 36 As long as the payer jurisdiction s tax rate is equal to or higher than the payee jurisdiction. One might argue that a well-advised group of companies always strives to create deductions in high-tax jurisdictions and receive income in low-tax jurisdictions. Therefore, those taxpayers should not be overly affected by the denial of preferences for income in low-tax jurisdictions, because most of their base erosion is accomplished without resorting to hybrid mismatches, but rather by rate mismatches to which the hybrid mismatch rules do not apply. The dollar-for-dollar approach is therefore preferable because it produces distortions only in the unlikely event the group is moving income from a low-tax jurisdiction to a high-tax jurisdiction (and even then, only in extreme cases), whereas the proportional approach almost always produces a distortive result whenever the rates differ. 37 Thus, neither approach to implementing the primary response (dollar-for-dollar or proportional) is perfect in all cases. Both will produce arbitrages to some degree in some cases, but a dollar-for-dollar approach will usually be less distortive. 2. How is base erosion measured? Why is the base erosion from a hybrid instrument seemingly measured by comparing that instrument with a debt/ debt instrument and measuring the degree of preference or exemption granted by the payee country, rather than by comparing the hybrid instrument with an equity/equity instrument and measuring the deduction granted by the payer country? Although there is intuitive appeal for measuring the degree of preference or exemption rather than the deduction, the issue is not as easy as it may first appear. Assume that Company B issues a perpetual bond to its parent, Company A, in exchange for cash. Country B treats the perpetual bond as a debt instrument and allows Company B a deduction for interest paid. Country A treats the perpetual bond as an equity instrument and allows Company A a preference for payments on that instrument to relieve the burden of double taxation. The hybrid report is neutral as to whether a perpetual bond is more properly treated as debt or equity. So why, in applying the primary response, do we measure the degree of base erosion by examining the benefit of the preference granted by Country A (compared with a debt/debt instrument) as opposed to the benefit of the deduction in Country B (compared with an equity/equity instrument)? Even if the rule 37 It might be argued that on the contrary, the proportional approach is preferable because it is usually more punitive in nature and therefore a better deterrent to the creation of hybrid mismatches. But there is no indication that this was the intention. If it had been, the same treatment should have applied to participation exemptions, for example, when recommendation 2 applies. 650 TAX NOTES, August 10, 2015

3 cannot favor the arguably more sensible equity characterization, it should at least grant that characterization equal dignity. One explanation is that even though the definition of a D/NI mismatch refers to the proportion of the payment that is deductible under the laws of a jurisdiction [HR para. 43], rarely are non-dividend payments (other than return of basis) only partially deductible. The interest denial rules for the portion of interest on an applicable high-yield discount obligation instrument that exceeds the applicable federal rate plus 600 basis points are one example. 38 But these types of rules are uncommon, at least compared with rules that allow holders of capital assets some relief from ordinary taxation on their earnings. It could therefore be argued that the payer deduction is more or less a constant and that the measurement of the mismatch is more properly made on the payee side. Another explanation is that there is some subtle judgment at work here. It could be argued that the normal rate of tax is always the ordinary rate and that all income should be subject to tax at this rate, whatever it might be. The one exception is that as long as there is a corporate income tax system, some allowance must be made to relieve the burden of taxing the same dollar of income multiple times (which can be either more than once, or more than twice, depending on how much of a classicist one is). Because that relief is never needed for deductible items, it would seem logical to measure the extent of the mismatch by reference to the payee country s deviation from its own ordinary rate schedule. If this explanation is correct, it departs somewhat from the neutrality principles that animate the hybrid report. Although the hybrid report is not explicit on this point, the commentary will apparently clarify that proportional application of the primary response is the rule. As noted above, expanding the scope of recommendation 2 in an attempt to minimize the problems created by the primary response runs the risk of narrowing the application of the bottom-up related-party approach of recommendation 1. Another way to minimize the problems caused by the primary response would be to expand the category of investment vehicles subject to the special exception. Both of those remedies will likely be imperfect fixes at best. 3. How to allocate the result of the primary response to the guilty payee? The problems created by application of the primary response are not limited to the fact that the numbers are almost 38 See section 163(e)(5). COMMENTARY / SPECIAL REPORT always wrong. In many cases, application of the primary response results in clearly inappropriate burden shifting as well. Assume Company A owns all of Company B and finances Company B s operations with a hybrid instrument. Whichever variant of the hybrid mismatch rules is applied to this example, the cost is ultimately borne 100 percent by the combined Company A-B Group and its ultimate beneficial owners in proportion to their interests in Company A. Suppose instead that Company B has two 50 percent owners: Company A located in Country A, and Company C located in Country C. Company B is required to maintain a specified amount of regulatory capital. Country B regulators regard some types of perpetual bonds as good regulatory capital for this purpose because the principal will never come due, while Country B tax authorities regard those bonds as bona fide debt instruments and therefore allow a deduction for interest paid thereon. Based on an analysis of Company B s needs, Company A and Company C decide they will each subscribe for an equal amount of Company B perpetual bonds to enable Company B to satisfy its regulatory capital needs. Meanwhile, Country A grants Company A a 60 percent exclusion on the payments received under this hybrid instrument, 39 while Country C treats all payments on the perpetual bond as ordinary income. The analysis under the hybrid mismatch rules would appear to be as follows: The perpetual bond is a hybrid instrument. Assuming a proportional application of the primary response, the result should be that Company B suffers a disallowance of 30 percent of its interest expense on the perpetual bonds (60 percent of 50 percent). So far, so good. However, unlike the wholly owned example, Company C bears 50 percent of the burden of this disallowance, even though as to Company C there was no hybrid element in the arrangement Assume that the exclusion is not intended to relieve double taxation and thus is not subject to recommendation The rule can be even more irrational than that. Suppose Country C in fact permits a 100 percent participation exemption (to relieve double taxation) on the payments received by Company C on the perpetual bonds. Assume that each of countries A, B, and C have implemented the recommendations of the hybrid report as written. In that case, under recommendation 2, Company C would have its participation exemption entirely disallowed. But it would still appear to bear (absent some special arrangement) 50 percent of the economic burden caused by the interest disallowance to Company B as a result of the application of the primary response to Country A s exclusion. In other words, Company C must bear the cost of correcting 100 percent of its own hybrid mismatch, as well as 50 percent of the cost of Company A s hybrid mismatch. Company C may feel it has been hit with a double whammy. (Footnote continued on next page.) TAX NOTES, August 10,

4 The problem arises because corporations do not usually have what we would call special allocations, and there is no requirement that the manner in which their tax liabilities are borne by their shareholders have substantial economic effect. 41 Consider a case in which Company A and Company C hold their interests in Company B through a 50/50 partnership that provides the hybrid financing to Company A. Assuming tax counsel had been especially clairvoyant, the partnership agreement would presumably have included a clause disproportionately allocating payments received between Company A and Company C to make up for the detriment caused by the future enactment of the hybrid mismatch rules and the consequent application of the primary response in this case. Somewhat similar provisions are in fact common in some partnership agreements, but they are often limited to separately allocating withholding tax liabilities for remittances from Company B to the partnership so that those taxes are charged to the partner on whom they are imposed. The clause that would be needed to equalize the cost of the reversal of the hybrid mismatch between Company A and Company C in this case would be much more intrusive than the normal withholding tax clause (and presumably controversial). Perhaps, in some cases when there are multiple investors who can be grouped into particular hybrid mismatch rule treatments, some investments could be tailored to eliminate the burden-shifting issue through the use of alternative investment vehicles (AIVs), multiple blockers, kick-out provisions, or a combination of the above. This issue is not entirely theoretical. For example, although most structures involving Luxembourg holding companies are not necessarily structured to take advantage of D/NI mismatches, instruments such as preferred equity certificates and convertible preferred equity certificates may be treated differently in different countries, and in some cases that treatment may trigger application of the primary One could argue that in that case, Company C should be allowed to keep a portion of its preferential regime, presumably on the grounds that some of its payment was made out of after-tax cash flows. However, there appears to be no basis in the hybrid mismatch rules for such a rule, which puts Company C in a better position than it was in the first case, when there was no hybrid element in the transaction at all insofar as it was concerned. It would be snatching defeat from the jaws of victory to permit Company C to get a better result in this case because it was the beneficiary of a preference. The better approach would be to switch the priorities of the primary response and the defensive rule. But it appears to be too late to do that. 41 Unlike partnerships, which may make special allocations if they have substantial economic effect. See section 704(b)(2); and reg. section (b)(2). response. To the extent that it does, this could result in the type of burden shifting discussed above. As the first example makes clear, this issue is not caused by the existence of partnerships (although it often arises in cases involving partnerships because partnerships frequently serve as a basis for pooled investment vehicles). In fact, it may be easier to solve it in a partnership context because of the flexibility to make disproportionate distributions and correlative allocations of income. The problem arises because the primary response has allocation neutrality that is, it is indifferent to whether the burden of reversing a D/NI mismatch falls on the party that benefited from the mismatch. 4. Primary response and exempt payees. It is unclear what effect, if any, the hybrid mismatch rules in general, or the primary response in particular, have on payees that are tax-exempt entities in their jurisdictions of tax residence. For example, it is clear that a payment of deductible interest to a tax-exempt entity is not alone a target of the hybrid mismatch rules because that arrangement does not rely on a hybrid element to produce a D/NI outcome [HR para. 46]. However, it is less clear whether a payment under an instrument containing a hybrid element that would result in a D/NI outcome if hypothetically made to a taxpayer of ordinary status should trigger the primary response even when the payment is made to a tax-exempt entity. As noted in Part 1, 42 the concept of a taxpayer of ordinary status arises only in the definition of a hybrid mismatch for purposes of recommendation 1 43 : The terms of the instrument result in a mismatch in the tax treatment of payments made under the financial instrument if the mismatch would have arisen had the same instrument been directly entered into between resident taxpayers of ordinary status under the laws of their respective jurisdictions [HR para. 65, recommendation 1(3)(b)]. This passage appears to require the construction of a hypothetical transaction in which the person being tested in both the payer and the payee jurisdictions is a resident taxpayer of ordinary status. Presumably, a tax-exempt entity is not a taxpayer of ordinary status. Under this test, any payment that exploits a hybrid element to produce a D/NI mismatch in the hands of a taxpayer of ordinary status would appear to create a hybrid 42 Section III.A. 43 Notably, the concept of a taxpayer of ordinary status is neither defined nor referenced again in the hybrid report. 652 TAX NOTES, August 10, 2015

5 mismatch, even when the payment is made to a tax-exempt entity and therefore does not have the effect of reducing the tax burden on the actual parties to the arrangement. However, even if that construction is correct, only hybrid mismatches that result in D/NI outcomes can trigger the application of recommendation 1. As illustrated in Part 1, 44 the definition of the hybrid mismatch tracks the language found in element 2 of the definition of a hybrid mismatch arrangement that is, it appears to identify when a financial instrument exploits the differences in the tax treatment of the instrument under the laws of two or more jurisdictions. This point is illustrated in the discussion draft: The hybrid financial instrument rule should only apply to those mismatches that are attributable to a hybrid element in the instrument itself. While the hybrid element is reasonably easy to isolate and define in respect of hybrid transfers, there is more of a challenge in the context of basic hybrid financial instruments that simply rely on differences in the tax treatment of debt instruments. The recommendation set out in this Consultation Document is that the hybrid element should be identified by elimination; focusing exclusively on the terms of the arrangement between the parties rather than any particular feature of the taxpayer. The test recommended in this Consultation Document is whether the terms of the arrangement in isolation would have been sufficient to bring about the mismatch in tax outcomes under the laws of the relevant jurisdictions. This formulation might suggest the following basic test as set out at paragraph (e) of the recommendations: A financial instrument will be hybrid if the same arrangement directly entered into between resident taxpayers of ordinary status in their respective jurisdiction, would have been sufficient to bring about the mismatch. [Discussion draft para. 97.] While not as clearly stated in the hybrid report, the language preceding the suggested formulation tracks the language in paragraphs 49 and 62 of the report. In paragraph 49, the focus is on identifying the causal connection between the hybrid element and the mismatch. It says that when that causal connection is not obvious, a financial instrument should be treated as a hybrid financial instrument if its terms cause a mismatch in tax outcomes. Com- 44 Section III.A. COMMENTARY / SPECIAL REPORT pare that with paragraph 62, which states that it is impossible in practice to identify all financial instruments that have hybrid elements and therefore recommends that a financial instrument be treated as a hybrid financial instrument when its terms are sufficient to cause a mismatch in tax outcomes. This reading implies that the comparison between taxpayers of ordinary status is used to determine whether the terms of the instrument alone are sufficient to cause a difference in tax outcome. But this rule does not negate the need for an actual difference in tax outcome a D/NI outcome whereby there is a deduction by the payer but no corresponding inclusion in the ordinary income of the actual payee before recommendation 1 can become operative. The requirement that there be an actual D/NI outcome (that is, a situation that requires aligning) appears both in the text of recommendation 1 and in paragraph 62: Rather than targeting these technical differences [in the way different jurisdictions tax financial instruments] the focus of this Report is on aligning the treatment of cross-border payments under a financial instrument so that amounts that are treated as a financing expense by the issuer s jurisdiction are treated as ordinary income in the holder s jurisdiction [HR para. 61]. In this case, the focus is on whether the income received by the tax-exempt entity is included in that entity s ordinary income. The question then becomes what included in ordinary income means. Logically, it should not require that tax be paid, merely that the income be included by the taxpayer in its reckoning of income. This interpretation is explicitly adopted in the discussion draft. As quoted in Part 1, paragraphs 94 and 95 of the discussion draft provide the following: What is included in ordinary income? 94. Similarly the corresponding requirement that a deductible payment be treated as ordinary income by the recipient means that the payment has been incorporated into a calculation of the recipient s net income under the laws of the relevant tax jurisdiction. Ordinary income in this context means income that is subject to tax at the taxpayer s full marginal rate and does not benefit from any exemption, exclusion, credit or other tax relief applicable to particular categories of payments (such as credits for underlying tax paid by the issuer). Tax exemptions granted to entities such as charities are not caught by the hybrid financial instrument rule because such exemptions are attributable to a TAX NOTES, August 10,

6 particular characteristic of the taxpayer, rather than a particular category of payment. 95. The jurisdiction specific details of how the taxpayer calculates its income tax liability and at what tax rate should not generally affect the question of whether a payment has been brought into account in calculating that taxpayer s income. Thus a payment that is offset against deductible expenditure or losses that have been carried forward would, on this definition, be treated as having been brought into account. Similarly a payment will be treated as having been brought into account as ordinary income even if it is derived by an entity that is exempt from tax on ordinary income or subject to tax on such income at a nil marginal rate [Discussion draft paras. 94 and 95, emphasis added]. The first italicized statement provides that exemptions attributable to a taxpayer s status are ignored; it is exemptions attributable to the category of payment that trigger the hybrid mismatch rules. The second italicized statement makes clear that as long as a tax-exempt entity brings an item of income into account, it has been included in ordinary income, even though it has not been subject to tax. In U.S. terms, for example, whether the item is required to be shown on the tax-exempt entity s Form 990, it has been brought into account and so presumably has been included in the entity s ordinary income. 45 This must be the right answer. Otherwise, imposition of the primary rule would deny a deduction to a payer in circumstances in which there is no D/NI outcome (and no base erosion). Moreover, imposition of the defensive rule would require a sort of international unrelated business taxable income regime for tax-exempt entities. 46 By contrast, the natural meaning of included in ordinary income would align the results for tax-exempt payees whether the primary response or the defensive rule was applied. The entire hybrid mismatch rule is a rule about tax base, not the imposition of tax. Invocation of either the primary response or the 45 See page 9 of the Form 990 and the instructions thereto. Even if the tax-exempt entity fails to include an item on the Form 990 because that item is not required to be included, it has still been brought into account, just as a person who is exempt from filing a tax return because his income is too low to trigger a filing has nonetheless recognized the income in question. 46 This is because the defensive rule requires that the payment be included in ordinary income, which is a defined term capable of only one meaning. If it is defined to require payment of tax when used in the definition of D/NI outcome, the same definition must apply when the same defined term is used in the defensive rule. defensive rule depends on the status of the actual taxpayer in question, not the status of an ordinary taxpayer. Suppose that tax-exempt A makes a perpetual loan to taxable B. Assume that in Country B the loan is treated as a loan, while in Country A it is treated as equity and is eligible for a preferential rate of tax. However, under Country A laws, A does not in fact pay tax on any payments received under the perpetual loan. There are two ways to analyze this situation. Under the no harm, no foul approach, we examine whether there was in fact an exclusion or preference granted by Country A given the category of payments made under the perpetual loan. Although there was no tax paid, this would have been the case regardless of the hybrid nature of the instrument, because of A s tax-exempt status. Therefore, under the analysis above, the hybrid mismatch did not cause a difference in tax outcome or base erosion, because there is no difference in tax outcome or base erosion in this case. Under this approach, recommendation 1 does not apply. Suppose instead that the existence of a hybrid mismatch was sufficient to invoke recommendation 1 regardless of whether there was a D/NI outcome. In that case, the analysis would appear to depend on what would have occurred had A been a taxpayer of ordinary status under the laws of Country A. If there would have been a mismatch in tax outcomes in that case, the perpetual loan between A and B would be considered to give rise to a D/NI mismatch even though (in reality) Country A s tax base has not been eroded compared with what would have been the case had there not been a D/NI mismatch. Again, whether we care depends on which rule applies the primary response or the defensive rule. If the defensive rule applies, A must simply forgo a preference on which it was not relying to exempt the payment from tax, at least if one interprets this rule to require only an inclusion and not a tax payment (the same result as recommendation 2). This approach then also results in no harm, no foul. However, if the primary response applies, it appears that B must have its deductions on the perpetual loan denied as punishment for a mismatch that never occurred. And, as the example above demonstrates, there is no rule that necessarily ensures that the economic burden of this punishment is visited upon A, as opposed to some other shareholder of B. Similarly, if one were to interpret the defensive rule to require the enactment (or expansion of) a local UBTI regime, the punishment would be imposed on the tax-exempt entity s intended beneficiaries. This does not seem to be a sensible (or intended) result. 654 TAX NOTES, August 10, 2015

7 It would be helpful if the commentary affirmed the explanation found in the discussion draft concerning how a tax-exempt entity includes an item in ordinary income and confirmed that inclusion of that item does not require the payment of tax thereon. Alternatively, if the OECD s intention is to invoke a remedy when no base erosion occurs, contrary to the analysis above, the commentary should confirm that the application of the defensive rule does not require payment of tax and that for a hybrid arrangement involving a tax-exempt payee, the defensive rule would automatically trump the primary response. In other words, the result would be exactly the same as the result currently applied to investment vehicles. To summarize, making the primary response the principal rule and applying it on a proportional basis creates multiple problems. It changes the overall impact of the hybrid mismatch rules in most cases compared with the base case, and it will often shift that burden inappropriately among owners when there are multiparty hybrid (or partly hybrid) instruments. 5. Suggestions to correct primary response distortions. The following are suggestions for mitigating this problem. First, consider allowing countries to implement the primary response on a dollar-for-dollar basis. In most cases this would make the financial equivalent of the different recommendations equal, although the burden would fall on different parties. However, it appears unlikely that the commentary will endorse this approach. Second, consider expanding the scope of recommendation 2 and the investment vehicles exception, thereby broadening the application of recommendation 2 and the defensive rule, respectively, without narrowing the application of the bottom-up related-party rule. Third, consider allowing pooled investment vehicles to obtain certifications from their investors that they will be ineligible for or will be allowed to (and will) forgo claiming any applicable preference for an item of income or payment allocated to or received by them from the pooled investment vehicle. If all the investors gave the certification, the pooled investment vehicle would be able to apply the defensive rule without fear of deduction disallowance for their investees (unless the manager of the pooled investment vehicle knew or had reason to know the certification was false). Pooled investment vehicles would then be able to group their investors into responsive and recalcitrant groups so that if some investors would not or could not so certify, they could be hived off into parallel sleeves or AIVs with separate blockers, Luxcos, etc., that could continue to be subject to the primary response. It would be a condition to the certification process that the pooled investment vehicle s manager be required to notify the relevant tax authorities of the certifications so that the tax authorities in each payee jurisdiction could match payee declarations to the respective payee tax return. B. Investment Vehicles An equity/equity financial instrument can give rise to a D/NI mismatch, typically when a jurisdiction permits a dividends paid deduction to prevent a pooled investment vehicle from incurring an extra level of tax on its investment income. Examples in the United States would include regulated investment companies and real estate investment trusts. 47 In that case, application of the primary response would frustrate the policy of the payer jurisdiction to encourage the formation of those vehicles to encourage savings and investments. Accordingly, the hybrid report contains a rifle shot provision designed to turn off the primary response for hybrid payments made by those vehicles in appropriate cases and instead make the defensive rule the only response. Unfortunately, the rifle shot is so narrowly drafted that it could well miss most or all of its target. The investment vehicle exception reads as follows: (a) Regimes where the tax policy of the deduction under the laws of the payer jurisdiction is to preserve tax neutrality for the payer and payee The primary response in paragraph 1(a) should not apply to a payment by an investment vehicle that is subject to special regulation and tax treatment under the laws of the establishment jurisdiction in circumstances where: (i) the tax policy of the establishment jurisdiction is to preserve the deduction for the payment under the financial instrument to ensure that: the taxpayer is subject to no or minimal taxation on its investment income; and 47 Section 852(b) (method of taxation for RICs). Section 857(b) (method of taxation for REITs). By contrast, real estate mortgage investment conduits are special nontaxable vehicles that are not treated as corporations, partnerships, or trusts. Section 860A(a). As with partnerships, taxable income or loss of a REMIC is passed through to the REMIC interest holders. Section 860C(a). Unlike in partnerships, however, an interest in a REMIC is treated as a debt instrument, and any taxable income or loss that flows through to the interest holders is treated as ordinary income or ordinary loss. Sections 860B(a) and 860C(e). TAX NOTES, August 10,

8 holders of financial instruments issued by the taxpayer are subject to tax on that payment as ordinary income on a current basis. (ii) the regulatory and tax framework in the establishment jurisdiction has the effect that the financial instruments issued by the investment vehicle will result in all or substantially all of the taxpayer s investment income being paid and distributed to the holders of those financial instruments within a reasonable period of time after that income was derived or received by the taxpayer; (iii) the tax policy of the establishment jurisdiction is that the full amount of the payment is: included in the ordinary income of any person that is a payee in the establishment jurisdiction; and not excluded from the ordinary income of any person that is a payee under the laws of the payee jurisdiction under a treaty between the establishment jurisdiction and the payee jurisdiction; and (iv) the payment is not made under a structured arrangement. Further guidance will be provided in the Commentary on the circumstances where the exception will apply and the requirements for the application of this exception. The defensive rule in 1(b) will continue to apply to any payment made by such an investment vehicle. [HR para. 65, recommendation 1(5)]. The first point to note about this provision is that the hybrid report does not define the term investment vehicle. By way of cross-reference to another OECD report, recommendation 12 defines the term collective investment vehicle as being limited to funds that are widely-held, hold a diversified portfolio of securities and are subject to investorprotection regulation in the country in which they are established....[p]rivate equity funds, hedge funds or trusts or other entities...do not fall within the definition of [collective investment vehicle] set out in this paragraph [HR para. 122]. 48 It is apparent that most equity REITs, for example, do not fall under this definition, because they do not hold diversified portfolios of securities and are not subject to any special investor protection regime (unlike RICs). Assume, however, that 48 OECD, The Granting of Treaty Benefits With Respect to the Income of Collective Investment Vehicles, at para. 4 (Apr. 23, 2010). the term investment vehicle in recommendation 1 is not intended to be synonymous with the term collective investment vehicle in recommendation 12 but is intended to be broader, so that no special investor protections are necessary to fall within the rule. 49 Consider the following example: Country B is the United States (or has REIT rules identical to the U.S. REIT rules). A owns 25 percent of the stock of B, a non-domestically controlled REIT. B is selfmanaged and owns residential real estate located in Country B, which it leases to thousands of tenants. B pays ordinary dividends and occasionally sells properties from its portfolio, in which case it pays capital gain dividends. Assume there is a perfect match between A s share of B s gain derived from its U.S. real property interests and the capital gain dividend paid to A (which is not always the case 50 ). Accordingly, A is subject to U.S. withholding tax on ordinary dividends at 30 percent, or, if applicable, the lower rate imposed by the Country A treaty with Country B, but is subject to full tax (and withholding) on all capital gain dividends under the 1980 Foreign Investment in Real Property Tax Act. 51 Assume that in Country A, A is taxable as a corporation and is eligible for a preferential rate on all the aforementioned dividends. According to the analysis above, the stock of the REIT is a hybrid instrument, even though A s stock in B is treated as an equity instrument under the laws of Country B and the laws of Country A. This is because dividend payments made by B are treated as deductible in computing B s taxable income while not included in ordinary income in Country A. Should the REIT be subject to the primary response (and therefore have its deduction for dividends paid denied), or should A instead be subject to the defensive rule (and be ineligible for a preferential rate)? If A were entitled to a participation exemption, recommendation 2 would deny A that exemption rather than deny B s deduction for dividends paid, regardless of the degree of A s 49 It is not entirely clear that this is so. The term collective investment vehicle is only referenced once in the hybrid report s recommendations. It is possible that it was an oversight of the working group to use two separate terms. In any event, the commentary should clear up any confusion in this regard. 50 See New York State Bar Association tax section, Notice and Possible Administrative Guidance Addressing Sections 897(h)(1) and 1445(e)(6), Report No (Jan. 7, 2014). 51 Under FIRPTA, capital gains by nonresidents from the sale or exchange of U.S. real property interests, including the stock of a U.S. real property holding corporation, are generally subject to full U.S. tax, enforced through a 10 percent withholding tax on the total amount realized by the nonresident on the disposition. See sections 897 and Under section 897(h)(1), capital gains distributions from a REIT are also subject to FIRPTA tax. 656 TAX NOTES, August 10, 2015

9 ownership in B. If, instead of a participation exemption, a preferential rate is available to A, recommendation 2 (unless broadened) is inapplicable, and the only question is whether the primary response or the defensive rule applies. A preferential rate situation could arise, for example, if B paid a capital gain dividend and Country A treated that dividend as a capital payment subject to a preferential rate (in other words, the two jurisdictions treated the payment identically). Whether such a counterintuitive result should apply depends on the scope of the investment vehicle exception. The relevant requirement for application of the investment vehicle exception is as follows: the tax policy of the establishment jurisdiction is to preserve the deduction for the payment under the financial instrument to ensure that: the taxpayer is subject to no or minimal taxation on its investment income; and holders of financial instruments issued by the taxpayer are subject to tax on that payment as ordinary income on a current basis [HR para. 65, recommendation 1(5)(a)(i), emphasis added]. In this context, it is clear that the taxpayer refers to B, the REIT (which does not actually pay taxes), rather than its shareholders. Is B s rental income investment income? 52 B is self-managed and appears to be conducting substantial activities through its active self-management of residential rental real estate. A REIT, for example, can be a party to a tax-free spinoff, which requires the conduct of an active business It may be argued that the status of B s rental income is irrelevant because the REIT rules would also allow B a deduction for payments out of its investment income, but this does not seem to be what the rule intended. For example, if Country Z would allow a deduction for all dividends paid, whether from investing in treasury securities or manufacturing widgets, it would clearly not meet the intent of this rule, even if one could argue it met the letter of the rule. 53 Rev. Rul , C.B (ruling that for purposes of Code Sec. 355(b) s requirement that distributing and controlled corps. engage in active conduct of trade or business immediately after distribution, REIT will satisfy the condition solely by functions concerning rental activity that produces income qualifying as rents from real property within the meaning of Code Sec. 856(d) ). See also Richard M. Nugent, REIT Spinoffs: Passive REITs, Active Businesses, Tax Notes, Mar. 23, 2015, p. 1513; and Nugent, REIT Spinoffs: Passive REITs, Active Businesses, Part 2, Tax Notes, Mar. 30, 2015, p In several recent IRS private letter rulings, including LTR (generally thought to be issued to Penn National Gaming Inc.) and LTR (generally thought to be issued to CBS Corp.), the IRS ruled that the transactions in question qualified for nonrecognition under section 355, based in part on the taxpayer s representations (and financial information) that both (Footnote continued in next column.) COMMENTARY / SPECIAL REPORT Further, B s capital gain dividends are eligible for a preferential rate of taxation insofar as individuals resident in Country B are concerned. This is true even though Country A and Country B both have identical rules for taxing the income (that is, in both jurisdictions, capital gain dividends maintain their character on a flow-through basis) and A has already paid full FIRPTA tax on the dividend. Therefore, it appears that the primary response applies because of the failure to include the capital gain dividend in ordinary income in Country B. Although counterintuitive, this result seems consistent with one of the underlying policies of the hybrid report, which is to generally ignore source country withholding taxes. All that appears to be relevant in determining whether this special rule applies is whether the arrangement creates a tax preference for Country B holders of B s stock that gives them a benefit, regardless of whether there are any Country B holders. Because B gets a deduction at ordinary rates while its Country B shareholders may benefit from lower capital gain rates, this prong of the investment vehicle definition does not seem to be met. Although not as clear cut, similar issues may arise with RICs, which pay out capital gain dividends. (Unlike REIT capital gain dividends, they are seldom subject to FIRPTA.) For example, if the ability to pay a capital gain dividend is inconsistent with the rule, as it appears to be, many RICs would likely be at risk of having their dividends paid deductions disallowed to the extent the primary response applied to them. In summary, the rules as drafted do not appear to apply to any U.S. investment vehicles if interpreted narrowly. This does not seem to have been intended by the drafters of the hybrid report. It is difficult to deal with this issue, because REITs in particular appear to be expanding the types of activities they undertake. The following steps would be helpful: make clear that regimes such as RICs and REITs are not per se structured arrangements; expand the scope of recommendation 2 as broadly as possible (this would preempt the possible application of the primary response more than would otherwise have been the case); make clear that RICs, REITs, and other entities subject to similar tax regimes would be within the scope of the investment vehicle definition and that any income earned by those types of vehicles would be considered investment income for purposes of the rule; the distributing and controlled companies would continue an active trade or business after the distribution. TAX NOTES, August 10,

10 allow character to pass through to the extent it would for a domestic shareholder, without causing the entity to fail to qualify for the investment vehicle exception; and call off the primary response for any payment subject to withholding tax at source at either the full rate of tax applicable to a domestic shareholder or at the statutory rate of withholding. C. Timing Differences One of the biggest areas of uncertainty regarding the substantive scope of the hybrid mismatch rules is whether timing differences should be included and, if so, how they should be integrated into the rules. The original 2013 base erosion and profit shifting action plan described the hybrid report agenda thusly: Develop model treaty provisions and recommendations regarding the design of domestic rules to neutralise the effect (e.g., double nontaxation, double deduction, long-term deferral) of hybrid instruments and entities. This may include: (i) changes to the OECD Model Tax Convention to ensure that hybrid instruments and entities (as well as dual resident entities) are not used to obtain the benefits of treaties unduly; (ii) domestic law provisions that prevent exemption or non-recognition for payments that are deductible by the payor; (iii) domestic law provisions that deny a deduction for a payment that is not includible in income by the recipient (and is not subject to taxation under controlled foreign company (CFC) or similar rules); (iv) domestic law provisions that deny a deduction for a payment that is also deductible in another jurisdiction; and (v) where necessary, guidance on co-ordination or tie-breaker rules if more than one country seeks to apply such rules to a transaction or structure. [Emphasis added.] Thus, even at the initial stage, combating longterm deferral was listed as one of the objectives of the hybrid mismatch rules, although the description of what was to be done focused exclusively on what would be considered character (or permanent) differences. Next came the discussion draft, which dealt with timing differences as follows: The recommendation is not intended to impact on questions of timing in the recognition of payments. Thus a hybrid mismatch does not arise simply because the issuer accrues original issue discount over the term of the bond while the holder only recognises the corresponding income as redemption premium once the bond is repaid [Discussion draft para. 88]. This is a very clear statement that does not appear to be qualified in any way. The example chosen effectively says that the difference between cash and accrual accounting is irrelevant to the hybrid mismatch rules. While the statement does not tell the reader what period the drafters had in mind (if any), bonds are generally not short-term instruments. Ten years is certainly typical, and 30 years is not unheard of. Michael Schler, for example, understood the discussion draft to mean what it said namely, that the hybrid mismatch rules do not apply to timing differences. 54 Unfortunately, rather than confirming this clear and sensible statement, the hybrid report created confusion. It lays out the general rule regarding timing differences as follows: The hybrid mismatch rules should not generally interfere with hybrid entities or instruments that produce outcomes that do not raise tax policy concerns. In order to fall within the scope of the rule, the arrangement should result in an erosion of the tax base of one or more jurisdictions where the arrangement is structured. For example, the hybrid mismatch rule limiting D/NI outcomes should not generally address differences in the timing of payments and receipts under the laws of different jurisdictions [HR para. 50, emphasis added]. This approach seems to follow the discussion draft rather closely, save for the word generally. This approach is also consistent with the hybrid report s definition of the term D/NI outcome : A D/NI outcome is not generally impacted by questions of timing in the recognition of payments or differences in the way jurisdictions measure the value of money. In some circumstances however a timing mismatch will be considered permanent if the taxpayer cannot establish to the satisfaction of a tax authority that a payment will be brought into account within a defined period [HR para. 122, recommendation 12, emphasis added]. One way to read this language is that only timing differences that are per se permanent (that is, not timing differences) should be considered. However, 54 Schler, BEPS Action 2: Ending Mismatches on Hybrid Instruments, Part 1, Tax Notes, Aug. 11, 2014, p. 697, at p. 702 ( There is no D/NI solely on account of timing differences between the deduction in the payer jurisdiction and the income inclusion in the payee jurisdiction. ). 658 TAX NOTES, August 10, 2015