The Article explains the various structures that permit offshore funds with U.S. managers to originate loans without the fund or its investors being subject to any meaningful amount of U.S. tax. The Article also discusses a legislative change that would render these structures moot.
Banks are no longer in the business of making and holding corporate loans. In 1994, banks purchased 30% of all primary leveraged loans. In 2018, they purchased 3%. In the mid-1990s, U.S. and foreign banks funded more than 70% of institutional leveraged loans. In the first half of 2019, they funded less than 11%.
The banks have largely been replaced as holders of loans by nonbanks, like collateralized loan obligation vehicles (CLOs) and offshore funds.
Excludes revolving credit-only loans as well as left and right agent commitments (including administrative, syndication and documentation agent, arranger). Data through second quarter 2019.
Figure 1
Banks continue to arrange loans and sell them in an “originate and sell” model. Offshore funds and CLOs often buy these loans.
Figure 2
Increasingly, offshore funds are engaging in direct lending activity: The fund (through its manager in the United States)—and not a bank—will originate the loan directly and hold it. Between 2010–2015, 32% of all loans were extended by nonbanks. Altogether the U.S. private debt market was about $700 billion in 2018.
When a U.S. bank, or the U.S. branch of a foreign bank, originates and holds a loan, it pays U.S. income tax on its origination fees and on the continuing interest income it receives. When banks engaged in this business of making and holding loans, if an offshore fund could engage in the same business through personnel located in the United States and avoid U.S. tax on its interest income, it could undercut U.S. banks. This is the policy basis behind taxing offshore loan-originating funds that have U.S. managers. However, when banks operate under an “originate and sell” model, they are taxed at the same rate on the same activities as the managers for offshore funds.
Banks that originate and sell loans receive fees. These fees are taxed. Because most banks are corporations, the fees are subject to a 21% federal corporate tax, and an additional tax of up to 23.8% in the hands of individual shareholders, for a total maximum marginal tax rate of 39.8%. CLOs that buy these loans from banks (i.e., that buy on the “secondary market”) pay no tax; they operate under guidelines that allow U.S. law firms to write opinions to the effect that the CLO “will” be entitled to the “stocks and securities safe harbor” and “will not” be engaged in a U.S. trade or business or otherwise subject to U.S. income tax on a net income basis. Offshore funds often operate under similar tax guidelines but generally under a lower standard of comfort.
These offshore funds and CLOs pay their managers for the service of choosing the loans to buy. If the manager is a corporation for U.S. tax purposes (as many of the publicly traded managers increasingly are), the fees received by the manager and distributed to its shareholders are taxed identically to the fees received by the bank and distributed to its shareholders. If the fees are received by an individual owner of a pass-through manager, they are taxable at a 40.8% rate (including the 3.8% Medicare tax). Lending activity does not generate any meaningful amount of capital gains that could be the subject of a carried interest and benefit from the lower capital gains rate. Mostly, the fees are currently taxed.
When an offshore fund with a U.S. manager originates a loan, the fund pays its manager fixed and variable compensation for the service of choosing and originating the loan (in essence combining the fees received by a bank for originating a loan and the fees received by the CLO manager for choosing it). In all cases, the United States fully taxes all of the compensation income received for services rendered in the United States at ordinary income rates. There is no reason to tax anything else. The managers compete with the banks. The funds don’t compete with anyone.
Now that banks no longer hold the loans they originate, the policy for taxing offshore loan-originating funds no longer exists. This Article begins by making the case that a legal basis for taxing offshore loan-originating funds never existed in the first place.
The “stock and securities safe harbor” and the regulations that determine when lending activities give rise to taxable “effectively connected income” (ECI) were written at a time when U.S.-source interest income was subject to a 30% U.S. withholding tax. The 30% withholding tax was much greater than the net income tax that would have been imposed on a foreign lender had the foreign lender been engaged in a U.S. trade or business. For this reason, the section 864 regulations are biased against finding a U.S. trade or business and ECI. In 1984, with the introduction of the portfolio interest exemption, Congress retained the 30% withholding tax for foreign banks but flipped the law on its head for foreign nonbank lenders—after 1984, foreign nonbank lenders that avoid ECI pay less tax. But Congress left section 864’s and its regulations’ bias against ECI for lenders intact. So both the underlying tax policy and the underlying law favor exemption for nonbank foreign lenders.
Still, the Service has insisted that a foreign nonbank loan-originating fund with a U.S. manager is engaged in a trade or business and has ECI. Loan fund sponsors and their advisors never openly challenged this assertion. They did the next best thing. They created no fewer than ten different structures that permit offshore funds with U.S. managers to originate loans without any meaningful amount of U.S. federal income tax being paid by the funds or their foreign investors.
This Article explains these structures. Business development companies (BDCs) offer the best example. BDCs were designed by Congress to make loans. Congress treated them as regulated investment companies so they wouldn’t be subject to U.S. income tax. And, in 2004, Congress specifically and deliberately exempted the interest-related dividends of a BDC that makes U.S. loans from U.S. withholding tax so that foreigners could invest in them without any U.S. tax.
So without any doubt, a foreign investor can hire a U.S. manager to make U.S. loans on its behalf without any U.S. tax by doing so through a BDC. However, BDCs, as ’40 Act companies, are heavily regulated, which makes them expensive. BDCs pay more in legal fees than they do in tax. BDCs are also subject to restrictions, such as borrowing limits. These and similar restrictions were designed to protect U.S. investors. This is protection the foreign investors don’t need, don’t want, but have to live with. BDCs also cannot be used to originate loans that pay non-U.S. source interest without subjecting their foreign investors to U.S. withholding tax. So while many foreign investors originate U.S. loans through BDCs, many more do not.
The different techniques described in this Article to avoid tax on loan origination are not perfect. Many of them (like BDCs) are very complicated and expensive to implement. Although much of the expense of implementing these structures is attributable to fees paid to tax lawyers, it is wasted money.
The U.S. private debt market is approaching a trillion dollars. It is time for Congress to rationalize the tax treatment of foreign nonbank lenders. To this end, the Managed Funds Association and the Real Estate Round Table have proposed amending the securities trading safe harbor to include origination, with some caveats, and to provide that origination that qualifies for the safe harbor not be treated as commercial activity for purposes of section 892. This Article concludes by suggesting some changes to that proposal.
This Part lays out the tax framework under which offshore funds operate. For ease of presentation, the discussion assumes a “master-feeder” type of structure where the underlying fund is a partnership for U.S. tax purposes, some investors invest directly in that underlying fund, and others invest through a foreign “blocker”—an entity treated as a foreign corporation for U.S. federal tax purposes. Foreign blockers are usually organized in a low-tax jurisdiction such as the Cayman Islands.
If a fund is engaged in a U.S. trade or business for tax purposes, a foreign “blocker” of the fund, and any foreign investors that invest directly in the fund, will be subject to U.S. federal net income tax on the income that is “effectively connected” with that trade or business. The fund will have the obligation to withhold that tax and remit it to the Service.
In addition, a foreign blocker (or a directly-investing foreign investor that is treated as a corporation) will be subject to a 30% branch profits tax on its “dividend equivalent amounts,” which generally are a measure of the fund’s after-tax effectively connected earnings and profits that are not reinvested in the U.S. business and are deemed repatriated offshore. The resulting effective tax rate on ECI that is repatriated (or deemed to be repatriated) offshore by these corporate investors is 44.70%. The fund will be subject to an interest charge, and possibly to penalties, with respect to any taxes for previous years that were not withheld and timely paid. If the fund has not filed a U.S. tax return, the statute of limitation never tolls for the Service to assess these taxes, penalties, and interest. Finally, states may also tax the fund or its foreign investors.
A fund and its foreign investors will be subject to U.S. federal tax on a net income basis only if, in the first instance, the fund is “engaged in a trade or business.” The following subparts explore when a fund is engaged in a trade or business.
The distinction between a person that conducts mere investment activities (i.e., an “investor”) and a person that is engaged in the business of trading (i.e., a “trader”) is that a trader’s activities are “frequent, continuous, and regular,” whereas the investor’s activities are “more isolated and passive.” Lending funds never experience the turnover rate that courts have established as the threshold for a trade or business for purposes of section 162.
Note, however, that the policies of section 162 are different from those of section 882, and for that reason the Service has suggested that different tests may be appropriate, and therefore that a lender that does not engage in sufficient activity to be engaged in a trade or business for purposes of section 162 may be engaged in a trade or business for purposes of section 882.
In Chief Counsel Attorney Memorandum 2009-010, the Service concluded that a foreign corporation whose U.S. agent originated loans on its behalf was engaged in a U.S. trade or business.
On the one hand, courts and the Service have held that active asset management activities (which arguably would include loan origination) do not cause an investor to become a trader. Under this view, even a fund that actively originates loans would not be engaged in a trade or business so long as the fund holds the loans for investment.
On the other hand, lending is a activity conducted by banks, financing companies, and other U.S. businesses, and the Service and courts have found taxpayers to be engaged in a trade or business for purposes of section 166 (relating to bad debt deductions) as a result of regularly and continuously making loans, even if the taxpayers retain and do not sell the loans. Moreover, some authorities have treated lending as a “service,” although courts generally have held that the activity of making loans for investment and not for purposes of resale to a customer or otherwise to earn a spread is not a service. Thus, there is no precedential authority under section 882 regarding whether making loans for investment constitutes a trade or business. So we really don’t know.
As part of the Foreign Investors Act of 1966, Congress enacted current section 864(b)(2)(A)(ii), which provides that a foreign person that is not a “dealer” is not treated as engaged in a trade or business within the United States as a result of “trading in stocks and securities for [its] own account,” regardless of whether the trading is done directly by the foreign person and its employees or “through a resident broker, commission agent, custodian, or other agent,” and regardless of whether the foreign person’s employees or agents have discretionary authority to effect the trades. Section 864(b)(2)(A)(ii) is known as the “securities trading safe harbor.”
The securities trading safe harbor applies to foreign persons that limit their U.S. activities to “trading in stocks or securities.” Regulation section 1.864-2(c)(2) expands the permissible activities to include buying, selling (whether or not by entering into short sales), or trading in stocks, securities, or contracts or options to buy or sell stocks or securities, on margin or otherwise, for the account and risk of the taxpayer, and any other activity closely related thereto (such as obtaining credit for the purpose of effecting the buying, selling, or trading).
For these purposes, the term “securities” is defined to include “any note, bond, debenture, or other evidence of indebtedness.” This definition of “securities” appears to include loans and, when the regulation was issued in 1967, the term securities was understood to include loans. In other places in the Code and regulations, the term “securities” clearly includes loans. Also, in 1967, there was no secondary market in loans. Therefore, if a foreign person held a loan, that person probably originated the loan.
However, the definition of securities under Regulation section 1.864-2(c)(2)(i) is substantially similar to the definition of “security” under section 2(1) of the Securities Act of 1933 (the “1933 Act”), which provides that, “unless the context otherwise requires,” the term “security” includes “any note . . . bond, debenture, evidence of indebtedness . . . or right to subscribe to or purchase, any of the foregoing.” Notes evidencing loans by commercial banks for current operations are not “securities” under this definition. Does this support the Service’s argument that the term “securities” for purposes of section 864 does not include loans?
No. The purpose behind the 1933 Act is very different than the purpose behind section 864. The 1933 Act was designed to protect investors that invest in securities. Loans are not securities for purposes of the 1933 Act because Congress did not intend to protect banks and other institutional investors that make loans.
Nevertheless, the Service asserts that the securities trading safe harbor is not available to a foreign person that makes “personal, mortgage, industrial, or other loans to the public.” The next section does its very best to defend the Service’s position based on the language of the regulations and their historic context.
An argument could be made that the term “trading,” as used in section 864(b)(2)(A)(ii) and Regulation section 1.864-2(c)(2), implies activity of a passive nature in the secondary markets and is semantically distinguishable from “originating” or “lending.” Analogous authorities under the defunct Financial Asset Securitization Investment Trust (FASIT) rules, the publicly traded partnership (PTP) rules, and the special ECI rules applicable to foreign persons that are engaged in the “active conduct of a banking, financing, or similar business” in the United States could be read to imply that passive “trading” does not include lending or origination activity.
But Regulation section 1.864-2(c)(2) specifically provides that the securities trading safe harbor extends to “the effecting of transactions in . . . stocks or securities,” including buying and selling (without evident restriction), and defines “securities” to include notes and other evidence of indebtedness (and not merely traded bonds). The statutory reference to “trading” thus is better understood as shorthand for “engaged in a trade or business of buying and selling securities,” and not as implying passive activity.
Moreover, the analogous authorities mentioned above are inapposite.
First, FASITs were permitted to issue “regular interests” that were automatically treated as indebtedness for U.S. tax purposes. To prevent abuse of this “debt safe harbor,” Congress tightly constrained the activities of FASITs, prohibiting them from selling assets to make a profit and rendering them even more passive than a mere investor. Since funds do not benefit from any such debt safe harbor, the FASIT prohibition on loan origination does not appear to be relevant in determining whether origination disqualifies an offshore fund from relying on the securities trading safe harbor.
Second, Congress intended the “qualifying income” safe harbor under the PTP rules to be limited to partnerships that engage only in “activities commonly considered as essentially no more than investments.” The securities trading safe harbor clearly extends beyond these bounds.
Finally, the special ECI rules applicable to foreign persons that are engaged in the “active conduct of a banking, financing, or similar business” in the United States apply only if, in the first instance, “at some time during the taxable year the taxpayer is engaged in business in the United States.” This suggests that a taxpayer might engage in U.S. financing activities without being engaged in a U.S. trade or business.
However, two more appropriate analogies to the language in section 864(b)(2)(A)(ii) can be found in the rules relating to tax-exempt organizations, RICs, and REITs. Tax-exempt organizations generally are not treated as engaged in an “unrelated business” if they earn interest from loans that they make to unrelated persons. RICs and REITs, which were designed to encourage capital investment rather than to conduct “active” commercial operations, can make loans without being subject to corporate-level tax. If Congress was so concerned about nontaxable entities unfairly competing with U.S. banks, why did they empower RICs and REITs and tax-exempts to do just that?
The securities trading safe harbor does not apply to dealers. Regulation section 1.475(c)-1(c)(1) provides that a taxpayer that regularly makes loans to customers in the ordinary course of a trade or business of making loans is treated as a dealer for purposes of section 475 unless the taxpayer “engages in no more than negligible sales of the securities so acquired.” This rule could be read to imply that origination activities could cause a fund to be treated as a dealer for purposes of the securities trading safe harbor.
But, as discussed below in Part V.D, Regulation section 1.864-2(c)(2)(iv) contains a special definition of “dealer” for purposes of the securities trading safe harbor, and this definition does not include making loans to customers. Moreover, the purposes of section 475 and section 864(b)(2)(A)(ii) are very different. Thus, a fund that originates and sells loans could be treated as a dealer under section 475 and not as a dealer under section 864(b)(2)(A)(ii). Also, funds rarely sell loans.
Congress’s policy rationale for excluding dealers from the securities trading safe harbor was to prevent foreign persons from unfairly competing with U.S. commercial banks. Nevertheless, while the legislative history to section 864(b)(2)(A)(ii) clearly indicates a concern with foreign dealer competition, neither the statute nor its legislative history contains any evidence that Congress was concerned about foreign competition with U.S. lenders. To the contrary, when the securities trading safe harbor was enacted in 1966, Congress and U.S. banks were eager to encourage foreign investment in the United States. Far from lobbying for greater protection from foreign competition, U.S. banks sought to retain the exemption from U.S. withholding tax on bank deposits to continue to encourage an inflow of U.S. dollars from abroad; they were unconcerned about competition from foreign lenders.
Moreover, in 1966, U.S.-source interest was subject to 30% withholding, as the portfolio interest exemption had not yet been enacted. Foreign banks and financial institutions lobbied Congress to treat all of their income (including income from trading in stocks and securities) as ECI because their net income tax rate (after deductions) would have been significantly less than the 30% gross-basis withholding tax. Thus, in 1966, many foreign lenders would have welcomed a U.S. corporate income tax on all U.S.-source interest income.
Notably, when Congress enacted the portfolio interest exemption in 1984, it retained the withholding tax only for foreign banks making loans in the ordinary course of their trades or businesses (but eliminated the withholding tax for nonbank foreign persons engaged in a financing or similar business), which suggests again that Congress was not concerned about competition from nonbank foreign lenders.
In any event, funds that originate loans from within the United States do not generate any goodwill in the United States because any goodwill generated in connection with their origination activities inures to their U.S. sponsors (and not to the funds). U.S. sponsors are compensated for their services at arm’s-length rates. Historically, corporate tax has been imposed only on entities that generate goodwill, but not on those that don’t (such as RICs, REITs, REMICs, and PTPs). Thus, arguably, even if loan origination from within the United States can give rise to a U.S. trade or business, it should be subject to corporate tax only if the business generates goodwill, which loan-origination funds do not do.
In short, there was no policy in 1966 to impose a net income tax on foreign lenders. In fact, foreign lenders had to lobby for that treatment, and it was granted only in limited situations. Most foreign lenders were left outside the net income tax regime and were subject to 30% withholding on interest paid on U.S. loans until Congress enacted the portfolio interest exemption in 1984 for all nonbank lenders. And finally, the only benefit foreign investors in offshore funds receive from loan-origination activity is a normal return on their investment. The sponsor retains all the goodwill and the sponsor is taxed fully on its economic return.
The sponsors of offshore funds assume that a fund that originates loans is engaged in a trade or business in the United States and employ a variety of strategies to avoid U.S. federal income tax being imposed on the fund, its feeders, and its investors. The following Parts describe these strategies.
The first strategy to avoid U.S. tax is to have all substantive activities conducted by people physically located outside the United States. Nonsubstantive activities that are ministerial, clerical, or preparatory in nature (such as preparing periodic reports, setting up bank accounts, and arranging and approving electronic money transfers) may be conducted in the United States without causing the fund to be engaged in a U.S. trade or business.
Some investment managers have personnel who are physically located both within and outside of the United States. An open question is whether the U.S.-resident investment managers could follow tax guidelines to ensure that their activities comply with the securities trading safe harbor (discussed in Part V, below), but the foreign personnel could be allowed to originate loans. These are sometimes referred to as “hybrid” guidelines. (The foreign personnel would not be permitted to engage in dealer activities because the securities trading safe harbor is unavailable to a dealer in stocks or securities, regardless of where the dealer conducts its dealer activities.)
There isn’t any authority that explains whether a fund that employs hybrid guidelines would be engaged in a trade or business in the United States.
But, the policy rationale for excluding dealers from the securities trading safe harbor—namely, to prevent foreign persons from unfairly competing with U.S. commercial banks—is the same one that the Service uses to argue against offshore fund loan origination. Therefore, it seems possible that the Service could deny the availability of the securities trading safe harbor to a manager that engages in loan origination through personnel located outside the United States by analogy to Regulation section 1.864-2(c)(2)(iv), which takes into account both domestic and foreign activities in determining whether a person is a dealer. For this reason, many sponsors that follow hybrid guidelines will have their offshore funds conduct their loan origination through a separate non-U.S. subsidiary that is treated as a corporation for U.S. federal income tax purposes.
What if all substantive activities are conducted by personnel who are physically located outside the United States, except that the investment committee includes one or more U.S. resident people? Regulation section 1.864-4(c)(5)(ii) provides that U.S.-source interest and gain from securities derived by a foreign corporation that is in the active conduct of a banking, financing, or similar business in the United States is treated as ECI only if the securities giving rise to the income or gain are (1) attributable to the U.S. office through which the business is carried on and (2) were acquired as a result of, or in the course of, making loans to the public.
For purposes of Regulation section 1.864-4(c)(5)(ii), a security is deemed to be attributable to a U.S. office only if the office “actively and materially participated in soliciting, negotiating, or performing other activities required to arrange the acquisition of the . . . security.” Notably, Regulation section 1.864-4(c)(5)(iii)(b)(4) provides that exercising “final approval over the execution of the acquisition of [the security]” does not cause the security to be attributable to the U.S. office.
Thus, on first blush, it would appear that if the only activity of a U.S. member of an investment committee is to vote yes for the fund to purchase a security, that activity is not enough to cause the security to be attributable to the U.S. office. The U.S. member has not actively and materially participated in the solicitation or negotiation of their loan or performed other activities required to arrange the acquisition of the loan. The U.S. member has merely exercised final approval over the execution of the acquisition of the loan.
The wording of Regulation section 1.864-4(c)(5)(ii) is specific and peculiar. The U.S. office must have actively participated in soliciting, negotiating, or performing other activities “required to arrange the acquisition of” the security. Is the decision to make a loan an activity “required to arrange” its acquisition? The activities required to arrange the acquisition of a loan would seem to be signing the loan documents and other forms and wiring the loan amount to the borrower. One would think the decision to make the loan would have been specifically noted. On the other hand, the decision to make a loan is the most fundamental activity connected to the loan’s acquisition. So while the wording of Regulation section 1.864-4(c)(5)(ii) is odd, it is hard to believe that if a person in the United States makes the decision to originate it, the U.S. office is not engaged in an activity required to arrange its acquisition. But Regulation section 1.864-4(c)(5)(iii)(b)(4) specifically provides that the exercise of final approval over the execution of the acquisition of a security does not cause the security to be attributable to a U.S. office. Doesn’t this allow U.S. personnel to sit on the investment committee for loan origination?
The wording of this regulation is odd, but it seems to cut against the taxpayer. The words permit “final approval over the execution of the acquisition of [the security].” The words do not permit final approval of the security; they permit only final approval over the execution of the acquisition. Approval over “execution of acquisition” sounds like a nondiscretionary action that would cause the security to be recorded in the name of the fund, or it might be like the “pro forma approval of the loans” mentioned in Example (5) of Regulation section 1.864-4(c)(5)(vii).
So there is no clear rule that would permit someone who is physically located in the United States to sit on an investment committee. If one or more U.S. people on the investment committee have a senior role in the organization, are listened to and respected, or happen to be the deciding vote on a particular loan, a trade or business could be found.
A second strategy to avoid a trade or business in the United States is for the investment manager to follow U.S. tax guidelines designed to ensure that the fund will satisfy the securities trading safe harbor and not engage in origination. An offshore fund that satisfies the securities trading safe harbor does not have a trade or business within the United States.
Practitioners unanimously conclude that the acquisition of U.S. Treasuries or other publicly offered bonds at initial issuance, instead of on the secondary market, falls within the securities trading safe harbor. A publicly traded bond is not a loan to the public, so purchasing a publicly traded bond at initial issuance is not the type of activity that the Service asserted could give rise to a U.S. trade or business in Chief Counsel Attorney Memorandum 2009-010. Even in the absence of the securities trading safe harbor, purchasing a publicly traded security should not give rise to ECI.
A fund may purchase debt securities that are typically denominated as “notes” and are issued to accredited investors in a private placement transaction or Rule 144A offering. Practitioners are also comfortable that the purchase of these securities are protected by the securities trading safe harbor. The rationale here is similar to the rationale for publicly traded securities. In addition, even though reliance on the similar language in the 1933 Act is questionable in light of the different policies under each regime, private placement transactions and Rule 144A offerings are almost invariably “securities” under the “family resemblance test” used by the Supreme Court in Reves v. Ernst & Young to identify a security under the 1933 Act.
To further distinguish a fund’s acquisition of private debt securities at initial issuance from loan origination, most tax guidelines require that the acquisition be made pursuant to an offering document and that any negotiations of terms of the securities be conducted through an independent investment banker or private placement agent (and not directly with the issuer).
Finally, most tax guidelines impose limitations on the percentage of the aggregate face amount of a private debt issuance that a fund can acquire. These limitations are intended to ensure that the fund could not amend the terms of the private debt securities without significant consensus from other noteholders, and thus is not in a position analogous to a “lead bank” or an “arranger” of a loan.
The securities trading safe harbor is unavailable to a dealer in stocks or securities. Regulation section 1.864-2(c)(2)(iv)(a) defines a dealer for this purpose as:
a merchant of stocks or securities, with an established place of business, regularly engaged as a merchant in purchasing stocks or securities and selling them to customers with a view to the gains and profits that may be derived therefrom.
This definition is very narrow. It (1) applies only to a “merchant of stocks and securities,” (2) requires that the merchant have an established place of business, (3) requires that the merchant regularly be engaged as a merchant, (4) applies only if the merchant purchases stocks and securities (presumably from anyone), and finally (5) applies only if the merchant sells the stock and securities “to customers with a view to the gains and profits that may be derived therefrom.”
In Chief Counsel Advice 2015-01-013, the Service interpreted the term “dealer” liberally for purposes of section 864(b)(2)(A)(ii), concluding that a person can be a dealer if it acts as a middleman or market-maker, even if it does not have a special relationship with vendors or vendees that distinguishes them as “customers,” and even if it does not have any customers (in the traditional sense of the word) at all.
Tax guidelines that are intended to satisfy the securities trading safe harbor contain stringent prohibitions intended to ensure that a fund neither holds itself out, nor acts, as a middleman or market-maker.
Notwithstanding the arguments in Part II.D, above, because the Service asserts that the securities trading safe harbor is not available to a taxpayer that makes “personal, mortgage, industrial, or other loans to the public” (either directly or through an agent or partnership) within the meaning of Regulation section 1.864-4(c)(5)(i)(b), practitioners generally assume that lending activities within the United States constitute a U.S. trade or business. Thus, at the heart of a fund’s U.S. tax guidelines is a detailed set of restrictions intended to ensure that any loans acquired, held, and disposed of in accordance with the guidelines are viewed as secondary market securities that benefit from the securities trading safe harbor and are not treated as originated by the fund.
In most respects, these restrictions are similar among law firms and have three overarching principles:
These principles are predicated on the traditional view of a “lender” as the party that negotiates the loan, funds the loan, bears first risk with respect to the loan, and holds itself out as the lender. However, there are several differences among law firms’ tax guidelines implementing these principles. The guidelines described below generally represent those required by counsel to CLOs to give an opinion to the effect that the CLO “will not be engaged in a trade or business in the United States” and therefore are conservative. In certain cases, less conservative guidelines are noted.
Tax guidelines universally prohibit funds from negotiating the terms of a new loan because negotiation is a core lending activity. However, tax guidelines generally permit customary due diligence questions and comments regarding errors or inconsistencies in the loan documents, because due diligence is a preparatory activity that alone does not give rise to a trade or business, and is “closely related” to the purchase of a loan on the secondary market.
Many tax guidelines also permit the manager to share its criteria for purchasing loans with banks before being shown a particular loan. These criteria must be generally applicable (i.e., not specific to one investment) and therefore cannot be used to negotiate the terms of a loan. Merely sharing investment criteria is not negotiation, and it is closely related to the purchase of a loan on the secondary market.
Some tax guidelines permit the manager to communicate information about its appetite for specific investments, including by responding to questions from a lead bank about the manager’s appetite for a specific loan under different conditions (e.g., “would you be more interested in this loan if we convinced the borrower to add a fixed interest coverage charge?”). Other guidelines prohibit any discussion around a specific loan on the theory that there would be too much risk of negotiation being found.
Finally, most tax guidelines contain special rules that permit funds to negotiate the terms of a loan modification that is not treated as a “significant modification.” These are discussed in Part V.E.7, below.
CLO tax guidelines historically prohibited funds from acquiring loans that are not customarily acquired on the secondary market. These loans are effectively “club deals”—loans negotiated by a lead bank with one or more specific investors in mind. More recent CLO guidelines permit the purchase of club deal loans. Funds that operate at a lower level of comfort may be permitted to acquire club deal loans.
Tax guidelines universally prohibit a fund from being listed in the loan transaction documents as an original lender.
In addition, many tax guidelines impose a 48-hour waiting period between the closing and full funding of a loan by the original lender and the fund’s purchase of the loan. This requirement, commonly referred to as the “48-hour rule,” creates a formal distinction between the fund and the original lender.
There are variations on the 48-hour rule. First, at least one law firm’s tax guidelines require that the fund wait only until the second business day after the loan is originated. This rule would permit a fund to purchase a loan earlier than 48 hours after origination if the loan is originated on Monday through Thursday (but it would require the fund to wait more than 48 hours after origination if the loan is originated on Friday). Second, a small minority of law firms impose (1) a 48-hour waiting period when the fund purchases the loan pursuant to a legally binding commitment, (2) a 24-hour waiting period when the fund purchases the loan pursuant to a legally nonbinding (or “soft”) commitment, and (3) no waiting period if there is no commitment. This variation on the rule appears to be premised on the notion that if the bank is fully exposed to the borrower’s credit risk for even a short period of time, then a formal waiting period is unnecessary. Finally, some law firms have no 48-hour rule (or variation) at all.
“Revolving” and “delayed drawdown” loans are not fully funded on their closing dates, but instead provide for advances after closing. For these loans, the 48-hour rule generally requires a waiting period that begins only on the closing date (and not also on the full funding of each advance). However, as discussed in Part V.E.6, below, these loans often must be acquired together with an associated term loan, and the term loan generally must be subject to a 48-hour waiting period.
Finally, as discussed in Part V.E.7, below, a “significant modification” of a loan is treated for U.S. tax purposes as a new origination. Some law firms therefore require the 48-hour rule to “reset” after a significant modification, so that a fund cannot acquire a loan from a lender that has significantly modified a loan with a borrower until at least 48 hours after the closing of the modification.
In a typical “underwritten” loan syndication, the lead bank unconditionally commits to the borrower that it will make a loan. Funds, in turn, commit to purchase a portion of the loan from the lead bank.
If a fund unconditionally commits to acquire a portion of a loan from the lead bank before the bank is, itself, legally committed to make the loan, then the fund will bear all economic risk with respect to the loan from the date of the commitment. This gives rise to the concern that the lead bank has not assumed the economic risk that one would expect a principal to assume and therefore is merely acting as the fund’s agent in originating the loan. Thus, most CLO tax guidelines historically prohibited the fund from unconditionally committing to purchase a loan before the lead bank is, itself, unconditionally committed to make the loan. (If the lead bank first unconditionally commits to make the loan in an underwritten syndication, then the lead bank will bear economic risk because it will not have any assurance that it will be able to sell the loan. Most law firms would then permit the CLO to unconditionally commit to purchase the loan from the bank.)
Some law firms also permit a CLO to unconditionally commit to purchase a loan before the lead bank has unconditionally committed to make the loan, so long as the fund’s commitment is subject to a materially adverse effect (MAE) clause. Under an MAE clause, the CLO is not obligated to purchase its portion of the loan if the condition of the obligor, the loan, or the market has changed between the date of the CLO’s commitment and the date of the purchase in a manner that would have an MAE on the CLO. The MAE clause effectively ensures that the lead bank bears economic risk with respect to the borrower.
In a typical “best efforts” loan syndication, the lead bank commits to use best efforts to build a syndicate of lenders to make a loan but is not unconditionally committed to make the loan. In the years following the 2008–2009 financial crisis, a growing number of law firms have permitted a CLO to commit to purchase a portion of a loan that is the subject of only a best efforts syndication (and not an underwritten syndication) so long as, in general, (1) the loan facility has an aggregate face amount of at least $100 million, (2) the investment manager has no reason to believe that the loan would not be executed on the same terms regardless of whether the commitment was made, (3) the CLO acquires less than five percent of the face amount of the loan, and (4) the CLO and all other commonly managed entities collectively acquire less than 15% of the face amount of the loan. In these situations, even though the fund might be unconditionally committed to purchase its portion of the loan before the lead bank is committed to make the loan, it would be more difficult for the Service to successfully argue that the lead bank negotiated and funded the loan as agent for the CLO because the CLO’s commitment has no bearing on the terms on which the loan was executed. Rather, the lead bank’s role in these situations is akin to an underwriter in a private placement transaction or Rule 144A offering—in exchange for an “origination fee,” the bank helps the borrower structure and market the loan. In this sense, the lead bank is not acting as an agent of the CLO. But, in this case, the fund would be bearing the first economic risk.
Finally, tax guidelines usually permit funds to commit to buy nonloan private debt securities pursuant to a best efforts or other customary marketing procedure. As discussed in Part V.C, above, nonloan private debt securities do not give rise to the same origination concerns as loans.
A “revolving” loan is a line of credit that may be drawn, repaid, and redrawn throughout its term in the borrower’s discretion. A “delayed drawdown” loan permits one or more draws upon the satisfaction of specified contractual conditions (e.g., time constraints or the satisfaction of certain financial tests or both), but does not provide for redraws after principal is paid down.
Because revolving and delayed drawdown loans require a fund to advance funds directly to the borrower upon each draw, and because each advance generally is treated as a new debt instrument for U.S. tax purposes, revolving and delayed drawdown loans resemble loan originations. Accordingly, most tax guidelines impose additional restrictions on a fund’s acquisition of these loans.
First, most tax guidelines require all of the terms of any advance to be fixed as of the date that the fund acquires the loan (or to be determinable under a formula that is fixed as of that date), and prohibit the fund from having any discretion as to whether to make advances. If the fund did not negotiate the terms of any advance and has no relationship with the borrower that could affect whether and when an advance is made, then the fact that the fund advances funds directly to the borrower should not cause it to be engaged in a U.S. trade or business.
Second, most tax guidelines prohibit a CLO from acquiring a revolving or delayed drawdown loan earlier than 60 days after the loan closes (i.e., the loan must be “old and cold”) unless (1) the purchase gives the CLO access to an associated term loan and the CLO in fact acquires the associated term loan, (2) the purchase price reflects a discount or premium of at least two percent since the closing date, or (3) a more than de minimis advance (often defined as the lesser of $5 million and ten percent of the maximum outstanding amount on the loan) is outstanding when the fund acquires the facility. These requirements ensure that the revolving or delayed drawdown aspect of the loan is “closely related” (or merely incidental) to an outstanding debt instrument (i.e., an associated term loan or a prior advance) or that the outstanding debt instrument is clearly a secondary market investment.
Finally, most tax guidelines limit revolving and delayed drawdown loans to a small percentage of a CLO’s assets (typically no more than ten percent, determined by reference to both funded and unfunded commitments) and also limit the amount of any single revolving or delayed drawdown loan that a CLO can acquire (typically to no more than 25% of the loan’s aggregate principal amount). These limitations are intended to ensure that, even if the CLO’s advancing of funds directly to a borrower pursuant to a revolving or delayed drawdown loan by a CLO is considered origination activity, the activity would be closely related (or merely incidental) to the CLO’s primary activity of trading securities on the secondary market. Funds’ tax guidelines may have some variation on these rules.
Under Regulation section 1.1001-3, a “significant modification” of a debt instrument is treated as a retirement of the pre-modified loan in exchange for a newly issued loan. The new loan that is deemed to arise upon a significant modification is potentially an origination that could jeopardize a fund’s ability to rely on the securities trading safe harbor.
A modification that is not a significant modification under Regulation section 1.1001-3 is not a loan origination because it does not give rise to a new debt instrument for U.S. tax purposes. Accordingly, tax guidelines usually permit the fund to negotiate and vote on insignificant modifications.
Some tax guidelines also permit a fund to vote for or against a modification that was proposed by the obligor or a creditor that is unrelated to the fund so long as (1) at acquisition, the loan was performing and the fund did not expect the loan to default, (2) the fund does not negotiate the modification, and (3) the modification does not require an advance of additional funds. This situation is a typical byproduct of being an investor and does not give rise to a customer relationship.
In addition, many tax guidelines permit the fund to participate in a loan workout if, at acquisition, the loan was performing and the fund did not expect the loan to default, but the loan subsequently defaulted and the workout was necessary to protect the fund’s investment. Although typically not stated in the tax guidelines, many tax advisors also permit the fund to advance additional funds pursuant to a workout in this context.
By contrast, tax guidelines almost universally prohibit a fund from negotiating or voting on a loan modification if the fund acquired the loan with the expectation that the loan would be worked out.
Finally, a loan generally must be retested as debt or equity if it is significantly modified. Generally, a fund would not directly hold a loan if, for example, after the modification, the loan is treated as equity in a partnership that is engaged in a U.S. trade or business. In this case, the fund would generally hold the loan through a blocker. Nevertheless, Regulation section 1.1001-3(f)(7)(ii) provides that the financial condition of the obligor is generally disregarded for purposes of the retest.
Many investment managers have affiliates that originate loans for their own account. To avoid the possibility that an affiliate of the investment manager could be viewed as having engaged in loan origination as an agent of a fund, many tax guidelines prohibit the fund from acquiring loans that were originated by affiliates (or with respect to which an affiliate engaged in dealer or other financing activities) unless “information walls” are established between the affiliate and the manager that prohibit the exchange of information and profits. Information walls ensure that the affiliate and manager act as independent parties with respect to each other, and thus minimize the risk that the affiliate could be treated as an agent of the fund.
Because the internal operations of each investment manager are different, there are many variations of the “affiliate-originated loan” provisions. For example, some tax guidelines require an independent review board to confirm that an affiliate-originated loan is purchased at fair market value and impose other requirements to ensure that the affiliate and investment manager act at arm’s length with respect to the loan.
Case law generally provides that a foreigner is not engaged in a trade or business in the United States unless the U.S. activities are “considerable, continuous and regular.” In Pasquel v. Commissioner, the Tax Court found that a taxpayer was not engaged in a trade or business in the United States when the taxpayer entered into a “single and isolated” financing transaction in the United States. Thus, a foreign fund that engages in a limited number of loan originations or workouts may not engage in sufficient activity to be treated as engaged in a trade or business. Based on Pasquel, most advisors believe that a single loan origination in a taxable year that represents a small percentage by value of a fund’s portfolio should not cause a fund to be engaged in a trade or business for tax purposes. Some advisors analogize to the rules for financial or insurance business under the PTP rules of section 7704(d)(2) to conclude that, based on a single private letter ruling, five or fewer loans per year should not constitute a U.S. trade or business for purposes of section 882. Funds refer to each of those loans as “bullets.”
Funds generally rely on the safe harbor in section 864(b)(2)(A)(ii) (i.e., the safe harbor for trading for one’s own account) to ensure that they are not engaged in a U.S. trade or business. This Part explains why funds do not rely on the alternative safe harbor contained in section 864(b)(2)(A)(i) (i.e., the safe harbor for trading through a U.S. resident broker or other agent).
Section 211(b) of the 1939 Code provided that a foreign person was not treated as engaged in a trade or business within the United States as a result of “the effecting of transactions in the United States in stocks, securities, or commodities through a resident broker, commission agent, or custodian.”
In 1948, in Adda v. Commissioner, the Tax Court held that a nonresident alien who granted discretionary authority to his brother in the United States to trade commodity futures for his account was engaged in a trade or business within the United States. Under Adda, any grant of discretionary authority by a foreign person to an agent physically located in the United States to engage in the business of trading stocks, securities, or commodities on behalf of the foreign person caused the foreign person to be treated as engaged in a U.S. trade or business.
The 1954 Code retained much of the language of section 211(b) as part of section 871(c). The Foreign Investors Act of 1966 repealed and replaced then existing section 871 and enacted section 864(b)(2)(A)(i) to provide that a foreign person is not treated as engaged in a trade or business within the United States as a result of effecting transactions in the United States in stocks or securities “through a resident broker, commission agent, custodian, or other independent agent.”
There is no definition of “independent agent” for purposes of section 864(b)(2)(A)(i). However, as discussed in Part VIII.A, below, an independent agent in other contexts generally is permitted to exercise discretionary authority on behalf of its principal. Thus, this amendment, on its face, appears to permit a foreign person to grant discretion to an independent agent to effect transactions in stocks, securities, or commodities from within the United States.
However, in light of the absence of any apparent Congressional intent to override Adda and the explicit reference to discretionary authority in section 864(b)(2)(A)(ii) (which, as described above, was enacted at the same time), most tax advisors have concluded that section 864(b)(2)(A)(i) does not permit an agent in the United States to exercise discretion on behalf of a foreign principal. The regulations provide some support for this conclusion, and the Service follows this view.
Accordingly, funds that grant discretionary authority to their investment managers do not rely on section 864(b)(2)(A)(i).
Part IV.A above described how a manager whose personnel are physically located outside the United States may originate loans on behalf of a fund without causing the fund to be engaged in a U.S. trade or business. Part V described how a fund can avoid a U.S. trade or business if its manager follows guidelines to ensure that the fund qualifies for the “trading in stocks or securities” safe harbor.
Nevertheless, many funds are direct lenders whose managers have personnel located in the United States who negotiate with borrowers. For them, the prior strategies do not work.
Tax treaties generally provide that the business profits of a resident of a foreign country entitled to the benefits of the treaty are not taxable by the United States unless the profits are attributable to a permanent establishment in the United States or the income is specifically addressed in another article of the treaty.
Under tax treaties generally, a person has a permanent establishment in the United States if it has a fixed place of business through which it carries on business in the United States. A permanent establishment generally includes a place of management, a branch, or an office. However, a permanent establishment does not include a fixed place of business that is used solely for preparatory or auxiliary activities. Administrators and placement agents generally perform preparatory or auxiliary activities.
Under most tax treaties, a person is treated as having a permanent establishment in the United States if a dependent agent has the authority to conclude contracts on behalf of the person and habitually exercises that authority through an office or other permanent establishment in the United States. Most funds conclude that the converse is also true: A fund with a dependent agent in the United States that does not have the authority to conclude contracts on behalf of the fund does not cause the fund to have a permanent establishment in the United States.
Thus, if a fund manager has an office outside the United States and a U.S. subsidiary with personnel in the United States, and all of the investors in the fund are entitled to the benefits of income tax treaties, the U.S. subsidiary’s personnel could engage in significant and substantive activities while physically present in the United States so long as these personnel do not have the power to bind the fund, and the power to bind the fund is exercised exclusively by the fund manager personnel who are physically located outside the United States. Some fund managers that rely on this structure will have the offshore personnel actively participate in negotiations to prevent the U.S. personnel from having apparent authority to act on behalf of the fund.
Also, under most treaties, a person is not treated as having a permanent establishment in the United States by reason of the activities of an “independent agent” in the United States that acts in the ordinary course of its business as an independent agent. Thus, if a fund manager is an independent agent, the manager’s personnel in the United States can engage in any activity on behalf of the fund, including discretionary decision-making, without causing a treaty investor to be subject to U.S. federal income tax on a net income basis.
A “bring your own treaty” fund (BYOT fund) is a fund whose investors are entitled to the benefits of a tax treaty between the jurisdiction in which they are resident and the United States and use the treaty to avoid being subject to U.S. federal net income tax. Usually, BYOT structures rely on the independent agent rationale. For those that do, a BYOT fund requires:
These requirements are discussed in turn.
As mentioned above, under tax treaties generally, a person is not treated as having a permanent establishment in the United States by reason of the activities of an “independent agent” in the United States that acts in the ordinary course of its business as an independent agent.
Income tax treaties do not define the term “independent agent.” The technical explanation to many treaties, however, without specifically defining the term, provides that an agent is treated as an independent agent if the agent is both legally and economically independent of its principal. Under most tax treaties, if a term is not specifically defined in the treaty, the term has the meaning ascribed to it by the laws of the country to which the tax relates. Therefore, U.S. law is relevant in understanding whether a person is an “independent agent” within the meaning of a treaty.
In Taisei Fire & Marine Ins. Co. v. Commissioner, the Tax Court considered whether an agent was an independent agent under the U.S.-Japan income tax treaty. The relevant provisions in the U.S.-Japan income tax treaty are identical to those in most tax treaties.
In Taisei, the taxpayers were four Japanese insurance companies that were entitled to the benefits of the U.S.-Japan tax treaty. Fortress Re, Inc., a U.S. corporation, acted as an agent in underwriting and managing reinsurance on behalf of the Japanese insurance companies and had entered into separate management agreements with each of them. Each of the insurance companies set a “net acceptance limit” which was the maximum amount of net liability with respect to any single reinsurance contract that Fortress could accept on behalf of a member. Fortress could underwrite reinsurance contracts that were greater than the net acceptance limit in the agreements, but then Fortress would have to arrange for retrocessions of the excess over the net acceptance limit. Fortress received a percentage of the gross premiums from the policies, which more than adequately covered Fortress’ operating expenses. The four insurance companies did not assume equal amounts of liabilities. One consistently had 40% and as much as 44% of the total.
Under the terms of the contracts, any of the insurance companies could leave on six months’ notice. There was evidence that hundreds of insurance companies could have been clients of Fortress and that Fortress would have been successful in finding a substitute insurance company if necessary. There was also evidence that Fortress had rejected the request by another insurance company to become a client.
Fortress had enough leverage to force the Japanese insurance companies to share reinsurance profits with Carolina Re, a reinsurance company owned by the Fortress owners, even though Carolina Re was not as well known or as financially secure as other potential reinsurers. In addition, there was evidence that one of the insurance companies had sought to require Fortress to impose a gross acceptance limit, which Fortress had successfully resisted.
In Taisei, the court held that in order for an agent to be an independent agent, the agent must be both legally independent and economically independent.
a. Legal Independence. The Taisei court concluded that legal independence means that the agent is free to perform its duties for the principal in the manner that the agent chooses. On the other hand, if the agent is subject to detailed instructions or comprehensive control by the principal, the agent is not independent. The court concluded that Fortress was free to perform its duties in the manner that Fortress chose. Although the insurance companies set net acceptance limits, the court held that an independent agent may be given authority to perform only specific duties for the principal without those restrictions constricting control.
Although Fortress consulted with its clients and reported to them more regularly than required under the terms of its agreements, the court viewed these facts as “actions of a company seeking to maintain good relations with longstanding clients, rather than one seeking approval.” On this basis, the court concluded that Fortress was legally independent of the insurance companies.
Field Service Advice Memorandum 1998-217 lists 11 factors to determine whether an agent qualifies as an independent agent, the first four of which are indicative of legal independence:
(i) the agent’s activities on behalf of the foreign enterprise generally are free from detailed instructions and comprehensive control by the enterprise; (ii) the agent controls the decision of whether to procure the services of third parties in the agent’s conduct of its commercial activities, and where such services are required, the agent hires and supervises the third parties and bears the cost of their services without reimbursement from the foreign enterprise;
(iii) the agent’s operations are generally separate from business operations of the foreign enterprise, and the two generally do not share functions (other than stewardship function performed by the foreign enterprise in relation to its investment in the domestic subsidiary, see Treas. Reg. § 1.861-8(e)(4)); and (iv) the agent is generally not required to submit regular reports to the foreign enterprise concerning activities on its behalf.
Most fund managers exercise the type of independent control described in Taisei and satisfy these factors.
As an initial matter, the fact that a manager sponsors a fund and therefore in some sense is related to it should not prevent the manager from being an independent agent. Tax treaties generally provide that a domestic subsidiary does not, itself, result in a permanent establishment to a foreign parent company even when that subsidiary is managed by the parent company. Under the OECD commentaries, similar rules apply to entities that are commonly owned. Thus, without more, the relationship between a manager and the fund it organized does not preclude the manager from being treated as an independent agent of the fund for U.S. federal income tax purposes. In reality, a fund manager has much more independence with respect to a fund and its investors than a subsidiary would have with respect to its parent.
Most fund managers have complete discretion over the manner in which they manage their funds. More specifically, subject only to the broad parameters described in the offering documents of the fund and in the investment management agreement between the fund and the manager, the manager will have complete discretion in its purchase, sale, and management of investments on behalf of the fund, will not be subject to any detailed instructions or control from either the fund or its ultimate investors, and will have discretion to hire any third parties. This is much more discretion than a subsidiary would have with respect to its parent and is similar to Fortress’s level of control.
Finally, a fund manager’s business operations are usually separate from the investments of the fund (and its ultimate investors), and the fund manager does not share any functions with the fund (or the ultimate investors). In addition, although a fund manager may submit reports to its investors concerning the fund’s investments, and a fund manager may submit reports regarding the fund and developments materially affecting the fund’s portfolio to the fund’s investors, most managers do not report to the fund or its investors on its own activities on behalf of the fund.
Accordingly, whether the principal is the fund or the ultimate investors, most managers satisfy the test for legal independence established by Taisei, all (or virtually all) of the factors for legal independence laid out in Field Service Advice Memorandum 1998-217, and the guidance in respect of legal independence set forth in the OECD commentaries. Thus, most managers should be treated as legally independent for U.S. federal income tax purposes.
b. Economic Independence. In Taisei, the Tax Court held that economic independence means that the agent bears the risk of loss for its business (i.e., the agent is not guaranteed revenue and is not protected from loss if it is unable to generate sufficient revenue). There appear to be two general concepts reflected in this test. First, the manager should generally bear its own overhead and expenses so that the manager bears the downside risk should its overhead and expenses exceed the fees it receives (more on this later). Second the manager should not be dependent for revenue on a single or small group of funds or ultimate investors.
In Taisei, Fortress was not guaranteed revenues and bore its own overhead, and if any of the insurance companies terminated the relationship, Fortress would have had to find new clients. While the management agreements provided sufficient revenue to cover Fortress’s expenses, the court noted that Fortress had to acquire sufficient business to generate the revenue. The court additionally noted that a large mutual fund company that charges an annual management fee would not be considered dependent on its thousands of investors. Finally, Fortress had been paid over $27 million during the three years at issue, which the court said, “is not the kind of sum paid to a subservient company.” On this basis, Fortress was economically independent.
Of the 11 factors listed in Field Service Advice Memorandum 1998-217 relevant to the determination of independence, the final seven factors are indicative of economic independence:
(v) the entrepreneurial risk of the agent’s activities on behalf of the foreign enterprise is substantially borne by the agent; (vi) the agent is not an exclusive agent and, over a substantial period of time, the agent represents clients other than the foreign enterprise and affiliated companies; (vii) the agent regularly offers its services to the general public; (viii) the agent is compensated for activities conducted on behalf of the foreign enterprise at the market rate for similar services; (ix) the foreign enterprise does not reimburse the agent for its ordinary and necessary trade or business expenses; (x) the agent contributes a significant part of the materials, equipment, and resources necessary to conduct its commercial activities on behalf of the foreign enterprise; and (xi) the agent makes long-term investments in the facilities and other resources used in its commercial activities on behalf of the foreign enterprise.
Many managers bear all of the overhead expenses (including compensation for employees, rent, utilities, and other similar items) they incur in connection with providing investment management services to the funds they manage and therefore will bear any losses if their expenses for managing a fund exceed the fees that they receive from the fund. (However, see Part VI.A.1.c, below, regarding reimbursement for certain noninvestment services.) Moreover, the services that managers provide to a treaty fund are usually identical to the services that the managers provide to their nontreaty funds, and managers are compensated for their activities on behalf of a treaty fund at an arm’s-length rate.
The issue that arises for start-up managers is reliance on a single fund or small group of funds. The manager would be dependent on that fund, and if the manager were terminated, it may be difficult in hindsight to demonstrate that the manager had the same ability as Fortress to obtain new clients (i.e., organize new funds) or manage additional assets.
Advisors seek to establish that the manager of a treaty fund manages a total of at least four funds and will continue to do so for the life of the fund (like the four clients Fortress had) and that the compensation the manager receives from the treaty fund will not exceed 25% of the compensation received from functionally similar funds (including the treaty fund). A functionally similar fund is a fund whose professionals have the skills to manage the treaty fund and vice versa. For purposes of these percentage computations, compensation must come from functionally similar funds because an agent must act in the ordinary course of its business in carrying out its activities on behalf of the fund to be an independent agent. A manager who manages long-short equity funds (which is not usually a trading strategy that is functionally similar to a lending strategy) would not be acting in its ordinary course of business of managing long-short funds when it manages a debt-origination treaty fund, and therefore the compensation that the manager receives from managing the long-short equity funds should not count in determining whether it is economically dependent on the debt-origination treaty fund.
Typically, fixed income funds are functionally similar to treaty lending funds even if the fixed income fund does not originate loans. Portfolio managers who buy and sell debt instruments typically have all of the skills necessary to originate loans. Preferred stock and airplane leasing may also be considered functionally similar; often managers who manage preferred stock or airplane leasing funds are equally capable of managing lending funds.
Arguably, the 25% requirement sought by advisors is more conservative than the Taisei case would demand. In Taisei, although there were four clients, one of the clients represented 40% or more of the agent’s business. Ultimately, the test should really be: were the treaty fund to terminate the manager, would the manager still survive? However, in Taisei, the court was convinced that Fortress had the ability to find other clients, even if it were fired, and it is less clear that a manager that is fired by a large fund it manages would be able to quickly raise another fund. Nevertheless, many managers can demonstrate that if they were terminated by a fund representing less than 25% of their revenue, they would survive on the compensation generated by the other funds during their remaining lives.
c. Reimbursement for Nonmanagement Services. Historically, fund managers have borne all of the overhead costs incurred in providing investment management services to the funds that they manage. As a fund manager grows, it may hire its own accounting, legal, compliance, and other support services, which, if provided to a fund by a third-party, would be separately paid for by the fund. Recently, as fixed management fees have declined, it has become common for large investment managers to require the funds that they manage to reimburse them for the in-house costs they incur in providing these types of support services. Although the payment by a principal of an agent’s costs is generally considered an indication of the agent’s economic dependence, the Service has indicated that where it is industry practice for a principal to reimburse an agent for certain operating expenses, the payment of the expenses does not indicate the agent’s economic dependence. Accordingly, so long as it is industry practice in the market in which the manager participates for funds to reimburse an investment manager for support services and the reimbursement represents no more than a small percentage, say ten percent, of the management fees received by the manager from the fund in any year, then the reimbursements should not be an indication of the manager’s economic dependence on the fund (because the manager still bears the risk that its other costs of providing services to the fund will exceed the management fee).
Many modern treaties include provisions to the effect that the benefits of the treaty apply only to income that is derived directly or through entities that are fiscally transparent in the treaty investor’s jurisdiction, and it is possible that this principle applies equally for those treaties that do not so specify. After all, the goal of treaties is to eliminate double taxation and not permit double nontaxation. Income received by an entity that is a pass-through entity for U.S. tax purposes—but opaque for purposes of the tax laws of the treaty partner—might escape tax by the treaty partner. Accordingly, funds seek to structure the master fund and any feeder in a treaty structure as limited partnerships (which are usually treated as fiscally transparent) rather than limited liability companies or local law corporations (which are usually treated as opaque).
Although all vehicles in a treaty structure must be treated as fiscally transparent for purposes of the tax laws of the investors, non-U.S. vehicles may elect to be treated as corporations for U.S. federal tax purposes. This is most common in a BYOT feeder structure. The sole purpose of this check-the-box election is to permit the feeder—rather than each investor—to file U.S. federal income tax returns, file any state and local tax returns, and pay any state and local taxes. Filing a federal tax return is a prerequisite to claiming treaty benefits.
Although it may seem odd that the blocker—an entity that is treated as a corporation for U.S. tax purposes and does not itself qualify for the treaty—could claim treaty benefits, this makes perfect sense. While the blocker is opaque for U.S. tax purposes, it is transparent for purposes of the tax laws of the treaty investor, and that is what matters. Because the blocker is transparent for purposes of the tax laws of the treaty investor, the treaty investor will be reporting and paying tax on the blocker’s income in the investor’s home country. Since the blocker’s income and gain will be taxed in the treaty jurisdiction, it makes sense that the United States should not also tax it. A single tax is the goal of treaties.
In Donroy, Ltd. v. United States, four Canadian corporations were limited partners in two California limited partnerships that operated beverage and transportation businesses. The general partner operated the partnership from an office in San Francisco. The Canadian corporations claimed that they did not have a permanent establishment in the United States because, as limited partners, the general partner did not have authority to conclude contracts on their behalf or, alternatively, because the general partner was a commission agent, broker, or independent agent.
The district court held that, even though the general partner did not have the power to bind the limited partners to third parties and creditors under the Uniform Partnership Act (which had been adopted in California), the general partner did have general authority to control and conduct the business of the partnership by making contracts and otherwise.
The court also held that an independent agent should be analogized to a “special agent” or “independent contractor” as opposed to a “general agent.” The court defined a special agent as “one authorized to do some particular act or to act upon some particular occasion” and who usually acts in accordance with specific instructions. A general agent is an agent with the power to “transact all of the business of [the] principal.” The court held that a general partner is a general agent and therefore could not qualify as an independent agent. The court also held that because a partnership is an association of individuals and the general partner is a general agent of the limited partners, a U.S. office of the partnership constitutes the permanent establishment of the limited partners.
In Unger v. United States, Robert Unger was a Canadian dentist who owned a limited partnership interest in a Massachusetts real estate partnership. The general partner operated the partnership from its offices in Boston. Following Donroy, the Tax Court held that the general partner was a “general agent” and, therefore, not an independent agent. Accordingly, the general partner’s activities and office were attributed to the partnership and then to Unger.
Donroy and Unger are fairly criticized for their restrictive definition of an independent agent. As discussed above, an independent agent acts with legal and economic independence and therefore is not analogous to a special agent, as defined by the courts. There is nothing inconsistent with a general partner being both a general agent and an independent agent with respect to the limited partners. If a party would be an independent agent with respect to a treaty resident, it should not matter whether that party acts directly on behalf of that treaty resident or acts as the general partner in a partnership of which the treaty resident is a limited partner. Allowing the general partner to qualify as an independent agent is entirely consistent with the aggregate theory of partnerships.
Many advisors read Donroy and Unger expansively and prohibit a treaty fund from paying carry to a sponsor entity under the theory that the sponsor’s U.S. office would be attributable to the fund and, ultimately, the limited partners. These advisors also typically require the sponsor to receive a performance fee rather than carry. However, Donroy and Unger are technical cases, firmly rooted in the ability of a general partner to bind the partnership under the Uniform Partnership Act.
For a U.S. sponsor that wishes to receive its compensation as carry, rather than a fee, one technical solution to the Donroy/Unger technical issue is to form a non-U.S. entity that is treated as a corporation for U.S. tax purposes as the “noneconomic” general partner of the treaty fund, ensure that it does not have an office in the United States, and ensure that all of its activities are conducted by people who are physically located outside the United States. (A noneconomic general partner does not contribute any more than nominal capital to the treaty fund and is not entitled to any profits from, and does not bear any losses of, the fund.) Then the sponsor may form a “special limited partner” that is entitled to carry. The special limited partner may have an office in the United States. Because this special limited partner does not have the power to act on behalf of the fund (and in fact will not do so), its office should not be attributed to the fund under Donroy or Unger.
Advisors will often insist either that the general partner for a treaty fund be independent of the sponsor (which often means that it is a foreign corporate entity and a majority of its board of directors consists of partners at a local law firm) or that the investors have some limited rights to fire the manager without cause. One of two reasons is often given for these requirements.
First, in Taisei, in discussing whether Fortress, the U.S. agent, was economically independent, the court noted that any one of the Japanese insurance companies could cancel its agreement on six months’ notice. Therefore, some advisors feel it is important that the investors also be able to cancel the management contract with the fund manager (i.e., fire the manager without cause upon the vote of a specified percentage of investors).
Second, some advisors worry that if a fund is locked into a contractual relationship with the manager, then the fund would be treated as the “alter ego” of the manager, which would either affect the independent agent status of the manager or otherwise interfere with the investors’ ability to claim treaty benefits.
These concerns are misguided, and the solutions undermine the strength of the independent agent argument.
First, in Taisei, in holding that Fortress was not economically dependent on the Japanese insurance companies, the court noted that Fortress was not guaranteed any revenue, any of the clients could leave on six months’ notice, and if one of the clients did leave, then Fortress would have to find a replacement to subscribe to that client’s share of reinsurance contracts.
However, had the insurance companies been bound to perpetual contracts, Fortress would have been less—rather than more—economically dependent on any one of them because it would have been certain to receive revenue from the others. So the fact that any of them could terminate at any time did not make Fortress any less dependent on any single insurance company and instead made Fortress more dependent on the others.
Second, an important reason that the Japanese insurance companies won in Taisei was because Fortress could exercise leverage over them. The court emphasized that Fortress was well compensated and that it could dictate terms to the insurance companies. If Fortress could prevent the Japanese insurance companies from firing it, Fortress’s case would have been even stronger.
The underlying rationale for the rule that the actions and office of an independent agent are not attributable to the principal is based on the premise that if an agent (such as the manager) is operating independently and is not economically dependent on a principal (such as the fund), then the agent can set a fair market value for its services, and its home jurisdiction will collect the proper amount of taxes for the services it provides. Since the agent’s home jurisdiction receives the proper amount of taxes, there is no need to collect more tax from the principal. In contrast, if the agent is subservient, then the principal can dictate terms and pay the agent less than fair market value for its services, and the agent’s home jurisdiction would then be justified in seeking additional taxes from the principal.
The fact that Fortress was well paid and could dictate terms to insurance companies indicated that it was being appropriately paid for its services.
Fund managers are quite analogous to Fortress. They act independently and not under the direction of investors, there is a market price for the services they provide, and if a single fund does not represent a significant portion of their revenue, they are not dependent on that fund for survival. (In fact, the Tax Court suggested that it may not be appropriate to view the fund as the principal, but look through to the individual investors. If this is correct, then a manager managing its first fund with as few as four investors may qualify as an independent agent.)
As part of the leverage that fund managers exercise over their investors, managers will often prevent the investors from firing them unless there is cause. A court following the Taisei reasoning should find that helpful in demonstrating that the manager is truly independent and receiving market compensation for the services it provides. The advisors that insist that investors be given the power to fire the manager without cause, or that the manager relinquish some control over the general partner to independent board members, undermine the leverage that a manager normally enjoys, which the Taisei court found persuasive in determining that Fortress was an independent agent.
Even for managers with “seed investors” who do exercise some economic leverage over the manager, the United States generally collects the proper amount of tax without having to resort to taxing the investors. Seed investors will often exercise their leverage by demanding a portion of the equity in the manager. Foreign seed investors will then be taxable in the United States on their share of the fee income of the managers (although they may escape tax with respect to their share of any carry).
The “Business Profits” article of many tax treaties provides that the business profits of a treaty resident are taxable only in the resident’s jurisdiction unless the resident carries on business in the other jurisdiction (i.e., the United States) through a permanent establishment in that jurisdiction. Nevertheless, tax treaties also generally provide that when business profits include items of income that are dealt with separately in other articles of the treaty and the resident has no permanent establishment in the other jurisdiction, then the provisions of those other articles are not affected by the business profits provision. Many treaties provide that interest arising in a contractual state (i.e., the United States) and beneficially owned by a resident of the other contracting state may be taxable only in that other state.
However, several tax treaties permit the United States to tax U.S.-source interest. For investors resident in one of these countries, it appears that the U.S. tax would be the lesser of the net income tax under domestic law and the gross-basis tax on interest under the treaty.
Assume that an Australian corporate investor is allocated $100 of interest income from originated loans and interest expense of $60. Absent the treaty, the investor would have net income of $40. At a 21% corporate rate, the tax would be $8.40, but there would also be a branch profits tax of $2.37 for a total tax of $10.77.
Under Article 10(6) of the Australia treaty, the branch profits tax may be imposed only on the profits of an Australian company that are attributable to a permanent establishment in the United States. If the Australian company has no permanent establishment, then the United States cannot impose the branch profits tax.
Under Article 7(6) of the Australia treaty, where business profits include items of income that are dealt with separately in other articles (e.g., Article 11 [interest]) and the resident has no permanent establishment in the other jurisdiction, those other articles apply. Under Article 11, the United States may impose up to $10 (10%) of tax on the gross amount of interest. But the United States would impose only $8.40 of tax on the net interest income under domestic law, and there is no provision of domestic law or the treaty that increases the tax for Australian residents beyond the normal withholding rate, although it would appear to be permitted under Article 23 (Nondiscrimination) of the Australia treaty. Therefore, it appears that, under the treaty, the tax would be $8.40. Nevertheless, a U.S. withholding agent paying interest to an Australian investor is going to withhold the full ten percent ($10). The Australian investor would have to apply for a refund of $1.60.
States and municipalities are not bound by treaties. Some states honor them, but many do not. So even an investor that qualifies for tax treaty benefits may be subject to state and local taxes.
BYOT feeders are for investors that are resident in a treaty jurisdiction. Many investors are not. Ireland and Luxembourg each have tax treaties with the United States and have special vehicles that can qualify for the benefits of the applicable treaty but pay no or relatively little local tax.
The Irish and Luxembourg treaties are similar. Each provides that a U.S. dependent agent that does not have the power to bind its principal, and an independent agent acting in the ordinary course of its business, do not cause an Irish or Luxembourg resident entitled to the benefits of the applicable treaty to be subject to U.S. income tax, and each provides for a zero rate of withholding tax on interest. Each also generally provides under its respective limitation of benefits provision that if (1) 50% or more of the aggregate vote and value of a resident corporation’s shares is held by “good” U.S. persons or at least 95% of the aggregate vote and value of the resident corporation’s shares are owned directly or indirectly by seven or fewer qualified residents of the EU or NAFTA and (2) a base erosion test is satisfied, then the vehicle qualifies for treaty benefits and is exempt from U.S. federal income tax on business profits and interest. There are, however, meaningful differences between the two treaties and differences between each country’s investment vehicles.
Ireland imposes a 25% corporate tax on the income of an Irish corporation from a business carried on wholly outside Ireland. A special regime exists for “section 110 companies,” however. Section 110 companies are special purpose companies that are permitted to invest in certain assets (including loans) and must satisfy certain rules. Section 110 companies may issue “profit participating notes” to their shareholders that reduce the effective Irish tax rate to nearly zero. Nevertheless, anti-avoidance provisions enacted in 2011 impose Irish withholding tax on interest payments under profit participating loans if the interest is paid to a resident of the EU or an Irish tax treaty partner that does not tax the payment.
Section 110 companies may qualify for benefits under the Irish treaty, but because they rely on interest deductions to reduce their Irish tax, they are subject to the base erosion test discussed below.
Irish Collective Asset-Management Vehicles (ICAVs) are regulated companies that (like U.S. mutual funds) are not subject to corporate tax or withholding tax on distributions. Because they are regulated by the Irish central bank, they are more expensive to organize and maintain but, because they do not rely on interest deductions to eliminate their corporate tax, they are not subject to the base erosion test discussed below.
Under the Irish treaty, a resident of Ireland includes any person who, under Irish law, “is liable to tax” in Ireland by reason of its place of incorporation. This describes section 110 companies and ICAVs. Even though neither of them pay much (if any) Irish tax, the Irish treaty Technical Explanation makes clear that an entity may be “liable to tax” without actually paying tax. Article 4(1)(d) of the treaty makes clear that ICAVs qualify. It provides that a “resident” includes, “in the case of the United States, a Regulated Investment Company and a Real Estate Investment Trust; in the case of Ireland, a Collective Investment Undertaking; and any similar investment entities agreed to by the competent authorities of both countries.” An ICAV is a Collective Investment Undertaking.
Under Article 23 of the Irish treaty (the limitation on benefits provision), only “qualified persons” are entitled to the benefits of the treaty. Under Article 23(2)(c) of the Irish tax treaty, an Irish resident company is a qualified person if:
i) . . . at least 50% of the aggregate vote and value of the company’s shares[] is owned, directly or indirectly, by qualified persons or residents or citizens of the United States, provided that such ownership test shall not be satisfied in the case of a chain of ownership unless it is satisfied by the last owners in the chain [the ownership test], and
ii) amounts paid or accrued by the [Irish company] during its fiscal year:
(A) to persons that are neither qualified persons nor residents or citizens of the United States and
(B) that are deductible in that fiscal year for [Irish] income tax purposes . . . (but not including arm’s length payments in the ordinary course of business) for
(1) services or tangible property, and
(2) payments in respect of financial obligations to a bank, provided that where such a bank is not a resident of either [the U.S. or Ireland] such payment is attributable to a permanent establishment of such bank, and the permanent establishment is located in [either the U.S. or Ireland],
do not exceed 50 percent of the gross income of the [company] [the base erosion test.]
The ownership test is designed to ensure that the treaty mostly benefits only Irish and U.S. taxpayers. This is achieved by requiring that at least 50% of the vote and value of an Irish company’s shares is held by “qualified persons” and U.S. residents (collectively referred to in this Article by its technical term, “good people”). The base erosion test is designed to limit nonqualified persons (and persons that are not U.S. residents) (i.e., “bad people”) from indirectly benefitting from the treaty by limiting the amount of base-eroding interest and other payments they may receive. Base-eroding interest is interest that is deductible for Irish tax purposes. There are three unusual aspects to the limitation on benefits provision of the Irish treaty.
a. U.S. Residents. For purposes of the ownership and base erosion tests under the Irish treaty, good people include “qualified persons or residents or citizens of the United States.” The inclusion of U.S. residents is very unusual. A resident under the Irish treaty includes any person who, under the laws of that state, is liable to tax therein by reason of its place of incorporation. As a result, all U.S. corporations are U.S. residents. In contrast, the equivalent Article of the Luxembourg treaty—Article 24(2)(c) and (d)—provides that only “qualified residents or U.S. citizens” are good people; for a U.S. corporation to be a qualified resident, it must be publicly traded. A nonpublicly traded U.S. corporation wholly owned by bad people is a good person for purposes of the ownership and base erosion tests of the Irish treaty but not the Luxembourg treaty. The Irish treaty Technical Explanation is clear on this. In fact, it appears that a bad person could organize a taxable C corporation to own 51% of an Irish section 110 company and own the other 49% directly.
b. The Proviso in the Ownership Test. The ownership test has an odd proviso. It says that the ownership test is not satisfied in the case of a chain of ownership unless it is satisfied by the last owners in the chain. Assume that an Irish corporation is owned by a U.S. corporation that, in turn, is owned by Bermuda individuals. Is the U.S. corporation part of a chain that requires us to look through the U.S. corporation to the Bermuda individuals and, because Bermuda individuals are not good people, the test is failed?
No. The technical explanation explains that ownership stops with a person who is entitled to benefits without reference to its owners, including a resident of the United States because neither the United States nor Ireland is concerned about treaty shopping through a U.S. resident. Thus, the chain ends with the U.S. corporation. Since the Irish corporation is owned by a U.S. resident, it satisfies the ownership test.
c. The Base Erosion Test and the Definition of Gross Income. Under the base erosion test, payments made to bad people—other than certain arm’s length payments made in the ordinary course of business—may not exceed 50% of the “gross income” of the Irish company. There is a special definition of gross income. It is the greater of (1) the gross income for the fiscal year preceding the current year and (2) the average of the annual amounts of gross income for the four fiscal years preceding the current fiscal year.
This definition causes two problems. First, under this test, the gross income for the first year is always zero. If this means what it says, then no Irish company with any bad person creditors would ever satisfy the base erosion test for the Irish company’s first year of existence because deductible bad person payments cannot exceed 50% of prior year gross income, which is zero for the first year. Despite the crystal clear language, most practitioners conclude that the provision cannot mean what it says, at least for the first year.
What about the other years? Assume that an Irish section 110 company with 50% U.S. residents and 50% bad people has issued profit participating loans to these investors proportionately. It earns $100 in its first year and $110 in the second and pays out all of that income on its profit participating loans. The payments to bad people in the second year may not exceed 50% of the first year gross income (or $50). In the second year, however, the payments to the bad people would be $55 (50% of $110). This section 110 company would not qualify for the treaty in the second year.
Some have argued that the drafters must have intended for this test to be elective because there’s no indication in the technical explanation or otherwise of any restrictive intent and, therefore, in the absence of an election to use this test, gross income is the gross income for the year being tested. However, with clear language and so much at stake, most advisors have not relied upon that argument. Instead, some section 110 companies’ profit participating loans provide that the interest is deductible only to the extent that it satisfies the base erosion test. To the extent the interest is not deductible, the Irish section 110 company will pay the 25% Irish corporate tax. Better to pay Irish corporate tax on some income than fail to qualify for treaty protection entirely for that year.
ICAVs are not concerned about this issue. Because they do not rely on interest deductions to eliminate their corporate tax, the base erosion test is not relevant.
d. Limitation on Benefits: The Derivative Benefits Test. Article 23(5)(a) of the Irish treaty provides that an Irish resident company is also entitled to the benefits of the treaty if:
i) at least 95 percent of the aggregate vote and value of all its shares is owned directly or indirectly by seven or fewer qualified persons or persons that are residents of member states of the European Union or of parties to the North American Free Trade Agreement (NAFTA) or any combination thereof [the ownership test]; and
ii) such company meets the [base erosion test described above], provided that a resident of a member state of the European Union or a party to NAFTA shall be treated as a qualified person for the purposes of that test [the base erosion test].
For these purposes, a “resident of a member state of the European Union” and a “resident of a party to NAFTA” means a person that would be entitled to the benefits of a comprehensive income tax convention in force between any member state of the European Union, or a party to NAFTA, as applicable, and the United States. Weirdly, a U.S. resident may not be a resident of a party to NAFTA.
Article 23(5)(b) imposes an additional requirement on an Irish resident company relying on the derivative benefits test to claim the benefits of the interest, dividends, and royalties articles of the treaty: 95% of its shares must be held directly or indirectly by one or more persons that are residents of a member state of the European Union or of parties to NAFTA or any combination thereof, and the treaty between their home jurisdictions and the United States must provide for a zero rate of withholding on interest (or the Irish treaty rate on royalties or dividends). Otherwise, the Irish treaty rate of withholding does not apply.
Therefore, under the derivative benefits test, if an insurance company resident in Europe and eligible for the benefits of a comprehensive income tax treaty between its jurisdiction of residence and the United States that provides for a zero rate of withholding on interest owns subsidiaries in Canada, Mexico, and Europe that qualify for the benefits of a comprehensive income tax treaty with the United States, and also subsidiaries in Asian jurisdictions, and the subsidiaries own proportionate interests in all of the stock and debt in an Irish section 110 company, the Irish section 110 company will satisfy the ownership test because the derivative benefits test requires only that seven or fewer qualifying European-resident companies directly or indirectly own at least 95% of the vote and value of the Irish company. The Irish company will satisfy the base erosion test if the interest paid to the Asian subsidiaries does not exceed 50% of the Irish company’s gross income (determined under the weird rule described in Part VII.B.2.c, above). (The base erosion test is applied directly.) Since investors usually hold stock and profit participating loans proportionately, this would generally require that the European, Mexican, and Canadian subsidiaries directly own at least 50% of the capital of the Irish company at all times.
An ICAV can be organized as an “umbrella” with a series of segregated portfolio companies (or “cells”). Different groups of investors may invest in each cell; the owners of one cell may not proceed against the assets of another. Cell companies raise the question whether the ownership and base erosion tests are applied at the umbrella level (i.e., treating all investors in each cell as if they were investors in the umbrella), the cell level (i.e., treating the investors of each cell as if they were investors in a stand-alone entity without regard to investors in another cell), or both (i.e., treating the ICAV as having satisfied the ownership and base erosion tests only if it satisfies them with respect to each cell and also with respect to the ICAV as a whole). Treating each cell as a separate entity that applies the ownership and base erosion test without regard to any other cell is likely the right answer, but in the absence of any guidance and with the stakes so high, advisors assume the worst and apply the test both at the cell level and at the ICAV level. This same issue exists with Luxembourg Sàrls that have compartments, and the resolution is the same.
The leading Luxembourg treaty vehicle is the société à responsabilité limitée (Sàrl). Sàrls are subject to Luxembourg corporate tax and typically issue “profit participating notes” or “asset linked notes” to investors to minimize that tax. These notes allow a deduction to the Luxembourg corporation equal to most of the Luxembourg corporation’s profits.
The 50% ownership and base erosion test of the Luxembourg treaty is substantially similar to that in the Irish treaty, except in four respects. First, the Luxembourg definition of a good person is much more limited than the Irish definition. Second, the Luxembourg treaty does not have the Irish treaty’s special (and bad) gross income test. Third, the ownership test under the Luxembourg treaty must be satisfied with respect to a Luxembourg company’s principal class of shares. Finally, qualified persons are not good people for purposes of the derivative benefits provision of the Luxembourg treaty.
a. More Traditional Definition of “Good Person.” First, as mentioned above, the Irish treaty has a very favorable definition of a good person because it includes all U.S. corporations. The Luxembourg treaty is much more typical and not nearly as beneficial. Under the Luxembourg treaty, good for-profit U.S. corporations include only those that are publicly traded.
b. Absence of a Special Gross Income Test. As mentioned above, the Irish treaty has a very unusual and troubling definition of gross income for purposes of the base erosion test. Luxembourg does not have that definition. Instead, gross income has its normal meaning—gross income for the year being tested.
c. The Principal Class of Share Test. The Luxembourg treaty provides that the ownership test must be satisfied with respect to the “principal class of shares.” The treaty does not define the term “principal class of shares,” but the Technical Explanation provides that this term refers “generally to the “ordinary or common shares of [the] company,” so long as that class of shares represents the majority of the voting power and value of the company. If more than one group of classes accounts for more than 50% of the shares, it is only necessary that one such group satisfy the requirements. If there is no single class that represents a majority of the voting power and value of the company, then the principal class of shares are those classes that in the aggregate own more than 50% of the voting power and value of the company.
This provision potentially provides, alternatively, a trap or a treasure. Assume that that a Luxembourg company has two classes of shares. The first class represents 51% of the vote and value of the company and is owned 51% by a Cayman corporation and 49% by a U.S. individual. The second class represents 49% of the vote and value and is owned by a U.S. individual. Under the treaty, the first class is the principal class. Because good people do not own at least 50% of that principal class, the ownership test is flunked, even though U.S. individuals own 73% of the vote and value of the Luxembourg company.
On the other hand, assume that a Luxembourg company has two classes of stock. The first has 51% of the vote and value and, in turn, is owned 51% by an individual U.S. citizen and 49% by a Cayman Islands corporation. The second class has 49% of the vote and value and is owned by a different Cayman Islands corporation. If the first class is the principal class of shares, the ownership test is satisfied even though the U.S. person owns a little more than 25% of the vote and value of the entire company and the Cayman Islands corporations own nearly 75%.
d. The Derivative Benefits Test. The Irish treaty treats “qualified persons” as good people for the derivative benefits test. The Luxembourg treaty does not. However, the Technical Explanation for the Luxembourg treaty leaves open the possibility that a U.S. resident is a resident of a party to NAFTA (so it would be easier for a Luxembourg Sàrl to argue that a U.S. resident is a good person for purposes of the Luxembourg treaty).
Irish section 110 companies and Luxembourg Sàrls look a lot like conduits. They receive interest income and pay out nearly all of it as interest expense (at least for local law tax purposes) on their profit participating notes or asset linked notes. If a Luxembourg or Irish company is a conduit for U.S. tax purposes, then it would not benefit from the treaty.
In Aiken v. Commissioner, a Bahamian parent corporation owned a U.S. subsidiary and an Ecuadorian subsidiary. The Ecuadorian subsidiary, in turn, owned a Honduran subsidiary. The Bahamian parent corporation loaned funds to the Honduran subsidiary which, in turn, loaned the funds to the U.S. subsidiary. The U.S. subsidiary paid interest to the Honduran subsidiary, which paid an equal amount of interest to the Bahamian parent corporation. U.S. withholding tax would have applied to interest paid by the U.S. subsidiary to the Bahamian parent corporation. At the time, a tax treaty existed between the United States and Honduras that eliminated withholding tax on interest, but no treaty existed between the United States and the Bahamas. The U.S.-Honduras income tax treaty did not contain a limitation on benefits provision.
The Tax Court held that because the U.S.-Honduras income tax treaty specifically defined a Honduran corporation and the Honduran subsidiary satisfied that definition, the Honduran subsidiary could not be disregarded. However, the treaty required that for interest to be exempt from U.S. tax under the treaty, the interest must be “received” by the Honduran subsidiary. The treaty did not define the meaning of “received” and provided that any terms not otherwise defined were to be given the meaning ascribed to them by the jurisdiction imposing the tax (i.e., the United States). Applying what it concluded were “the genuine shared expectations of the contracting parties” (i.e., the governments of the United States and Honduras), the Tax Court held that, because the Honduran subsidiary “was committed to pay out exactly what it collected, it made no profit” on the transaction, and because there was no valid nontax economic or business purpose for the transaction other than to avoid U.S. tax, the Honduran subsidiary did not “receive” the interest income for purposes of the U.S.-Honduras income tax treaty. Instead, the Honduran subsidiary was merely “a collection agent” acting for the benefit of the Bahamian parent corporation with respect to the interest paid by the U.S. subsidiary. Accordingly, U.S. withholding tax was imposed on the interest payments.
A Luxembourg or Irish company should not be treated as a conduit under Aiken. In contrast to the Honduran subsidiary in Aiken, the Irish or Luxembourg company is not “committed to pay out exactly what it collect[s]” with respect to its investment portfolio, and will not have “the same inflow and outflow of funds”; instead, during the investment period of a typical fund, the Luxembourg or Irish company will be free to invest and reinvest proceeds from the disposition of investments, and both during the investment period and thereafter, payments under the profit participating or asset linked notes will be net of (1) an arm’s length profit spread, (2) expenses, and (3) taxes incurred by the Luxembourg or Irish company. Thus, the Irish or Luxembourg company should have “dominion and control” over its income.
Second, the treaty between the United States and Honduras did not have a limitation on benefits provision and, in the absence of such a provision, the Tax Court in Aiken concluded that the governments of the United States and Honduras did not intend for an item of income to be treated as received by the Honduran subsidiary where it was paid out. By contrast, the Irish and Luxembourg treaties have limitation on benefit provisions, and Article 23 of the Irish treaty and Article 24 of the Luxembourg treaty specifically provide that an Irish or a Luxembourg company, as applicable, may qualify for benefits under the relevant treaty even if some of its owners and creditors are not U.S. citizens or qualified residents and provide specifically how much interest (and other deductible amounts) may permissibly be paid to bad people. The specificity of the limitation on benefits provisions in the Irish and Luxembourg treaties also argues against the application of Aiken.
Regulation section 1.881-3 permits the Service to disregard one or more intermediate entities in a financing arrangement for purposes of section 881 if the entities are acting as conduit entities. If one of the ultimate investors is a bank, these regulations could apply. For this reason, sponsors will often prohibit banks from investing in Irish or Luxembourg vehicles.
For purposes of these rules, a “financing arrangement” is a series of transactions pursuant to which one person (a “financing entity”) advances money, and another person (the “financed entity”) receives money, if the advance and receipt are effected through one or more other persons (“intermediate entities”) and “financing transactions” link the financing entity, each of the intermediate entities, and the financed entity.
If an Irish section 110 company or a Luxembourg Sàrl is treated as an intermediate entity that is linked to the ultimate investors by a financing transaction (i.e., the profit participating notes or asset linked notes issued by the Irish or Luxembourg company), and the obligors underlying the Luxembourg or Irish company’s investments are financed entities that are linked to the Luxembourg or Irish company by a financing transaction (i.e., the loans) then, under Regulation section 1.881-3, the Luxembourg or Irish company would be treated as a conduit entity and disregarded for purposes of section 881. In this case, to the extent an ultimate bank investor receives U.S.-source interest income and does not benefit from a treaty that provides for a zero rate of withholding on that interest, the investor would be subject to a withholding tax. However, the conduit rules should not apply to other investors.
Under the regulations, an intermediate entity (such as the Irish or Luxembourg company) may be treated as a conduit entity only if the entity’s participation in the financing arrangement reduces the tax imposed under section 881. As discussed below, assuming that none of the investors are banks (or, if they are, they benefit from a treaty with a zero rate of withholding on interest), if the Irish or Luxembourg company were disregarded, no tax would be imposed under section 881 because either (1) the interest income would be effectively connected with the conduct of a trade or business in the United States and subject to tax under section 882 or (2) the investors would qualify for the “portfolio interest exemption” under section 881(c), for benefits under a treaty, or for both. (The loans held by a loan-origination loan fund are generally limited to loans that qualify for the portfolio interest exemption when held by nonbank investors.)
a. Effectively Connected Income. Section 881(a) generally imposes a 30% withholding tax on U.S.-source interest income, “but only to the extent the amount so received is not effectively connected with the conduct of a trade or business within the United States.” If an Irish or Luxembourg company is treated as engaged in a trade or business in the United States, then the interest income it receives is treated as effectively connected with the conduct of a trade or business in the United States and, therefore, would be subject to tax under section 882 and not section 881. Likewise, if an Irish or Luxembourg company is disregarded under Regulation section 1.881-3, the investors would also be subject to tax under section 882 and not under section 881. Therefore, the participation of the Irish or Luxembourg company does not reduce the tax imposed by section 881.
b. The Portfolio Interest Exemption. Under the portfolio interest exemption of section 881(c), U.S. federal withholding tax is not imposed on any interest (1) that would otherwise be subject to tax under section 881(a), (2) that is paid on an obligation in “registered form” for U.S. federal income tax purposes, (3) that is not “contingent interest” described in section 871(h)(4), (4) with respect to which the paying agent receives an IRS Form W-8, (5) that is not paid to a bank on an extension of credit made pursuant to a loan agreement in the ordinary course of the bank’s trade or business, and (6) that is not received by a “10% shareholder” of the obligor or by a controlled foreign corporation (CFC) from a related person. Loans held by a loan-origination fund are generally limited to those that satisfy these requirements, except possibly item (5) with respect to a bank investor.
If banks that do not benefit from a zero rate of withholding under a treaty are prohibited from investing in the Irish or Luxembourg company, then all of the interest received by the investors in the fund should either qualify for the portfolio interest exemption or be exempt under a treaty. Because the participation of the Irish or Luxembourg company would not reduce the tax otherwise imposed under section 881, the conduit regulations should not apply to impose U.S. withholding tax.
In Chief Counsel Attorney Memorandum 2009-010, the Service explained the legal analysis that permits a U.S. manager that qualifies as an “independent agent” to originate non-U.S. loans on behalf of a foreign fund without causing the foreign fund to have ECI. (The foreign fund would technically be engaged in a trade or business in the United States by reason of the U.S. manager’s participation in the origination of loans to non-U.S. borrowers.)
Under section 864(c)(4)(A), foreign source interest income is generally not treated as effectively connected with the conduct of a U.S. trade or business. Foreign source interest income of a foreign corporation that is derived from the active conduct of a banking, financing, or similar business within the United States is, however, treated as effectively connected with the conduct of a U.S. trade or business if the foreign corporation “has an office of other fixed place of business within the United States to which such income, gain, or loss is attributable.” For purposes of determining whether a foreign corporation has an office or other fixed place of business under this rule, the office or other fixed place of business of an agent is disregarded unless the agent (1) has the authority to negotiate and conclude contracts in the name of the foreign corporation and regularly exercises that authority and (2) is not a general commission agent, broker, or other independent agent acting in the ordinary course of business.
Regulation section 1.864-7(d)(3) defines an independent agent in the same manner as treaties, and therefore the analysis in Taisei (discussed above in Part VI.A.1) should apply equally for purposes of Regulation section 1.864-7(d)(3). Thus, if a U.S. manager qualifies as an independent agent, then it may be given full discretionary authority to make non-U.S. loans on behalf of a foreign fund without causing that foreign fund to have ECI.
As mentioned above, because the foreign fund would be engaged in a trade or business in the United States, it avoids ECI only with respect to its foreign source income. A technical issue arises if the foreign fund also enters into currency swaps to convert non-U.S. currency into U.S. dollars. Regulation section 1.988-4(c) provides that exchange gain or loss that, under principles similar to those set forth in Regulation section 1.864-4(c), arises from the conduct of a U.S. trade or business is sourced in the United States, and the gain is treated as effectively connected to the conduct of a U.S. trade or business.
Under the asset-use test in Regulation section 1.864-4(c)(2), fixed or determinable annual or periodical gains from an asset are deemed effectively connected with the conduct of a trade or business in the United States if the asset is held in a direct relationship to the trade or business in the United States. An asset is held in direct relationship to the trade or business in the United States if the asset is held to meet the operating expenses or other present needs of that trade or business. If management fees are paid in dollars and a foreign currency swap converts non-U.S. dollar currency into U.S. dollars so that the fund can pay its management fees, the asset-use test would appear to be satisfied. A direct relationship is also presumed to exist if personnel who are present in the United States and actively involved in the conduct of the trade or business exercise significant management and control over the investment of the asset. So, even if the swap is not held to meet the operating expenses or other present needs of the fund’s trade or business, if the swap is controlled by the U.S. personnel, the asset-use test would appear to be satisfied. Thus, under this test, the foreign currency gain or loss would appear to be effectively connected with the U.S. trade or business.
That said, it is entirely counterintuitive that the interest on a loan would be exempt from U.S. tax, but income from a foreign currency swap to hedge that interest would be subject to U.S. tax. An inconvenient solution would be to have the master fund that owns the foreign subsidiary enter into the swap. The master fund would qualify for the “trading in derivatives” safe harbor under Proposed Regulation section 1.864(b)-1.
Under section 865(e)(g), gain of a nonresident from the sale of personal property attributable to a U.S. office is generally U.S.-source. So any gain would be ECI. However, loan funds rarely have gain from the sale of loans.
For purposes of the rules described above, the office or other fixed place of business of an agent is disregarded if the agent does not have the authority to negotiate and conclude contracts in the name of the foreign corporation or does not regularly exercise that authority, even if the agent is a dependent agent. Assume that a non-U.S. manager with discretionary authority hires an unrelated U.S. party to find non-U.S. borrowers and negotiate loans with them, but the U.S. party does not have the authority to enter into the loans on behalf of the fund. Only the non-U.S. manager would have that authority. In this case, the fund would not appear to have ECI. However, the U.S. party should take great pains to avoid the appearance of being able to bind the fund. Apparent authority may give rise to actual authority under local law.
Assume instead that a manager has a U.S. office and a non-U.S. office, but doesn’t qualify as an independent agent. The investment committee members live and work out of the non-U.S. office, and the investment committee makes all of the decisions. The U.S. office finds the non-U.S. loans and negotiates them, but the U.S. personnel do not have the authority to bind the fund. Can the fund rely on section 864(c)(5)(B) and Regulation section 1.864-7(d)?
The answer is not so clear. Regulation section 1.864-7(d) provides that the office of an agent is disregarded unless the agent has the authority to negotiate and conclude contracts in the name of the fund. In the example above, only one manager has the authority to negotiate and conclude contracts on behalf of the fund, even though the U.S. office of the manager does not.
The language of the regulation is different from the language in treaties, which provides that a dependent agent causes its principal to have a permanent establishment only if the agent “is acting on behalf of an enterprise and has and habitually exercises in a Contracting State an authority to conclude contracts in the name of the enterprise . . . .” So if the investor benefits from a treaty and invests in a treaty fund, this structure seems to work.
Assume the reverse: The individuals who are physically outside the United States conduct all substantive activities with respect to loan origination to non-U.S. borrowers, except that one or more U.S. individuals sit on the investment committee that approves those loans and the manager is not an independent agent with respect to the fund. Does the fund have ECI?
Regulation section 1.864-6(b)(2)(ii)(b) applies the principles of Regulation section 1.864-4(c)(5)(ii) for purposes of determining whether non-U.S. source interest income from any transaction, or gain from the sale or exchange of a security, is ECI. As discussed in Part IV.C, it is unclear what “execution of the acquisition of a security” means. Therefore, it is unclear whether this exemption is available.
Section 864(c)(4)(D)(ii) and Regulation section 1.864-5(d) provide that any foreign source income of a CFC that is Subpart F income is not ECI, even if the CFC is owned by non-U.S. people and even if no amount is included in the gross income of a United States shareholder. Without this rule, if a CFC entirely owned by United States shareholders was engaged in a trade or business, the CFC’s home jurisdiction could tax the income, the CFC would be taxed by the United States on the ECI, and the United States shareholders would also be taxable on the Subpart F income, leading to a potential immediate triple tax without deemed paid foreign tax credits for individual shareholders. Ostensibly, the tax policy behind this rule is to provide relief from the triple tax by foregoing U.S. corporate income tax when foreign source income is also Subpart F income of a CFC. Presumably as a matter of administrative convenience, the regulations ignore the fact that the CFC may have foreign shareholders.
Assume non-U.S. investors invest in a Delaware partnership that, in turn, owns a Cayman Islands corporation. The Cayman Islands corporation is managed by a U.S manager and originates non-U.S. loans. In its most extreme application, this regulation would allow the Cayman Islands corporation to avoid ECI on its foreign-sourced interest income. The Delaware partnership is a United States shareholder. Therefore, the Cayman Islands corporation is a CFC, and the interest would be Subpart F income, even though no person will report it. The potential triple tax contemplated by the regulations is a zero tax in this case, and still the regulations forego the U.S. tax.
If U.S. investors are involved in the Delaware partnership, this structure is not ideal for them under the current regulations. The Delaware partnership would be a United States shareholder, and all U.S. investors would report Subpart F income of the CFC. Accordingly, any long-term capital gains would be converted to ordinary income, and losses would be trapped in the CFC. Under Proposed Regulation section 1.958-1(d), however, Subpart F income would be reported only by those U.S. investors that indirectly own ten percent of the vote or value of the CFC. Other U.S. investors would treat the Cayman Islands corporation as a passive foreign investment company. Accordingly, the Delaware partnership could make a qualifying electing fund election with respect to the Cayman Islands corporation. While losses would be trapped and unusable until the liquidation of the Cayman Islands corporation, long-term capital gains could pass through.
The CFC exception applies only to foreign source income. Therefore, foreign currency and gain are potentially treated as ECI, as discussed in Part VIII.A, above.
Mezzanine (or mezz) funds are funds that originate a limited number of mezzanine or second lien loans, often acquiring equity or options in the borrower, and sometimes with some governance rights. Mezz funds take the position that they are not engaged in a trade or business in the United States under two rationales. First, while they do originate loans, they argue that they are not in the business of making loans because they make relatively few loans, and second, they are not competing with domestic banks. Mezz fund investments are generally more speculative than bank loans: the subordinate nature of their loans and equity investments subjects them to more risk, and these loans and equity provide more yield and greater return than banks receive when they make loans.
However, as discussed above in Part II.D, the Service argues that the phrase “trade or business” has different meanings depending on the context, and therefore the phrase as used in section 864 may not have the meaning of the phrase for purposes of sections 162, 166, and 7704. Second, there is no official or unofficial Service or other guidance that distinguishes mezz funds from more traditional senior loan funds. In fact, in Example (3) of Regulation section 1.864-4(c)(5)(vii), a foreign bank makes a loan to a U.S. borrower and receives back equity in exchange. Therefore, the drafters of Regulation section 1.864-4 contemplated that lenders could receive equity from borrowers. The example holds that the dividends and any gain on the equity (along with interest on the loan) is ECI.
Mezz fund guidelines typically will limit the number of loans that the fund can make each year and over the term of the fund. Some guidelines limit the fund to an average of five loans a year.
Other guidelines limit the number of loans to ten a year and 40 over the term of the fund. Guidelines that allow a higher number of loans each year typically require that the loans be accompanied by a significant equity investment in the obligor. For example, such a set of guidelines might require that the fund (1) reasonably expect when it makes an investment that the equity component of the expected pre-tax internal rate of return will be more than 20% with respect to the investment or (2) hold at least five percent of the equity or equity-type rights of the borrower.
Nevertheless, these guidelines might also permit the fund to make up to three investments a year and 12 investments during the life of the fund that do not satisfy the equity component requirement so long as the loan component of these nonconforming investments do not exceed 30% in principal amount or number of all loans and the equity component represents at least five percent of the fair market value of the accompanying loan or else that the loan is expected to produce equity-like returns. The guidelines might also allow up to two “bullets” a year—loans that do not satisfy any criteria—provided that such a loan is made in consultation with the fund’s tax advisors.
Guidelines may also require that the fund not be registered as a bank, dealer, underwriter, or market-maker and not share in fees for services. They also might require that (1) all loans be subordinated and unsecured and have a minimum five-year term (but generally a six- to eight-year term), (2) the fund hold all loans for at least two years and not reinvest the proceeds of the loans, and (3) the fund seek governance rights in the underlying issuer, such as the right to consult with and advise issuer management, the right to receive regular financial reports, or the right to observe and participate on the board of directors of the issuer.
Some mezz guidelines limit the amount of debt the fund can incur to a 1:1 debt-to-equity ratio, plus additional short-term debt. Generally, the guidelines prohibit the fund from acquiring a revolving loan, but some permit the fund to enter into a delayed drawdown term loan so long as the fund’s future funding obligations are fixed and the fund does not have any discretion as to whether or not to fund the drawdown.
In a “horizontal” season-and-sell structure, an “onshore fund” with only U.S. and, possibly, treaty eligible investors will originate a loan, wait a period of time (typically 60 days), and then offer to sell a portion of the loan to an “offshore fund” managed by the same manager. If the offer is accepted, the actual sale of the loan typically may be executed no sooner than 90 days after the loan origination. Only the onshore fund will earn origination fees. The offshore fund will employ an “independent investment professional” (IIP), who is independent of the manager and acts on behalf of the investors in the offshore fund. The IIP will typically be given absolute discretion to purchase the loan, reject the purchase, or renegotiate the terms of the purchase (typically price). Often the offshore fund will obtain an independent valuation of the loan to help aid the IIP’s decision.
The theory underlying season and sell is that, because the offshore fund has no obligation to purchase the loan offered to it by the onshore fund, the onshore fund is not acting as the agent of the offshore fund when it originates the loan, and a 90 day period is sufficient for the onshore fund to establish its ownership. Therefore, when the offshore fund purchases the loan, it should qualify for the trading in securities safe harbor.
Season-and-sell structures pose three commercial issues. First, because the onshore fund will hold loans for at least 90 days before selling them to the offshore fund, the onshore fund will earn the origination fees and will earn more interest income than the offshore fund; therefore, the investment returns of the two funds will diverge. Second, because the onshore fund will have to purchase more of each loan than it intends to ultimately retain, and is required to hold that portion of the loan for at least 90 days before it can sell that portion to the offshore fund, the onshore fund will be unable to purchase loans as frequently as it might otherwise prefer. This problem is sometimes referred to as a “capacity issue” because the onshore fund may lack the capacity to buy the loans it wants to buy, by reason of holding loans for sale to the offshore fund.
Finally, if an adverse event occurs with respect to the borrower or the market within the 90 day period, the IIP may refuse to purchase the offered portion of the loan, in which case the onshore fund will be left holding both its own portion of the loan and the offshore fund’s portion, at the worst possible time.
Season and sell works if the IIP is making independent investment decisions on behalf of the offshore investors, and can so convince a fact finder. Season and sell does not work if the IIP blindly approves each offer. The best IIPs have a history of evaluating loans, apply that experience to each loan, act independently, document their decisions, and act as credible witnesses if called to testify. IIPs are usually paid either a fixed fee for each loan or a fixed fee each year regardless of the number of loans. They may act for more than one offshore fund but should, in each case, act only for buyers whose interests are not adverse.
Typically, the general partner of the fund will name the initial IIP, but a majority of investors may remove the IIP for any or no cause. The general partner may then propose a replacement IIP who will be appointed unless a majority of the investors object within 30 days. If the IIP resigns, the general partner appoints his or her replacement.
The mechanics of a season-and-sell structure might work as follows. The onshore fund purchases a loan without consulting the IIP. The IIP is not consulted because were the IIP consulted prior to loan-origination approval either the origination activity could be imputed to the IIP or else, psychologically, the IIP would be far less likely to reject the loan after the 90-day period. The object of the season-and-sell structure is for the IIP to be given complete discretion to accept or reject the loan following the seasoning period.
After the onshore fund has funded and closed a loan, it would hold it for at least a 60-day period. After that 60-day period, the onshore fund would offer a portion of the loan to the offshore fund (often in the form of a participation interest). The offer would be accompanied by all of the materials (such as the investment memorandum) that were used by the onshore fund to make its purchase determination. These materials should describe the proposed investment and the terms of the offer, and the manager of the onshore fund should be available to provide market color to the IIP and answer questions. Also, at that time, a third-party valuation would be commissioned by the offshore fund. The IIP would be required to make a decision on the loan no sooner than the ninetieth day following the funding and closing of the loan, and any valuation would be updated to that date. Valuations may be ranges or specific amounts. The IIP would not be bound by the valuation and could decide to purchase a loan for an amount in excess of the valuation. However, in this event, the IIP should document the reasons for the decision. The onshore fund would keep any fees it received for originating the loan, and any interest that accrued while it owned the loan.
In the case of subsequent closings of the onshore or offshore funds where “rebalancing” must occur between the two funds (i.e., portions of loans are bought or sold between the funds), the IIP must approve any additional purchases by the offshore fund (but only after the 90-day period has ended), but need not approve sales. (Sales do not pose origination risk.)
For U.S. tax purposes, the offshore fund will be treated as engaging in any activity in which its agents engage. If the onshore fund is acting as the agent of the offshore fund and the onshore fund originates loans, then the offshore fund will be treated as originating loans. Season and sell rests upon basic agency principles.
An agent for this purpose is defined under common law principles. The Restatement (Third) of the Law of Agency defines agency as the consent by one person (the “principal”) that another person (the “agent”) act on the principal’s behalf, subject to the principal’s control. Agency has been held to require (1) the ability of the agent to bind the principal, (2) the existence of a fiduciary relationship between principal and agent, and (3) the ability of the principal to control the conduct of the agent with respect to the matters entrusted to it.
These elements are not present in the relationship between the onshore fund and the offshore fund. There is no formal or informal agency agreement between the onshore fund and the offshore fund. The onshore fund will not have the power to bind the offshore fund, and the onshore fund will purchase loans for itself, without any assurance that the offshore fund will purchase any portion of the loans. In fact, the IIP (representing the offshore fund) will not typically be shown the identity or terms of any loan until at least 60 days after the onshore fund has funded and closed the loan, and the onshore fund will be entitled to all fees and interest that accrue during the period that it owns the loan and will be subject to all market (and other) risks during that period.
Moreover, the offshore fund will be under no obligation to purchase any loan from the onshore fund. The IIP will represent only the offshore fund and its investors with respect to the purchase of loans from the onshore fund and, in his or her sole discretion acting on behalf of the offshore fund, may accept or decline the offer to purchase a loan from the onshore fund or negotiate the terms of any purchase. In addition, the offshore fund will obtain from an independent third-party expert a valuation of a loan that is offered by the onshore fund and provide the valuation to the IIP. Finally, the offshore fund will not have any right to control the onshore fund and will not in fact control the onshore fund.
In Amalgamated Dental Co. v. Commissioner, a U.K. corporation owned approximately one-third of the stock of a U.S. corporation. As a result of difficulty shipping goods to its customers in the United States during World War II, the U.K. corporation entered into an agreement with the U.S. corporation whereby the U.S. corporation would fill orders from the customers of the U.K. corporation and pay the U.K. corporation the difference between the retail price and the wholesale price of the orders.
Rather than treat the U.S. corporation as the agent of the U.K. corporation, which would have subjected the U.K. corporation to U.S. federal income tax, the Tax Court held that because there was no formal agreement between the two companies and the U.S. corporation was not paid as an agent, no agency existed. Instead, the court treated the U.K. company merely as having purchased the goods and directing the U.S. corporation to ship them to the U.K. corporation’s customers.
Likewise, no agency agreement will exist between the onshore fund and the offshore fund, and the onshore fund will not be paid as an agent. Therefore, under Amalgamated Dental, the onshore fund does not appear to be the agent of the offshore fund.
Similarly, courts have held that an agency relationship does not exist when the purported agent does not have the power to bind the purported principal. For example, in Griffin v. United States, the court held that an agency relationship did not exist when the purported agent did not have the power to commit the purported principal to business relationships with third parties. And in Cadwallader v. Commissioner, in which a U.S. corporation sent orders to a Filipino corporation but the Filipino corporation was free to accept or reject them, the court held that the U.S. corporation was not the agent of the Filipino corporation.
As in Griffin and Cadwallader, the onshore fund will not have the power to bind the offshore fund. Instead, the offshore fund (acting through the IIP) will independently decide whether to purchase or reject any loan and will have full discretion to negotiate price. Accordingly, under these cases, the onshore fund should not be considered the agent of the offshore fund.
Of course, agency cases are decided on their facts. If the manager of the onshore fund can bind the offshore fund by, for example, forcing it to take a loan or otherwise controls the IIP, or if there is an understanding that the IIP will simply accept each loan that is offered by the onshore fund, an agency relationship could be found.
In the typical season-and-sell structure, as described above, a single manager manages both the onshore and offshore fund, but an IIP is given discretion to make purchase decisions on behalf of the offshore fund. In this situation, a 90-day seasoning period is typical.
Assume instead that an originating party (which could be a bank or an onshore fund) merely offers loans to an unrelated non-U.S. investor and the investor has the sole discretion to purchase, reject, or negotiate the terms of the loans. Under the best of facts, it is possible that such an arrangement could satisfy the guidelines discussed in Part V.E, above, in which case the seasoning period could be as little as 48 hours. If no strings are attached to the decision to purchase or reject, then the existence of an agency relationship should not be much of a concern.
Unfortunately, the facts are rarely so good. For example, the parties may agree that if the non-U.S. investor were to reject more than a fixed number of loans, the originating party will stop offering them. Or the parties may have an understanding that the non-U.S. investor will accept all offered loans that satisfy certain criteria (although the understanding is not legally binding). These factors may lead the parties to increase the seasoning period. The next usual stopping point is 14 days. Analogous authority is found in a series of private letter rulings relating to spinoffs. Under these private letter rulings, an intermediary is respected as the owner of the debt of a distributing company if the intermediary assumes meaningful risk for a minimum time period, which is established as five days of price and event risk and 14 days of execution and credit risk. Although the private letter rulings are now obsolete in light of Revenue Procedure 2018-53, the concept originated in the section 108 context and appears to reflect a broader principle of tax ownership and agency.
In a vertical seasoning structure, a fund establishes a U.S. subsidiary that is treated as a corporation for U.S. tax purposes. The fund capitalizes the subsidiary with equity (and sometimes also debt). The U.S. subsidiary originates loans, holds them for a period of time, and then sells the loans to the parent for a price usually determined under a transfer pricing methodology to be fair market value.
In a 2014 Tax Club paper, Stuart Leblang, Daniel Paulos, and Brett Fieldston developed the rationale for vertical season and sell. In a number of cases, arm’s length transactions between parents and their subsidiaries have been respected, and the subsidiaries were not treated as agents of the parents. Leblang, Paulos, and Fieldston conclude that it is harder to find agency between parents and their subsidiaries than between unrelated persons.
The holding period for vertical seasoning is usually longer than that for horizontal seasoning—180 days versus 90 days (some sponsors use shorter periods for each)—because of increased concern regarding agency in a related-party context. In addition, to best align with the authorities respecting transactions between parents and their subsidiaries, often the subsidiary is required to have its own employees. And, finally, to avoid a circular cash flow when the U.S. subsidiary sells a loan to its parent (which might encourage a court to disregard the structure), the subsidiary must “trap” and not distribute the cash immediately. These requirements add some administrative inconvenience to the vertical season-and-sell structure.
While strong authorities exist that respect corporate subsidiaries and decline to find them as the agents of their parents, the vertical season-and-sell structure does not really reflect the parent-subsidiary relationships of those cases. Rather than a parent engaged in business and conducting transactions with its subsidiaries, which are themselves engaged in independent businesses, in a vertical season-and-sell structure, the sponsor manages each of the parent and the subsidiary to minimize the tax of the subsidiary. That difference might bring the vertical season-and-sell structure outside the typical parent-subsidiary agency cases. And as Leblang, Paulos, and Fieldston explain, outside the parent-subsidiary context, courts are far more likely to find agency when control exists. The sponsor does control both the parent and the subsidiary.
Under a leveraged blocker structure, non-U.S. investors invest in an offshore feeder, typically organized as a Cayman Islands partnership and treated as a partnership for U.S. tax purposes. The offshore feeder, in turn, organizes a U.S. blocker, frequently as a U.S. limited partnership or limited liability company that checks the box to be treated as a corporation for U.S. federal tax purposes. The U.S. blocker invests capital into a master partnership, usually organized as a U.S. limited partnership, into which U.S. investors also invest. As capital is called by the offshore feeder from investors, the offshore feeder simultaneously makes loans and invests equity capital into the U.S. blocker, which the U.S. blocker, in turn, invests as equity into the master partnership. U.S. investors may invest directly in the master partnership or invest in an onshore feeder, typically organized as a U.S. limited partnership that is treated as a partnership for U.S. tax purpose, which in turn invests in the master partnership.
Over the life of the structure, the U.S. blocker pays (or accrues) interest to the offshore feeder. Optimally, the U.S. blocker would pay no dividends, but would instead reinvest any profits until it liquidates. The interest payments made by the U.S. blocker to the offshore feeder reduce the U.S. blocker’s taxable income. Those interest payments are structured to benefit from the portfolio interest exemption so they are not subject to U.S. withholding tax. The debt-equity ratio and the interest rate of the U.S. blocker’s debt is typically set by an accounting firm under a transfer pricing model.
If a fund’s investments yield is 14%, the U.S. corporation is funded 75% with debt and 25% with equity, the interest rate on the debt is 8%, and the aggregate federal and state tax rate is 24%, then a leveraged blocker can reduce the taxes that would otherwise be imposed on an unlevered corporation by 43% from 24% (taking into account state taxes) to 13.7%.
Typically, no foreign investor will own ten percent or more of the offshore feeder. This is the easiest way to ensure that the interest paid by the U.S. blocker to the offshore feeder qualifies for the portfolio interest exemption from withholding tax.
If a foreign investor owns ten percent or more of the offshore feeder (and therefore ten percent or more of the voting stock of the U.S. blocker), interest paid by the U.S. blocker to the offshore feeder and allocable to the foreign investor would not be eligible for the portfolio interest exemption. To address this, the U.S. blocker could be “decontrolled.” Decontrol requires the sponsor to invest some amount (often one to two percent of the debt and equity of the U.S. blocker) in exchange for all of the voting stock of the U.S. blocker. If no foreign investor owns ten percent or more of the voting stock of the U.S. blocker (and no foreign investor is a CFC that is related to the U.S. blocker), the interest would appear to qualify for the portfolio interest exemption.
By investing an amount equal to one to two percent of the capital of the U.S. blocker in exchange for all of the U.S. blocker’s voting stock, the sponsor hopes to demonstrate that it is the real owner of the voting stock and that it will exercise its vote solely to benefit itself and not to benefit the investors. Nevertheless, a sponsor that exercises its vote to benefit itself rather than its investors might hear loud complaints from investors and might have trouble raising the next fund. If the sponsor is deemed to exercise its vote on behalf of the investors, the investors might be deemed to hold the voting stock of the U.S. blocker, thereby subjecting the ten percent or greater foreign investors to U.S. withholding tax on interest.
The repeal of downward attribution in the Tax Cuts and Jobs Act (TCJA) presents a further challenge to a decontrol structure. Section 881(c)(3)(C) denies the portfolio interest exemption for interest received by a CFC from a “related person” as defined in section 864(d)(4). Section 864(d)(4), in turn, provides that a related person includes any “United States shareholder” of a CFC, and any person who is a related person (within the meaning of section 267(b)) to such a United States shareholder.
So, assume a foreign parent corporation has a foreign subsidiary and that foreign subsidiary owns all of the nonvoting stock of a U.S. corporation.
A foreign corporation is a CFC if more than 50% of the total combined voting power of all classes of stock entitled to vote or the total value of the stock of the corporation is owned directly, indirectly, or constructively by “United States shareholders” on any day during the taxable year of the foreign corporation. A “United States shareholder” is a U.S. person that owns directly, indirectly, or constructively ten percent or more of the total combined voting power or value of the foreign corporation.
Section 958 provides indirect and constructive stock ownership rules that deem a taxpayer to own stock that it does not own for purposes of determining whether it is a “United States shareholder” of a foreign corporation and whether a foreign corporation is a CFC. For these purposes, section 958(b) requires taxpayers to apply the attribution rules of section 318(a), which generally treat a shareholder as constructively owning stock owned by taxpayers with whom the shareholder has certain specified relationships. Under the “upward attribution” rules of section 318(a)(2)(A), (B), and (C), as modified by section 958(b)(2), if 50% or more in value of the stock of a corporation is owned by an entity, that entity is considered as owning all of the stock owned by that corporation. Under these rules, in the hypothetical described above, the foreign parent is deemed to own the stock of the U.S. blocker owned by its foreign subsidiary.
Under the “downward” attribution rules of section 318(a)(3)(A), (B), and (C), stock owned by a person (such as the foreign parent) is attributed to certain partnerships, estates, trusts, and corporations in which the person has an interest. Under these rules, because the foreign parent owns the stock of the foreign subsidiary and is treated as owning the stock of the U.S. blocker under the upward attribution rules, the downward attribution rules treat the U.S. blocker as owning the stock of the foreign subsidiary owned by the foreign parent.
Prior to the TCJA, section 958(b)(4) prevented the “downward attribution” of stock held by a foreign person to a U.S. person. Section 958(b)(4) was repealed by the TCJA. Section 958(b)(4) would have prevented the U.S. blocker from being treated as owning the stock of the foreign subsidiary.
In contrast, following the repeal of section 958(b)(4), the U.S. blocker is treated as owning the stock of the foreign subsidiary. As a result, the foreign subsidiary is a CFC, the U.S. blocker is a United States shareholder of the foreign subsidiary, and the two are related. Therefore, the interest paid by the U.S. blocker to the foreign subsidiary fails to qualify for the portfolio interest exemption and is subject to a 30% U.S. withholding tax. This result was not intended, but it has not been repealed.
For investment structures where this is an issue, the can could be kicked down the road by providing that any interest will be payable upon maturity in the hope that Congress will eliminate downward attribution in these circumstances before the interest is payable. That, however, raises the question of whether the U.S. blocker is entitled to deductions before the amounts are paid.
Business development companies (BDCs) are ’40 Act companies that are treated as “regulated investment companies” (RICs) for federal income tax purposes. BDCs offer the most perfect tax vehicle for U.S. loan origination by foreign investors. As a RIC, a BDC is a corporation for U.S. federal income tax purposes and therefore serves as a “blocker,” meaning that foreign investors need not file tax returns or pay taxes for the activities of the BDC. So long as a BDC distributes all of its income each year, it can avoid a corporate-level tax. And so long as its loans qualify for the portfolio interest exemption (which they almost always do), a BDC’s dividends attributable U.S. source interest earned by the BDC also escape U.S. withholding tax. However, this exemption does not extend to dividends attributable to foreign source interest earned by the BDC. Accordingly, a BDC is not an appropriate vehicle to originate non-U.S. loans. As mentioned above in Part I, BDCs are tax-free U.S. loan-originating machines, but the tax freedom comes with regulation. And regulation is often worse.
Nearly 20 years ago, I asked whether a foreign fund that originates loans is engaged in a trade or business in the United States. The Service eventually answered yes. But that answer has not stopped foreign funds from originating loans and has not yielded significant tax revenue.
Existing law principles do not provide a coherent framework for taxing origination activity. First, there is no defensible policy basis for treating the secondary market purchase of a loan differently than a loan origination. Lending is no longer a capital-intensive business, and regardless of whether a bank originates a loan and then sells it to a fund or a fund originates the loan itself and holds it, U.S. taxpayers (the bank in the first case and the U.S. investment manager in the second) are taxed equivalently at demonstrably arm’s-length prices for the services rendered. And in neither case does the fund generate any goodwill. The bank in the first case and the investment manager in the second hold all of the goodwill generated by the activity.
Second, there is no legal or historic basis on which to tax foreign loan-originating funds. The trading in securities safe harbor defines a security to include “any note, bond, debenture, or other evidence of indebtedness.” That definition includes a loan for good reason. In the 1960s, there wasn’t the slightest bit of concern with foreign lenders competing with U.S. banks, and because any foreigner that held a U.S. loan would have been subject to a 30% withholding tax on interest, the government was happy to exempt foreign lenders from net income tax in order to subject them to a gross withholding tax. There was a concern with competition in the 1980s, but only with respect to foreign banks. In 1984, Congress subjected only foreign banks to interest withholding.
The absence of any comprehensive guidance or any coherent policy, legal, or historic basis to tax foreign nonbank lenders has led to a vacuum that was filled by the various work-arounds described above. Most of them rely on very technical readings of authorities, principles developed decades ago in different contexts, or arbitrary distinctions. Seasoning structures rely on the absence of agency, a common law principle dating back hundreds of years. Treaties are also decades old. The concept of bullets derives from authorities that the Service has disclaimed. Treating a loan originated 48 hours previously as a secondary market loan is arbitrary.
Since the past is the best predictor of the future, it is highly likely that the current patterns will continue: Foreign funds will increasingly originate loans, new structures will arise, and no new guidance will be issued. Nevertheless, alternative paths are possible.
It is possible that the Service will challenge foreign funds that directly lend without relying on one of the structures described above. Such a challenge happened in Chief Counsel Attorney Memorandum 2009-010. But it did not deter foreign funds from using the structures described above. If the Service were to challenge one of the structures described above, it would deter sponsors and investors from that particular structure and shift them to a different one.
Finally, Congress could step in and clarify or expand (depending on your point of view) the securities trading safe harbor so that it includes origination activity.
The Managed Funds Association and the Real Estate Roundtable have proposed the following changes to include origination activity in the securities trading safe harbor explicitly, exclude it explicitly from the definition of commercial activity for purposes of section 892, and provide a conforming change to section 475:
1. Amend section 864(b)(2) by adding the following new sub-paragraph:
(2)(C) Lending or restructuring loans for taxpayer’s own account.
(i) Making a new loan or otherwise extending new credit to a borrower for the taxpayer’s own account, including negotiating and structuring the terms thereof, whether by the taxpayer or his employees or through a resident broker, arranger, or other agent, and whether or not any such employee or agent has discretionary authority to make decisions in effecting the transactions.
(ii) Negotiating and restructuring the terms of an existing loan or other extension of credit to a borrower for the taxpayer’s own account, including a loan or extension of credit that was not originally made by the taxpayer, whether by the taxpayer or his employees or through a resident broker, arranger, or other agent, and whether or not any such employee or agent has discretionary authority to make decisions in effecting the transactions.
This sub-paragraph (2)(C) shall not apply in the case of (i) a dealer in securities or (ii) a creditor that receives fees or other compensation from other creditors for performing services in the ordinary course of a trade or business, such as originating, syndicating, participating, or otherwise arranging, managing, or restructuring a loan or extension of credit for such other creditors. For purposes of this sub-paragraph (2)(C), a “dealer in securities” means a merchant of securities, with an established place of business, that regularly engages as a merchant in purchasing stocks or securities and selling them to customers with a view to the gains and profits that may be derived therefrom.
2. Re-number existing sub-paragraph 864(b)(2)(C) as 864(b)(2)(D).
3. Modify section 475(f)(1)(D) as follows:
(D) Other rules to apply. Rules similar to the rules of subsections (b)(4) and (d) shall apply to securities held by a person in any trade or business with respect to which an election under this paragraph is in effect. Subsection (d)(3) shall not apply under the preceding sentence for purposes of applying sections 864(c)(2), 1402 and 7704.
4. Add new sub-paragraph 892(a)(2)(C) as follows:
(2)(C) Lending or restructuring loans for taxpayer’s own account.
The activities described in section 864(b)(2)(C) (whether within or outside the United States) shall not constitute commercial activities.
It makes sense to amend the trading in securities safe harbor. Nevertheless, this proposal is both too broad and too narrow.
First, as drafted, the proposal would allow foreign banks to set up a branch in the United States and offer mortgage loans (and generate goodwill) without being subject to any U.S. federal income tax. On the other hand, under the proposal, if a foreign fund originates a loan, participates it out, and retains a fee, not only would that fee be subject to U.S. federal income tax, but the fund would be kicked entirely out of the safe harbor (and presumably all of its income could be taxed).
I suggest a different approach. The defining feature of credit funds is that they are passive investment funds managed by independent investment managers and do not generate goodwill. REITs are similar passive vehicles. To ensure that REITs do not engage in an active business, the REIT rules require, in certain circumstances, that the potentially active business be conducted through an “independent contractor” that (1) bears the cost of the services, (2) is adequately compensated, and (3) makes a separate charge for the services that are provided and retains the compensation. The purpose and operation of the independent contractor test is similar to the independent agent test in treaties. If a principal does not economically control its agent, then the market and not the principal will set the price for the agent’s services, and the agent will be adequately compensated for its services. Because an independent agent is fully compensated for its services (and the independent agent’s jurisdiction can tax that compensation), the principal cannot capture any value generated by the services, and the independent agent’s jurisdiction need not tax the principal. By the same token, since an independent contractor is fully compensated for its services, a REIT cannot profit from those services and remain a passive investor.
The independent contractor test is a little less restrictive than the independent agent test and is more appropriate in the manager context. The independent agent test requires a manager to have multiple clients, and the compensation from any one client (at least under common practice) cannot represent more than 25% of the manager’s revenue. Start-up managers cannot qualify even though there is a large and established market price for their services. The independent contractor test as applied to REITs has some quirks. For example, a corporation that serves as the investment manager to a REIT cannot qualify as an independent contractor. In contrast, a wholly-owned subsidiary of the REIT’s investment manager can qualify as an independent contractor. To avoid these anomalies, I would modify the independent contractor test so that it is based on economic interest alone. To avoid confusion with the REIT test, and to better describe it, a manager that satisfies the test would be called a “nondependent agent.”
An entity would be deemed a nondependent agent only if (1) it does not own, directly or indirectly, a greater than 35% economic interest in the fund, (2) not more than 35% of the economic interest of the entity is owned, directly or indirectly, by one or more persons owning a 35% or greater economic interest in the fund, and (3) the fund does not derive or receive any material amount of net income from the entity.
Requiring origination activity to be conducted through a nondependent agent would prevent a fund from conducting a banking business in the United States or otherwise generating goodwill. It could also allow the safe harbor to be extended to funds that participate out the loans that they originate. Participating out a loan is a service for which the U.S. investment manager would be fairly compensated, and taxed. So long as the participants are not customers of the fund (and they ordinarily would not be), the fund need not be taxed on the income. In fact, if U.S. investment managers are conducting the activity, are receiving fair market value fees for the service, and are taxed on the income, then there would be no reason why the safe harbor could not be extended to dealer activity.
I would, therefore, modify the proposal to amend section 864(b)(2) as follows:
(2)(C) Lending or restructuring loans for taxpayer’s own account.
(i) Making a new loan or otherwise extending new credit to a borrower for the taxpayer’s own account, including negotiating and structuring the terms thereof, through a nondependent agentwhetherby the taxpayer or his employees or through a resident broker, arranger, or other agent, and whether or not any such nondependent agent has discretionary authority to make decisions in effecting the transactions.
(ii) Negotiating and restructuring the terms of an existing loan or other extension of credit to a borrower for the taxpayer’s own account, including a loan or extension of credit that was not originally made by the taxpayer, through a nondependent agentwhetherby the taxpayer or his employees or through a resident broker, arranger, or other agent, and whether or not any such nondependent agentemployeeor agenthas discretionary authority to make decisions in effecting the transactions.
This sub-paragraph (2)(C) shall not apply in the case of (i) a dealer in securities or (ii) a creditor that receives fees or other compensation from other creditors for performing services in the ordinary course of a trade or business, such as originating, syndicating, participating, or otherwise arranging, managing, or restructuring a loan or extension of credit for such other creditors.For purposes of this sub-paragraph (2)(C),a “dealer in securities” means a merchant of securities, with an established place of business, that regularly engages as a merchant in purchasing stocks or securities and selling them to customers with a view to the gains and profits that may be derived therefrom.”a“nondependent agent” is a person (1) that does not own directly or indirectly, a greater than 35% economic interest in the taxpayer, (2) not more than 35% of the economic interest of which are owned, directly or indirectly, by one or more persons owning a 35% or greater economic interest in the taxpayer, (3) with respect to which the taxpayer does not derive or receive any material amount of net income, (4) that bears the cost of its negotiating, restructuring, and investment management services, and (5) is adequately compensated for those services.
Re-number existing sub-paragraph 864(b)(2)(C) as 864(b)(2)(D).
The author thanks Jean Bertrand, Janicelynn Asamoto Park, and Tony Meyer for their comments and Xiaoyang (Winnie) Ma for her cite checking.
