Planning for an existing foreign non-grantor trust that has one or more U.S. beneficiaries presents complex challenges for trustees and professional advisors. Merely understanding the relevant U.S. tax and reporting rules is daunting enough. But, trustees and advisors must proceed further and incorporate these difficult rules into the overall fiduciary decision-making process on appropriate investment and distributional policies for the trust.1 While abundant literature exists on the details of the U.S. tax rules applicable to a foreign non-grantor trust,2 little or no guidance is available regarding how the U.S. tax rules tend to interact over time with a trust's investment allocation and distributional policies, or whether any opportunities or pitfalls that exist are significant or relatively small.3

We'll introduce a quantitative analysis of foreign non-grantor trusts and investigate the interaction of investment allocation, currency selection, distributional policy and U.S. tax rules in the context of a case study. Our research investigates tax-effective methods for transferring wealth from a foreign trust to U.S. beneficiaries.4

Our research indicates that distributing current year income from a foreign trust represents a more tax-effective strategy than many trustees or practitioners realize, potentially moving 90 percent of the foreign non-grantor trust's assets onshore in approximately 40 years. We also found that increasing the foreign non-grantor trust's allocation to equities may accelerate the amount and pace of transfer. But perhaps most significantly, we found that foreign non-grantor trusts facing a short time frame can significantly accelerate the transfer of wealth onshore by use of the so-called “default” method of distributing current year income.

The Santos Family

To bring this analysis to life, we'll focus on a fictional trust and its beneficiaries. The simple yet common fact pattern will allow us to set the general planning stage and highlight the operative variables. Of course, you must be cognizant that even seemingly small variations in the underlying circumstances may produce very different considerations or issues.

The members of the Santos family have long been citizens and residents of Nirvana, a fictional country that has no income or transfer taxes.

  • U.S. beneficiaries: Virginia Santos moved to the United States many years ago, where she married a U.S. citizen and had three children (all U.S. citizens). Virginia, now age 50, has a green card, which makes her a U.S. resident for federal income tax purposes. She also qualifies as a U.S. resident for U.S. transfer taxes by virtue of being domiciled in the United States. She is a successful lawyer whose income is more than sufficient to meet the needs of her immediate family.

  • The foreign non-grantor trust: Aunt Imelda, Virginia's maternal aunt, died 40 years ago wealthy and childless in Nirvana, leaving behind several large trusts, including one for her niece Virginia, known as the Virginia Santos Family Trust (VSF trust). The VSF trust is a foreign non-grantor trust that currently holds $100 million, invested in a 60 percent stocks/40 percent bonds allocation. (See “All in the Family,” p. 55.)

Virginia has recently accepted that she's likely to remain a U.S. taxpayer for the rest of her life. And the future beneficiaries of the VSF trust — her children — are currently U.S. citizens who are subject to worldwide U.S. taxation no matter where they reside. So, Virginia has decided to assemble a team of advisors to work with the Nirvana Trust Company to lay out the U.S. tax planning options available for the VSF trust. While a complete review of the relevant U.S. tax rules lies beyond our purpose here, a quick primer on the basic concepts will help set the table for the analysis that follows.

Overview

As summarized in “Downside of a Foreign Trust, p. 56,” the U.S. tax rules applicable to a foreign5 non-grantor trust effectively divide the assets of the foreign trust into three portions: 1) current year income, 2) accumulated income, and 3) the remaining balance of the trust.6 Distributions from the foreign trust to a beneficiary (whether the beneficiary resides in the United States or not) are generally deemed to be sourced as follows: first, from the trust's current year income until exhausted; next, from accumulated income; and finally, from the balance of the trust.

Receipt of current year income by a U.S. beneficiary is subject to normal U.S. income tax treatment. That is, the U.S. beneficiary pays U.S. income taxes on the items of current year income as if she had received such items directly, with normal tax rates applying to ordinary income, capital gains and so on.

Receipt of accumulated income by a U.S. beneficiary incurs two special U.S. taxes:

  • The so-called “throwback tax” that basically subjects a distribution of accumulated income to taxation at ordinary U.S. income tax rates (assumed at 39.6 percent).7 For all items of income (other than tax-exempt income), the throwback tax entails a loss of the character the income had when originally accumulated in the foreign trust. Thus, an item of capital gain accumulated in the foreign trust will not benefit from the lower tax rates normally allowed for capital gains.8

  • A nondeductible interest charge that addresses the time value of the delay in U.S. tax payment caused by the offshore accumulation in the foreign trust.9

Receipt by a U.S. beneficiary of a distribution of the remaining balance of the trust generally incurs no U.S. tax consequence.

Virginia and her U.S. advisors quickly conclude that the throwback tax and interest charge represent an enormous threat to the VSF trust. Indeed, the VSF trust, after 40 years of offshore investment will be deemed to have accumulated so much income subject to the throwback tax and interest charge that a current distribution of all trust assets to Virginia (or any other U.S. person) would be entirely consumed by those special taxes.10

Luckily for Virginia, she doesn't need a large distribution. But with the imminent marriage of her eldest child, she's begun to think about a long-term plan for maximizing the benefit of the foreign trust.

Virginia, her advisors and the trustee begin to focus on the fact that the throwback tax and interest charge apply only to distributions of accumulated income, and don't apply to distributions of current year income. Since distributions from a foreign trust are deemed to come first out of its current year income, a foreign trust may safely distribute current year income every year to U.S. beneficiaries without incurring adverse U.S. tax consequences.

To understand how distributing the VSF trust's current year income might work over time, Virginia and her team need to understand that the relevant U.S. tax rules allow at least three different options for identifying the current year income of a foreign trust. Here we will only look at two such options: (1) distributable net income (DNI), and (2) the so-called “default” method.11

Annual Distribution of DNI

The most common option for calculating the current year income of a foreign non-grantor trust is to use standard U.S. tax law concepts to identify all items of the income for U.S. tax purposes earned in the trust's current calendar year. We'll refer to this definition of current year income as DNI, using the Internal Revenue Code's terminology. For a trust holding a typical investment portfolio, DNI will include all dividends, interest and realized capital gains.12

Before looking at what distributions of DNI from the VSF trust might accomplish annually, let's briefly consider U.S. transfer tax planning.13 If the VSF trust is structured so that its assets wouldn't be includible in Virginia's estate (or any other U.S. beneficiary's estate for that matter) for U.S. estate tax purposes at death, then the trustee will want to avoid making outright transfers to Virginia or other U.S. beneficiaries of assets not needed for immediate spending.14 While foreign trust instruments and applicable trust laws differ in the powers provided to trustees, it's often true, as we assume with the VSF trust, that the trustee may make distributions either outright to beneficiaries or to new or existing trusts for the beneficiaries. In this case, we want to begin to make distributions of current year income into the United States, but not lose the U.S. transfer tax benefits of keeping these assets in trust. We have analyzed what happens if the VSF trust begins to make annual distributions to a new U.S. trust for Virginia's benefit.

To understand this strategy's probable success, we used a proprietary system designed to project wealth over time, using historical data, research and sophisticated modeling software. Our system creates a vast range of potential market returns, taking into account the underlying drivers and linkages in the capital markets, as well as a degree of randomness. It generates 10,000 possible outcomes, which we show as a probability distribution known as a “box and whiskers” display. In “Distributing Current Year Income,” p. 57, the median of outcomes, or 50th percentile, is indicated by the blue circle in the middle of each box; the 90th percentile — which represents very poor market performance, is shown at the bottom of the box; and the 10th percentile at the top of the box. The whiskers extend the distributions to the 95th and 5th percentiles. (See Notes on Wealth Forecasting System in “Distributing Current Year Income,” p. 57.)

“Distributing Current Year Income” shows the range of outcomes over 40 years assuming the trustee of the VSF trust makes annual distributions of its DNI to a U.S. trust. As the time series demonstrates, the annual distribution of DNI effectively prevents the foreign trust from growing, setting its inflation-adjusted value on a mild decline. The U.S. trust, in turn, receives each distribution, pays normal U.S. income taxes on the distributed income (no throwback tax or interest charge), invests the after-tax proceeds in a 60/40 portfolio and begins to grow rapidly. The overall result appears to accomplish Virginia's goals — at least partially — since she and her descendants gain access to the benefits of a burgeoning U.S. trust that is funded from the foreign trust without the imposition of any punitive U.S. throwback taxes and interest charges.

But how successful is the strategy, really? It seems that annual distributions of the trust's DNI don't significantly reduce the size of the VSF trust, which remains very large at the end of 40 years. Virginia and her descendants would prefer to have access to the entire value of the foreign trust, not just its current year income over time. Counterintuitively, however, this perspective misses an important point. If, instead of looking at the size of the foreign trust in dollars, we look at where the total wealth comes to be located over time — whether in the foreign trust or the U.S. trust — we can isolate the real benefit of distributing DNI annually. “Shifting Wealth,” p. 58, shows the percentage of the total wealth held in trust that comes to be held in the U.S. trust over time.

Perhaps surprisingly, distributing the foreign trust's DNI annually for 20 years will succeed in moving more than 60 percent of the total wealth into the United States in the median case. Making these distributions for 40 years will shift almost 90 percent of the total wealth into the United States in the median case. So, a strategy that at first appears to leave most of the starting value of the foreign trust trapped offshore in fact allows a growing percentage of the total wealth to be brought into the United States. And since distributions of DNI don't incur the throwback tax or interest charge under current law, the U.S. tax rules that were designed to address the benefit of offshore growth in foreign trusts don't come into play.

Clearly, time represents an underappreciated but critical tool in managing any existing or prospective foreign trust that has, or will have, large amounts of accumulated income. Virginia, if her time horizon for benefiting from the trust can be measured in decades, not years, may request that the foreign trust distribute DNI only, thereby preserving much of the benefit of the VSF trust that her aunt created many years ago.

Default Method

As shown in the preceding discussion, a strategy of making annual distributions of a foreign trust's DNI requires time — potentially decades in our analysis — to shift significant wealth from the foreign trust into the United States. But what if Virginia or her family can't wait? Is there a faster way to shift wealth from a foreign trust into the United States without punitive taxation?

Happily, there is. The IRS allows a second option for defining the current year income of a foreign trust — this option is commonly known as the “default method.”15 This option employs a formula that ignores the actual income of the foreign trust and instead focuses on the history of distributions. The formula holds that, where the total distributions received by a U.S. beneficiary in a calendar year don't exceed the average annual distribution received over the preceding three years, plus 25 percent, the distribution represents the foreign trust's current year income.16 The formula effectively permits distributed income to increase every year by roughly 12 percent. Because such an increase is unlikely to be matched by a corresponding increase in the growth rate of the foreign trust's investment portfolio, a U.S. beneficiary's election into the default method will often allow distributions of income to completely empty the foreign trust.

Consider “Increased Pace of Wealth Transfer,” this page, which shows the VSF trust distributing maximum income under the default method for 40 years. Again, we've assumed that distributions are made to a U.S. trust and not to Virginia outright, to preserve the U.S. transfer tax benefits of keeping assets in trust.

As predicted, the distribution of default method income completely exhausts the foreign trust by year 20 in the median case. This might, by itself, seem to promise an improvement in how much wealth can be transferred to Virginia and her family. Yet, a comparison with “Distributing Current Year Income” indicates that the U.S. trust receives less wealth ($307 million versus $365 million in the median case) at year 40 under the default method than it did receiving DNI.

The default method is unfavorable at year 40 due to its U.S. tax cost. All income calculated under the default method is deemed to be ordinary income in the hands of the U.S. beneficiary, taxable up to the highest marginal income rate (assumed at 39.6 percent).17 The default method entails a loss of the character of the income earned by the foreign trust. And after 40 years, this cost overtakes the benefit of the default method — the speed at which it empties the foreign trust — producing an unfavorable result for the U.S. beneficiaries.

What about a shorter time frame? Compare “Distributing Current Year Income,” and “Increased Pace of Wealth Transfer,” again, focusing this time on the results at year 20. At this point, again looking at median outcomes, the default method has transferred $152 million, while the distribution of DNI has transferred only $137 million. As you can see, the benefit of the default method begins to shine as we focus on shorter time frames.

But we still haven't isolated the greatest benefit of the default method, which lies in its adaptability to different time frames. This adaptability arises from the fact that the level of wealth transfer allowed by the default method is keyed to a three-year rolling average of distributions. With our modeling, it's possible to determine whether any given starting point for this rolling average will be sufficient to initiate a stream of current year income distributions that will empty the foreign trust within the available time frame. If not, consider priming — or increasing — that rolling average by making immediate additional distributions. Indeed, the trustee of Virginia's foreign trust can reset the rolling average by making an immediate distribution to her. In her case, such a distribution will largely be lost to the throwback tax and interest charge. But the distribution will nonetheless enter into the three-year rolling average and immediately elevate what can be distributed as current year income going forward under the default rule.

To showcase the adaptability of the default method, let's assume that the VSF trust will terminate 10 years from now, so Virginia wants a plan to derive maximum benefit over that time period. As shown on the left side of “Maximizing Wealth,” this page, if the trustee simply distributes the DNI for 10 years, Virginia will receive approximately $55 million in the median case.18

Interestingly, the default method employed with no planning produces a worse result, dropping her median outcome to $47 million. However, our modeling can determine whether priming the pump of the default method might improve results. So we tested four scenarios in which the trustee makes increasingly larger up-front distributions (shown as a multiple of the trust's DNI) to prime the rolling average of distributions prior to election into the default method. And it turns out that a distribution of four times the VSF's current year income (4 x prime) initiates a rolling average of distributions that optimally depletes the VSF trust over the 10 years available and produces by far the best result, improving Virginia's median outcome to $87 million. Priming less fails to empty the VSF trust optimally, eroding the outcome, while priming more empties the VSF trust too quickly, producing too high a U.S. tax cost.

“Optimal Planning,” this page, identifies the optimum distribution strategy over a range of possible time frames.19 As we've seen, the default method with some level of priming will often produce the optimum result over shorter time frames. For longer time frames, the default method becomes too costly and strategies based on distributing DNI will be superior.

We hope that this brief quantitative analysis of the complex subject of optimizing the distribution of current year income from foreign trusts to U.S. beneficiaries has at least demonstrated that there are more options and tools than commonly realized. Within the context of any particular trust or situation — such as our case study with the VSF trust — we believe that this kind of analysis is critical to understanding the interaction of asset allocation, distributional policy and U.S. tax laws and fashioning the set of solutions most appropriate for meeting the fiduciary obligations of the trustee while optimizing the benefit of the trust to the U.S. beneficiaries.

Bernstein Global Wealth Management does not provide tax, legal or accounting advice. In considering this material, readers should discuss individual circumstances with professionals in those areas before making any decisions.

Endnotes

  1. As of this writing, (October 2010), the future direction of U.S. tax laws and policy remains highly uncertain. In this study, we've assumed that the U.S. income tax rates and rules will conform to the law as written at the beginning of 2010, including any changes that were then scheduled for 2011 and after. In general, this means that U.S. income tax rates on ordinary income are assumed to be 39.6 percent, qualified dividend treatment is unavailable and the rate on capital gains is 20 percent. All federal tax and legislative changes enacted since Jan. 1, 2010, haven't been included in our analysis, including the 3.8 percent payroll tax on unearned income scheduled to begin in 2013 as part of the recent universal healthcare reform legislation (Patient Protection and Affordable Care Act).
  2. See, e.g., Amy P. Jetel, “When Foreign Trusts Are Non-Grantor,” Trusts & Estates, April 2008, at p. 53.
  3. An important caveat relates to the recent crackdown by Congress and the Internal Revenue Service on the tax rules and reporting requirements applicable to certain foreign accounts and entities directly or indirectly used by U.S. taxpayers. Although Congress has paid periodic attention to this issue for decades, the current environment has seen the passage of some of the most important offshore tax legislation in a generation, particularly the Foreign Account Compliance Act provisions recently enacted as part of the Hiring Incentives to Restore Employment Act (the HIRE Act) on March 18, 2010. The prudent planner should expect greater than normal legislative uncertainty going forward, with a strong trend towards the application of more rigorous U.S. tax reporting and penalty provisions to foreign non-grantor trusts over time.
  4. We emphasize that this study doesn't focus on the numerous U.S. reporting provisions applicable to the U.S. beneficiaries of foreign non-grantor trusts.
  5. To qualify as a U.S. trust under U.S. tax law, a trust must satisfy both of the following tests: 1) a court within the United States is able to exercise primary supervision over the trust's administration; and 2) one or more U.S. persons have the authority to control all substantial trust decisions. See Internal Revenue Code Section 7701(a)(30)(E). Any trust that fails either or both of these tests qualifies as a foreign trust for U.S. tax purposes. See IRC Section 7701(a)(31)(B).
  6. Current year income includes interest, dividends and capital gains realized in the current calendar year. Accumulated income includes all interest, dividends and capital gains realized by the foreign trust in prior calendar years and not previously distributed. The balance of the trust includes the original principal transferred to the trust and unrealized gains.
  7. See supra, note 1.
  8. The “throwback” rules (IRC Sections 665 through 668) are extremely complex, and Congress repealed them for domestic trusts — but not for foreign trusts — in 1997. Much of the complexity of the rules is geared toward making fine adjustments for changes in a beneficiary's marginal rates of tax over time. Thus, when one assumes — as we do in this study — that the U.S. beneficiary is subject to U.S. income tax at the highest marginal tax rates, then the effect of the throwback tax is relatively simple, transforming accumulated income distributed from the foreign trust into ordinary income taxed at the highest marginal rate, as noted in the text.
  9. For periods since 1996, the interest charge is based on the floating rates charged under IRC Section 6621 on underpayments of tax, compounded daily. Figuring the number of years of interest requires a calculation of the dollar-weighted average of years for all accumulated income held by the foreign trust and not previously distributed. Different interest calculations are required for periods up to 1996, when the current law was enacted.
  10. Because the trustee holds trust records that go back only seven years, nobody knows what percentage of the trust assets reflects the original principal contributed by Aunt Imelda 40 years ago or how much income and gains have been accumulated since then. The absence of records would mandate use of the “default” method for calculating the U.S. income tax consequences of a complete distribution of the Virginia Santos Family trust (VSF) to Virginia. As a result, nearly all of the liquidating distribution would be treated as accumulated income subject to the throwback tax, and the interest charge would be calculated over one-half the number of years the trust had existed as a foreign non-grantor trust — or 20 years. The resulting tax and interest charge would consume the trust.
  11. A third option for calculating a foreign trust's current year income is to use the definition of “income” used for trust law purposes, as set forth in the trust instrument or under the trust laws of the relevant jurisdiction. Many professionals refer to this definition of income as “fiduciary accounting income.” Subject to restrictions, if a trustee of a foreign non-grantor trust distributes no more than the trust's current year fiduciary accounting income, the distribution will not be deemed to include any accumulated income subject to the throwback tax and interest charge. See IRC Section 665(b). We didn't include an example of fiduciary accounting income in this study for reasons of space and simplicity, but note that, with appropriate guidance from professional advisors, our Wealth Forecasting System can employ several different definitions of fiduciary accounting income for purposes of investigating this option.
  12. While capital gains are typically not included in distributable net income (DNI) for domestic trusts, foreign trusts are required to include realized capital gains in DNI. See IRC Section 643(a)(6)(c).
  13. While this article focuses on U.S. income tax planning for foreign non-grantor trusts, our broader research determined that planning for U.S. gift, estate and generation-skipping transfer (GST) taxes represents the most significant planning opportunity for the foreign donor wishing to benefit U.S. individuals. (See our publication, “Bring It On Home”, June, 2010 at www.bernstein.com). As noted in the text, trustees of existing foreign trusts should typically avoid making outright transfers to U.S. beneficiaries when possible since an outright transfer to a U.S. beneficiary who doesn't need it for spending is potentially exposed to future U.S. transfer taxation when the U.S. beneficiary later gives the wealth away or dies. By contrast, wealth left by a foreign donor in a properly drafted trust, whether such trust is foreign or domestic, should escape all future U.S. transfer taxation for as long as the trust can be made to endure. The U.S. transfer tax savings made possible by a foreign donor's effective use of a multigenerational trust become very large over time and contrast sharply with the limited planning opportunity available to U.S. donors seeking to accomplish the same objective within the constraints of the GST tax.
  14. Note that the VSF trust isn't subject to GST tax because Aunt Imelda, the only transferor of the VSF trust, wasn't a U.S. citizen or domiciliary at the time of any funding transfers to the trust and these transfers didn't include any U.S. situs property subject to U.S. transfer tax. See Treasury Regulations Section 26.2663-2.
  15. The default method is set forth in Notice 97-34, 1997-1 C.B. 422 and IRS Form 3520 (2009) and was designed to address cases in which a U.S. beneficiary doesn't receive sufficient information from a foreign trust to determine the proper tax treatment of a distribution. However, the method is available to any U.S. beneficiary who makes a proper election. Once the election is made, it's irrevocable as to subsequent tax years (except the final year in which the trust terminates).
  16. Here's the formula expressed as an equation: Current year income received by U.S. beneficiary in year four = (total distribution received by U.S. beneficiary in year one + total distribution from year two + total distribution from year three) X 1.25/3. For example, consider a foreign trust that distributes $100 every year for three years to a U.S. beneficiary. If the U.S. beneficiary elects into the default method in year four, then the maximum amount that the foreign trust may distribute in years four through seven as current year income would be determined as follows: Year: 1, 2, 3, 4, 5, 6, 7/Distribution: $100, 100, 100, 125, 135, 150, 171.
  17. See supra, note 1.
  18. Note that Virginia will receive at most the total current year income distributions made during the 10 years prior to termination, since on the trust's termination, any remaining trust assets payable outright to her or to any other U.S. individual will include the trust's accumulated income and thus will be lost to the throwback tax and interest charge, as noted in note 8, supra.
  19. While we use the word “optimal” in our comparison of the distribution strategies considered in this analysis, the reader should understand that, depending upon the facts and circumstances of each particular foreign trust, it's possible that a number of other investment and distribution strategies may be available that could improve the outcomes.

Andrew Auchincloss is a director in Bernstein's Wealth Management Group in New York City