An Adobe PDF version of this document is available here.

August 15, 2013

Dear Ladies and Gentlemen:

Thank you again for meeting with representatives of ACTEC to discuss selected provisions of the Green Book proposals.

Enclosed are the follow up memos you requested.  Please let me know if you have any questions or would like to schedule a call or meeting to explain our comments in more detail.

Sincerely yours,

President

Enclosure

The Administration’s Green Book proposals include a provision that would amend current law to explicitly require that the fair market value used for estate and gift tax purposes also must be used to determine the income tax basis of property for income tax purposes.  The proposal notes that section 6034A also imposes a consistency requirement for beneficiaries of trusts and estates.  The beneficiary must report on his or her income tax return the exact information included on the Schedule K-1 for the trust and/or estate or provides appropriate notification of the inconsistent treatment.

The proposal includes both a consistency and a reporting requirement.  The consistency requirement means that the value used for estate and gift tax purposes must be used by the recipient to determine income tax basis unless the value used for gift or estate tax purposes exceeds the accurate value.  Under the proposed reporting requirement the executor of the decedent’s estate and the donor of a lifetime gift must provide the necessary valuation and basis information to both the recipient and the Internal Revenue Service.

Regulatory authority would be granted to provide details regarding the implementation and administration of these requirements, including rules for situations where a gift or estate tax return is not required to be filed, exceptions for situations where the surviving joint owner or other recipient may have better information about value than the executor, and the timing of required reporting where there has been an adjustment to value after the filing of an estate or gift tax return.  The scope of the exception where estate tax returns are not required to be filed should take into account the abbreviated reporting for estates that elect to use portability but are not otherwise required to file a federal estate tax return.

Case law already requires consistency when there is privity between the executor and the beneficiary of a decedent’s estate.  This rule curtails the use of calculated and deliberate attempts to reduce estate tax by claiming a low value for estate tax purposes and reduce income tax by claiming a higher value for basis purposes with respect to the same assets.  However, disparities of reporting are not necessarily deliberate.

Consider an estate that leaves asset A to individual A, asset B to individual B and the residue to C.  The estate tax is charged to the residue.  Assume that X is executor.  Individuals A, B and C are not entitled to intervene in the process of reporting values of estate assets.  Typically, the executor furnishes A, B and C the values reported for estate tax purposes for assets each receives and typically A, Band C use those values for reporting basis.  However, in fact the executor may not have valued the assets correctly and is not required by law to furnish reported values to A, Band C.  A, Band C under current law are required to use fair market value of assets received from a decedent for basis and are not required or even entitled to use values as reported for federal estate tax purposes unless those values are in fact fair market value.  Thus A, B and C may report a basis for an asset received from a decedent that differs from the value reported by the executor.  Requiring A and B to use the value obtained by X highlights the conflict of interests between A and B on the one hand and C on the other.  A and B benefit from a higher valuation and individual C benefits from a lower valuation.  If any of A, B or C is executor, then consistency already is required to that extent under current law.

The conflicts of interest are exacerbated if there is an audit.  In determining whether and how to settle an audit, X may compromise on some issues in order to avoid a dispute on other issues, looking to what is most beneficial to the estate as a whole, both in terms of tax costs, the costs of pursuing a controversy and delay in settling the estate.  The give and take of negotiating a settlement will be more difficult if X’s settlement negatively affects the determination of income tax basis of inherited assets for A but not B and C, or if C is unwilling to bear the costs of disputing valuation for either A or B where the proposed settlement doesn’t significantly affect the total estate tax liability or where a dispute would delay distribution to C.

In addition, where an audit resolves the estate tax due, current law does not make the audit report binding on the executor with respect to the details on which the estate computation of the government is based.  In other words, when an executor agrees to a deficiency, the executor is not agreeing to any specific changes in values proposed by the government in the audit report, absent a closing agreement, which is rarely obtained.  Instead, the executor is merely agreeing to pay the estate tax agreed on.  Introducing a requirement that the executor and the beneficiaries must use audit report valuations is likely to result in unduly complicating the settlement of audits and could result in state laws granting rights to beneficiaries to intervene in an audit.

If the audit is not resolved and the dispute is taken to court, none of A, B or C is entitled to participate in the litigation.  Under current law, only X is bound by the court’s determination, not A, B or C who were not parties.  Thus under current law, A, B and C may obtain their own valuations to determine basis of inherited assets.  Similarly, under current law, the government has always been free to disagree with the income tax basis reported by a beneficiary.

If the law were to change, as proposed by the Green Book, so that A, B and C would be bound by the valuations determined by X, or by the audit or the court proceeding, it would be reasonable to allow A, B and C to participate in some way in the preparation of the tax return, the audit, and the adjudication of estate tax liability.  The Green Book appears to contemplate this in its proposal of regulatory authority to address "situations in which the surviving joint tenant or other recipient may have better information than the executor." But that highlights the dilemma that it is possible that nothing short of the full legal rights of parties would appropriately protect those recipients, relief that would go well beyond normal bounds of rulemaking authority because it would implicate even the jurisdiction of Article III courts, and relief that in any event would threaten to make both administrative and judicial proceedings completely unmanageable.

And if the law is changed to require consistency to any extent, then state law may change to impose a fiduciary duty upon X not only to minimize estate taxes for the estate as a whole, but also to consider the effect of the valuation determinations on each beneficiary.  This would also be likely to greatly complicate the resolution of may audits and tax disputes.

Despite the existence of privity on the government’s side, the proposal does not require consistency for the government.  The government may argue on audit of an income tax return that the value used to determined gift and/or estate tax was too high.  There is no apparent reason for the government and taxpayers to be subject to different rules regarding consistency, and it should be taken into account that such perceived double standards can have a uniquely adverse effect on public acceptance of tax rules.

As possible alternatives, we recommend the following:

(i)    that executors and donors be required to file information returns advising donees and beneficiaries of the basis used for gift and estate tax reporting and of the amount of gift tax paid on lifetime gifts which enters into the determination of basis;

(ii)   that donees of lifetime gifts and beneficiaries of decedent’s estates be required to file information returns or schedules notifying the Internal Revenue Service when inconsistent valuations are used for income tax purposes, or in the case of an audit that is settled, notifying the Internal Revenue Service that an audit report increased the estate tax that was paid compared to the estate tax originally reported, and that the taxpayer is not using the values in the audit report for income tax purposes;

(iii)  that the privity rules be codified to make it clear when a beneficiary is bound by the values used to determine gift and estate tax liability; or

(iv)  that section 7491 be revised to always place the burden of proof on whichever party seeks to assert as basis a value different from the estate tax value.

The Green Book proposes to prevent contributions being made to the qualified retirement plan or IRA of an individual whose combined retirement plans exceed certain thresholds.

The proposed rule would require coordination of plans maintained by different sponsors as well as individual accounts and coordination of plans of different types, e.g.  defined contribution and defined benefit plans.  Each defined benefit plan administrator would need to separately calculate and disclose the present value of each participant.

If the proposal intends that an amount contributed by an employer to a defined benefit plan that results in an excess accrual are to be withdrawn from the plan, we have concerns about the resulting complexity of plan administration that might arise.  The Green Book provides:

If a taxpayer received a contribution or an accrual that would result in an accumulation in excess of the maximum permitted amount, the excess would be treated in a manner similar to the treatment of an excess deferral under current law.  Thus, the taxpayer would have to include the amount of the resulting excess accumulation in current income and would be allowed a grace period during which the taxpayer could withdraw the excess from the account or plan in order to comply with the limit.  If the taxpayer did not withdraw the excess contribution (or excess accrual), then the excess amounts and attributable earnings would be subject to income tax when distributed, without any adjustment for basis (and without regard to whether the distribution is made from a Roth IRA or a designated Roth account within a plan).

We recommend that the treatment of excess accumulations be clarified.

How would an employer be required to treat an employee who fails to report the employee’s balances in other plans under the proposal?

The proposal may interfere with policies favoring automatic enrollment and automatic employee contributions to 401(k) plans.  Those policies promote retirement savings as a result of employees’ inaction.

The compliance costs and reduced benefits to owner participants may discourage smaller employers from maintaining retirement plans.

The large majority of employees have savings well under the proposed cap.  Administrative costs might decrease and automatic enrollment facilitated if only highly­ compensated employees were subject to this proposal.

Many small employers maintain qualified retirement plans because the owners wish to have a tax-qualified vehicle to save for their own retirement.  Rank-and-file employees tend not to fully fund their retirement benefits, which has led federal agencies and Congress to adopt measures encouraging participation by rank-and-file employees.  If the owners become unable to benefit from maintaining qualified plans, they might very well discontinue making contributions to them, thereby diminishing the future retirement savings of rank-and-file employees.  Thus, this proposed legislation may have unintended consequences, effects that are the opposite of the policy that federal agencies and Congress have been promoting for many years.

Defined benefit plans impose investment performance risk on employers, whereas defined contribution plans impose investment performance risk on employees. If a defined benefit plan turns out to have a shortfall due to investment performance or changes in actuarial assumptions, the employer would contribute as much as necessary to shore up the plan.  However, annual limitations on contributions to defined contribution plans or IRAs might prevent restoring the expected benefit for defined contribution plans that suffer investment losses.

Perhaps the formula should consider how the annual cap on defined contribution plan contributions and the investment risks affect retirement savings.  Congress could consider either providing a higher cap or allowing make up contributions without regard to the annual cap on contributions if a defined contribution plan is underfunded in a subsequent year, for example, due to poor investment returns.

The Green Book proposes to require distributions from retirement plans within five years of the death of the participant in a retirement plan (an IRA or other qualified plan) except where the beneficiary is a spouse, a minor child, a disabled beneficiary or a beneficiary not more than 10 years younger than the decedent (a "qualified beneficiary").  Under current law, distributions from retirement plans may be made over the life expectancy of the designated beneficiary, which is referred to as "stretching." The legislative proposals do not address the treatment of trusts for the benefit of qualified beneficiaries.  The statute should address this issue because trusts are especially appropriate for minor and disabled beneficiaries and for spouse beneficiaries where the decedent has children from a prior marriage and because accelerated distributions to trusts exposes the withdrawals to tax at a higher rate than might otherwise be applicable.  Moreover, the current regulations dealing with required minimum distributions where a trust is the designated beneficiary of a retirement plan are both unclear and impracticable.

We suggest that the legislation provide that withdrawals be allowed to be made to a trust for the benefit of qualified beneficiaries over the life of the eldest qualified beneficiary when all beneficiaries to whom a distribution is required or permitted to be made during the calendar year ("current beneficiaries") are qualified beneficiaries, whether or not income can be accumulated in the trust for distribution to a non qualified beneficiary at a later time.  We suggest that the five year payout rule be imposed when any current beneficiary is not a qualified beneficiary unless the beneficial interest of the non qualified beneficiary was paid out of the trust within a reasonable period of time, either by payment to another trust for such non qualified beneficiary or outright to him or her. Such segregated amount would subject to accelerated timing for payout, e.g.  within a few months’ of the person becoming a qualified beneficiary or on the fifth anniversary of the decedent’s death, whichever is later.

As explained in the enclosure, existing regulations allow withdrawals over the life expectancy of the eldest current beneficiary of a trust only under limited circumstances. If income can be accumulated in the trust for distribution to a contingent beneficiary after the death of the current income beneficiary, the age and/or status of the contingent beneficiary has to be taken into account in determining the required distributions. This is true even if the likelihood of the contingent beneficiary taking any benefit is extremely remote.  This limitation is particularly problematic in the case of minors and disabled beneficiaries, but is also difficult in other contexts where distribution to a trust beneficiary would be unwise.  There are many legitimate non tax reasons for using a trust to limit a beneficiary’s access to funds, and we suggest that the tax laws support the use of such limitations where they are not imposed for tax avoidance purposes.

Under current regulations, the beneficiaries of a trust are determined on September 30 of the year following the calendar year in which the account owner died.  This allows the plan to make distributions to nonqualified beneficiaries of the retirement plan so that the assets remaining in the plan could take advantage of the stretching. Perhaps that rule also should apply to this legislation.

Consideration might be given to ameliorating the income tax burden on trusts that receive retirement plan distributions.

A trust that has undistributed taxable income in excess of $11,950 is subjected to federal income tax at the same rate as individuals who earn $400,000 if single or $450,000 if married.  Any income that is distributed to a beneficiary is taxed to that beneficiary at that beneficiary’s income tax rates rather than at the trust’s income tax rates. A trustee can reduce trust taxes by distributing part or all of the current year’s income to or for the benefit of beneficiary.  If the beneficiary is in a lower tax bracket, the combined tax cost is lower when distributions are made to shift income to a beneficiary. However, there are a number of non-tax reasons why distributions to a beneficiary may be inappropriate, especially for qualified beneficiaries - minors and disabled persons and spouses where there are children from another marriage.

Forcing retirement plan benefits to be distributed over a short amount of time will likely cause the trust to pay tax on the benefits at rates much higher than the trust beneficiary would pay.  Congress enacted the compressed rates due to concerns that trusts were being used by wealthy individuals to shift income within family units and lower the overall tax burden.  That concern could be addressed by other means than imposing tax on trusts at the highest rate under compressed brackets.

For example, Congress might consider allowing the accelerated retirement plan benefits payable to a trust to be taxed at the beneficiary’s income tax rates, rather than at the trust’s income tax rates.  This might be calculated in a manner similar to the "Kiddie tax," which determines tax by adding a child’s income to a parent’s income.  If the trust has more than one current beneficiary, the calculation might use the highest adjusted gross income of a beneficiary, perhaps determined in the same manner as Internal Revenue Code section 152(c)(4)(A)(ii), dealing with allocating the exemption for dependents.

The term HEET is generally used to describe a trust that avoids GST tax by (i) making distributions directly to health care providers and/or to educational institutions to pay medical expenses and tuition costs of trust beneficiaries that are exempt from tax under section 2611(b)(l); and (ii) having at all times at least one beneficiary who is a non-skip person so that the initial funding of the trust is not a direct skip and the termination of a beneficiary’s interest in the trust will never be a taxable termination.^[1] The non-skip person is often a charity.  However, section 2652(c)(2) allows the IRS to ignore interests of trust beneficiaries used primarily to postpone or avoid GST tax.  If the beneficial interest of the charity is ignored, then GST tax would not be avoided either on initial funding (if the trust had no other non-skip person as a beneficiary) or on the termination of the interest of a beneficiary if no other non-skip person had a "real" interest in the trust.  That is, BEETs would not "work" to avoid GST tax.  On the other hand, if the charity’s interest is not ignored, then the trust can benefit multiple generations of the grantor’s descendants without incurring GST tax.  Allowing a trust to pay tuition and medical expenses of multiple generations of descendants without payment of GST tax apparently is perceived to be abusive.

The HEET Proposal set forth in the "General Explanations of the Administration’s Fiscal Year 2014 Revenue Proposals" (the "Green Book") seeks to repeal section 2611(b)(1) on the grounds that the statute was never intended to apply to trust distributions but only to direct gifts by an individual to pay medical and educational expenses.  The Green Book states that "[T]he intent of section 2611(b)(1) is to exempt from GST tax only those payments that are not subject to gift tax, that is, payments made by a living donor directly to the provider of medical care for another person or directly to a school for another person’s tuition." This assertion cannot be correct because the statutory language is quite clearly applicable only to trust distributions.  Section 2611(b)(1) provides that "[T]he term ‘generation-skipping transfer’ does not include any transfer which, if made inter vivos by an individual, would not be treated as a taxable gift by reason of section 2503(e) (relating to exclusion of certain transfers for education or medical expenses)." [Emphasis added.] Thus, the statute is applicable not to transfers made inter-vivos by an individual but rather to transfers "which if made" by an individual would not be taxable gifts.  Moreover, section 2642(c) provides an exemption from GST tax for an individual’s inter­ vivos gifts for medical and tuition expenses.  If section 2611(b)(l) applied only to gifts made by living individuals, the statute would be unnecessary because section 2642(c) already exempts such gifts from GST tax.

Even if the proposal is to change rather than "clarify" legislative policy, it appears that the more appropriate focus should be on strengthening the rules to prevent avoidance of direct skips and taxable terminations such as by including non-skip beneficiaries for tax avoidance purposes.  There are several reasons why the proposal, as written, is misdirected:

Because the proposal, as written, is aimed at a certain kind of distribution, not at taxable terminations, it would still permit a trust to continue for several generations without a taxable termination, and therefore without any GST tax, merely by employing a charity as a permanent non-skip beneficiary.  Thus, the very technique that is thought to be what arguably makes HEETs inappropriate or even abusive would not be affected.

Because the proposal, as written, is aimed at a certain kind of distribution, not at the definition of an "interest in property held in trust," it would not affect the technique described in footnote 1.  If it is believed that the type of HEET described in footnote 1 is abusive, section 2652(c) could be revised not only to disregard interests included primarily to avoid or postpone the GST tax but also to disregard a delay in trust benefits provided primarily to avoid giving a person an interest and thereby to avoid or postpone the GST tax.

Despite the implication of the Green Book proposal, current section 261l(b)(l) clearly permits distributions to directly pay medical expenses and tuition without GST tax.  Therefore, such distributions may be the current practice or expectation of trustees and beneficiaries of trusts that are not designed as HEETs or otherwise designed or used to avoid GST tax on taxable terminations.  Because the Green Book proposal, as written, would adversely change that tax treatment whether or not the proposal is expanded to address real abuses, the enactment of that proposal could upset the reasonable plans and expectations of trustees and beneficiaries, as well as grantors, of such trusts, with no apparent policy justification.  While these may be reasons not to enact the proposal at all, if it is enacted we recommend consideration of an exemption for existing trusts, or at the very minimum relief from the abrupt proposed effective date for medical and tuition payments made before the proposal is enacted.

If, apart from possible abusive uses of trusts that could be addressed only by different remedies, it is still viewed as appropriate to repeal section 2611(b)(1), a trustee of a trust with broad discretion over distributions could get around that by making a distribution to a non skip person (such as a child of the grantor) who could in turn make the direct medical or tuition payment for a skip person (such as a grandchild of the grantor) free of gift tax under section 2503(e).  Indeed, this would also be true for future generations, because under section 2653 the grandchildren would in effect become non skip persons after the taxable termination caused (in a non abusive trust) by the death of a child or of the last surviving child, and so on for each generation.  In our experience, tracing rules or "step transaction" rules are clumsy and often ineffective ways of changing such a result.

While the workaround described in the previous paragraph would be available to trustees with broad discretion over distributions, it might not be available to a trustee with narrower discretion often found in older or more "traditional" trusts.  For example, if the trustee’s discretion to distribute principal to the non skip person were limited to that person’s need for support, maintenance, health, and education, it could strain or violate that trustee’s fiduciary duties to justify such a distribution with reference to the health or education of the skip person who is the intended ultimate beneficiary.  There are many reasons why trusts are designed as they are, and it is often an unfair trap to permit some trustees to protect a tax result that other trustees cannot protect.  It can be particularly unfair when the reward in such situations goes to the grantors and trustees who receive the most sophisticated tax advice.

Section 2652(c)(2) provides the IRS with the authority it needs to prevent multi generational HEETs.  The Administration does not need legislation to accomplish its policy objectives.

The General Explanations of the Administration’s Fiscal Year 2014 Revenue Proposals (the so-called "Green Book"), proposes amendments to address the "lack of coordination between the income and transfer tax rules" which "creates opportunities to structure transactions between the deemed owner and the trust that can result in the transfer of significant wealth by the deemed owner without transfer tax consequences."  Since the proposal is limited to grantor trusts,^[1] presumably the intent is to prevent value shifting without either income or transfer tax consequences.  However, it should be considered that the avoidance of income tax on the transfer of assets to a grantor trust is only temporary because the assets transferred to the grantor trust retain the same income tax basis as the grantor had so that the unrealized gain will eventually be recognized.

1.    Alternatives to avoid inappropriate value shifting without transfer and income tax consequences: (a) conform income tax rules to transfer tax rules; (b) conform transfer tax rules to income tax rules; (c) limit proposal to abusive leveraged transactions and define those qualified purchase money loans that are not abusive.

One way to address the "lack of coordination" would be to make the income and transfer tax rules the same.  For example, if a gift to a trust were complete for gift and estate tax purposes, the trust would not be a grantor trust for income tax purposes; or, alternatively, if a trust is a grantor trust for income tax purposes, then gifts to the trust would be incomplete for transfer tax purposes.  Such a solution would involve broad-reaching changes in the income and/or transfer tax rules that may have unintended consequences. A narrower rule is probably sufficient to address value shifting techniques that inappropriately avoid transfer tax payments.

Although the proposed rule change is appropriately limited to grantor trusts that engage in sales, exchanges, and comparable transactions with the grantor, it still would reach transactions that do not result in value shifting without payment of transfer tax.  We suggest that where appropriate transfer tax has been paid, the new rules should not apply because the value shift is not abusive.  For example, if a grantor gifts cash to a trust and pays gift tax on that gift, and subsequently sells an asset to the trust in a value-for-value exchange for the cash, although appreciation on the asset will shift, the shift of value is not inappropriate because gift tax has been paid on the cash gift.  On the other hand, if the donor gifts a small amount of cash to the trust ("seed money") and sells an asset to the trust in exchange for a note bearing interest at the applicable federal rate, value shifts to the trust with payment of gift tax only on the seed money.  This sort of leveraged transaction presents a potential for the abuse identified in the proposal - the shift of value without payment of transfer tax.  But the reach of an "all or nothing" approach is not likely to be understood, and therefore such an approach can risk the loss of legitimacy.  More appropriate would be a rule that limits value shifting to transactions between a grantor and a grantor trust that lack a prescribed level of economic substance, such as where the loan terms do not reasonably reflect the prevailing standards of lending.  While we recognize that a "bright line test" is appropriate as a rule of administrative convenience, the practices commonly used now arguably do not reflect market realities.

Limiting the new rules to highly leveraged transactions that depart significantly from third party lending practices is consistent with legislative solutions that previously have been applied to combat or restrict other transfer tax avoidance techniques. For example, section 2701 of the Code doesn’t ban all corporate freeze techniques but only those that have terms that allow inappropriate value shifting.  If the preferred interest is a "qualified interest" (cumulative and the preference is paid in accordance with its terms), the freeze will be respected. Similarly, section 2702 doesn’t give value to a grantor retained income trust if the remainder beneficiaries are related and the trust doesn’t require a fixed annuity payment.  The same sort of approach would seem to be appropriate to stop the use of highly leveraged sales to grantor trusts from allowing value shifts without payment of transfer taxes. That would involve legislating a bright line test that more closely reflects economic realities

The proposal seems not to distinguish between value shifts that avoid transfer tax and those that do not and thus appears to be overly broad.  The proposal reads in part:

If a person who is a deemed owner under the grantor trust rules of all or a portion of a trust engages in a transaction with that trust that constitutes a sale, exchange, or comparable transaction that is disregarded for income tax purposes by reason of the person’s treatment as a deemed owner of the trust, then the portion of the trust attributable to the property received by the trust in that transaction (including all retained income therefrom, appreciation thereon, and reinvestments thereof, net of the amount of the consideration received by the person in that transaction) will be subject to estate tax as part of the gross estate of the deemed owner, will be subject to gift tax at any time during the deemed owner’s life when his or her treatment as a deemed owner of the trust is terminated, and will be treated as a gift by the deemed owner to the extent any distribution is made to another person (except in discharge of the deemed owner’s obligation to the distributee) during the life of the deemed owner.  The proposal would reduce the amount subject to transfer tax by any portion of that amount that was treated as a prior taxable gift by the deemed owner.  The transfer tax imposed by this proposal would be payable from the trust.

As discussed below, the proposal also seems not to cover some arrangements that arguably are abusive.

If the perceived abuse is a sale, exchange, or comparable transaction that allows value to shift from the donor to the trust without payment of transfer taxes, then the new rules should not apply to sales from the trust to the grantor for a note.  This sort of "one-way" rule was used in section 2701 as well.  (The creation of a preferred interest in the younger generation members of the family and the creation of a common interest in the senior generation is not subject to the anti-abuse rules of section 2701 because it does not allow shifting of value to younger generation family members.)  There could be many legitimate and substantial non-tax reasons for the grantor to repurchase an asset from a grantor trust.  For example, the grantor may want to regain control of a family business owned by a grantor trust if the trust beneficiaries ceased to participate in the business.

The new rules probably should not apply to leases or loans between a grantor and a grantor trust (except where the loans are used to purchase assets from the grantor as a way of avoiding the new rules if applicable only to sales for a note).  This exception for leases and loans parallels exceptions to section 2701, which only applies the anti-abuse rules to "distribution rights" and does not apply them to leases and loans.  Leases and loans are not specifically mentioned in the proposal, but it should be made clear that they are not "comparable transactions," except perhaps in special cases where Congress determines that the policy behind the proposal should apply.

It is puzzling in a number of respects that the proposal would exempt trusts that are grantor trusts solely by reason of section 677(a)(3), which makes a trust a grantor trust if its income may be used by the grantor or a nonadverse party to pay premiums on policies of insurance on the life of the grantor or the grantor’s spouse.

For example, the proposal would seem to exempt from the proposed rule a grantor’s sale of an asset to- a qualifying life insurance trust.  It is not intuitive why such an exemption furthers the policy underlying the legislative proposal.  Moreover, it is unclear whether the exemption is intended to be limited to a sale of a life insurance policy to a qualifying life insurance trust or whether the exemption also is meant to cover sales of other assets.

A life insurance trust may be a grantor trust both by reason of section 677(a)(3) and another section of subpart E of part 1 of Subchapter J as well.  The proposal seems to limit the exemption to trusts that are grantor trusts solely by reason of section 677(a)(3).  There seems to be no policy reason to exempt a life insurance trust that is a grantor trust only because it allows income to be used to pay life insurance premiums and not for any other reason.  For example, a trust that benefits the grantor’s spouse can be a grantor trust under section 677(a)(1) or (2).  Why should a life insurance trust that doesn’t benefit the grantor’s spouse qualify for an exemption and an otherwise similar life insurance trust that does benefit the grantor’s spouse not qualify for the exemption?

In addition, section 677(a)(3) by its terms applies to a trust whose income "may" be used to pay life insurance premiums on a policy on the life of the grantor or the grantor’s spouse.  It is not clear whether a trust that may use income but does not in fact use income, or all of its income, to pay such premiums is a grantor trust, either in whole or in part.  If the exemption for life insurance trusts is broadly applied, the exception could provide a loophole for the new rule.  That is, merely by authorizing the trustee to use income to pay life insurance premiums and including no other grantor trust feature, any sale of assets by the grantor to the trust would be exempt from the proposed rule.  We are certain that this wasn’t the intent, of course.

The proposal expressly provides that it would not change the treatment of any trust that is already includable in the grantor’s gross estate under existing provisions, including but not limited to, grantor retained income or annuity trusts, personal residence trusts, and qualified personal residence trusts.  The implication - although this is only our guess - is that the treatment of such trusts does not have to be changed because such trusts already are treated consistently with the proposal.  In fact, the value of grantor retained income or annuity trusts, personal residence trusts, and qualified personal residence trusts is includable in the grantor’s gross estate only if the grantor fails to survive the term of her retained interest, which is quite different from the result under the proposal.

In addition, it is not clear whether the proposed rules would be applicable to such a trust after the end of the retained term, when the value of the assets is no longer includable in the gross estate of the grantor, if during the retained term the trust had engaged in a sale, exchange, or comparable transaction with the grantor.  For example, suppose the grantor created a 10-year GRAT.  Three years into the term, the grantor exchanges the index fund owned by the trust for bonds of equivalent value.  Assume that the grantor is living at the end of the term of the GRAT and the assets pass outright to the grantor’s children as remainder beneficiaries.  It is not clear whether the proposed rule would apply to impose gift tax at the end of the term and if so, how the proposed rule would calculate the amount of the taxable gift.  Presumably the gift would be limited to that portion of the yield on the bonds represented by the remainder of the trust, assuming that could be determined.

Alternatively, if a grantor makes a home improvement loan to a qualified personal residence trust to pay for improvements to the residence, when the term of the trust ends and the residence is distributable to the remainder beneficiary subject to the home improvement loan, it is not clear whether the proposed rule would apply to impose gift tax on the value represented by the home improvement.

The proposed rule should clarify how it would apply to transactions between a grantor and a trust that was treated only in part as a grantor trust, such as a trust that is a grantor trust only as to income, or only as to principal.  Would the rule be limited only to transactions with a wholly owned grantor trust? Would the fact that a portion of the trust was a non-grantor trust escape the new rules in full or only in part? How would that part be determined?

The gift tax imposed by the proposed rule may be harsher than regular gift tax rules.

For example, it should be clarified whether all usual exclusions and exemptions that would be available for actual gifts will apply equally to deemed gifts made under the proposed rule. This would include the annual exclusion and the marital and charitable deductions, for example.

In addition, the gift tax imposed under the proposed rule appears in general to be imposed at a higher effective rate than the regular gift tax because, under the proposal, the tax on deemed gifts is payable by the trust, rather than by the donor as in the case of actual lifetime gifts.  The gift tax normally is tax exclusive, except for gifts made within three years before death, which become tax inclusive by reason of section 2035.  But the gift tax under the proposed rule appears to be tax inclusive in all cases.

The gift tax reporting will require coordination between the grantor and the trustee of the grantor trust.  The amount of gift tax due will depend upon prior gifts by the same donor.  Normally the donor files the gift tax return.  However, the donor/grantor may not know that a taxable event (such as a distribution to a beneficiary) has occurred to create a reporting obligation.  In addition, it is not evident that the grantor can force the trustee to pay the gift tax imposed on the trust.

Any drafting of this sort of legislation would be an occasion to consider what adjustments to basis would be appropriate and consistent, undersection 1015(d)(6) with respect to gift tax and under section 1014(b)(9)with respect to estate tax.  The proposal does not mention basis.

If a grantor’s spouse sells assets to a grantor trust created by his spouse, the sale avoids gain under section 1041 because the sale is treated as made between spouses.  The proposal does not appear to address that sort of value shifting technique, although it could be amended to reach that case.

The proposal also may not appropriately capture techniques where grantor trust status terminates before assets acquired by purchase from the grantor have appreciated in value.  Due to the offset for the consideration furnished by the trust, the amount of the gift at that point in time would be quite low.

For example, suppose that a grantor trust purchases an asset from the grantor for its fair market value of $1MM and immediately ceases to be a grantor trust.  There would be no gift tax due because there had been no appreciation and no value shifting. If the purchase were not leveraged with debt, that is not a loophole because there has been no shift in value without payment of transfer tax.  If the purchase was leveraged, on the other hand, there likely would be gain realized when the trust ceased to qualify as a grantor trust.  If income tax is paid as a result of the termination of grantor trust status on a leveraged transaction, however, presumably no gift tax should be due because the trust would not have been used to avoid both income tax and transfer tax on the gain. In other words, the tax consequences should be the same whether the sale for a note is made to a grantor trust which immediately after the sale ceases to be a grantor trust or the sale for a note is made to a non-grantor trust.

A trust that is includable in the grantor’s estate under sections 2036-2038is exempted from the proposed rule even where the gift is complete for gift tax purposes. For example, if a trust gives the trustee discretion to distribute to beneficiaries but the donor has a limited testamentary power of appointment (whether retained or later conferred), the gift may be complete but the assets would be includable in the grantor’s gross estate.  If such a trust is exempted from the new rule, however, distributions of the profit from leveraged sales could be made to beneficiaries (including other nongrantor trusts) free of gift tax so that the amount includable in the grantor’s estate would be limited to the trust’s remaining undistributed assets.  This would continue to allow value shifting without payment of transfer taxes.

The proposal seems to create a very broad rule penalizing sales to grantor trusts and then, as stated in the last sentence of the proposal, relies on regulations to implement appropriate exceptions. However, this would have a chilling effect even on tax planning contemplated by the exceptions, until those exceptions are promulgated in regulations, which seems to be an appropriate approach only for very abusive transactions. Pending the issuance of regulations creating exceptions, it is hard to see the policy justification for giving sales to grantor trusts harsher gift tax treatment than gifts themselves.

In sum, the most efficient way to combat the perceived abuse may be to define those purchase money loans made by senior generation grantors to grantor trusts benefitting younger generation members that will be respected as loans and to treat other loans as having no value for gift tax purposes.  Qualified purchase money loans could require compliance with bright line standards comparable to third party purchase money loans.  This would be more consistent with prior legislative fixes to the use of retained interests to reduce transfer taxes, particularly in sections 2701 and 2702.

In 2013, the threshold for the 3.8% Net Investment Income Tax ("NII Tax") for individuals is set at $250,000 of Net Investment Income ("NII") for married taxpayers filing jointly and $200,000 of NII for individual taxpayers.  We believe that these high thresholds reflect Congressional intent that the NII Tax would be applied only to the NII of high-income taxpayers.^[1] However, as a result of the system of taxation to which trusts and estates are subject, the NII Tax will apply to trusts and estates NII at levels much lower than the NII levels at which high-income individual taxpayers would be subject to NII Tax.  This rule affects middle income taxpayers in a manner that may have been unintended.

The 2013 threshold for the NII Tax for purposes of a trust or an estate is set at $11,950 of NII.^[2] A trust or estate is not taxable to the extent amounts are distributed to the trust’s or estate’s beneficiaries.  Instead, the taxable income shifts to the beneficiaries.  Thus, a fiduciary who has lower income beneficiaries could avoid the NII Tax by making distributions to the beneficiaries.  However, there are many considerations other than taxes that a fiduciary must consider when determining how much to distribute to beneficiaries.  If the trust or estate does not make distributions to beneficiaries, the NII Tax will apply to deplete the trust or estate after realizing only $11,950 of income.  For trusts and estates benefitting lower income beneficiaries, this tax system creates a strong tax incentive to distribute to beneficiaries even where other considerations suggest that this is not wise.  By contrast, the trust’s or estate’s lower income threshold for applying the NII Tax will not affect high income beneficiaries who would incur the NII Tax even if distributions were made.^[3] The lower threshold will frustrate the non-tax objectives of a middle income taxpayer who funds a trust to provide for his or her family where none of the beneficiaries’ incomes exceed the thresholds for the NII Tax to apply to individuals.  Because middle-income taxpayers have limited resources, they are more likely to set up one trust for the benefit of their entire family, rather than separate trusts for each child, thereby magnifying the cost of the NII Tax, in light of the threshold of $11,950 in 2013 to be used against all of the trust’s or estate’s NII for purposes of the NII Tax.  In an estate with NII of over $11,950 in 2013, the NII tax is unavoidable, at least in the early years of the estate, because estate assets may have to remain in the estate, and subject to tax at the estate level, during the administration period.

Moreover, in a trust, if the trustee makes distributions to reduce or eliminate the NII Tax, the distributions would need to go out of trust, perhaps to custodianships for minor children that would become their property at age 21.  These distributions will deplete the capital available to the trust beneficiaries later when they start college, graduate or professional school.

If Congress decided that as a matter of policy it was unfair to expose estates and trusts to the NII tax at such a low threshold set for trusts and estates because it unfairly burdens middle income taxpayers, one possible statutory solution is to raise the threshold for purposes of the NII Tax for a trust and estate to the same threshold for unmarried individuals.

This solution may be inconsistent with the policy behind taxing trusts and estates at the highest bracket after realizing only $11,950 of income: to discourage the use of multiple trusts for income shifting in order to allow more income to be taxed at lower brackets.  Therefore, if this is a policy obstacle to this simpler solution, the alternative solution, discussed below, may be a more appealing alternative.

An alternative statutory solution is a joint election by the executor or trustee and the beneficiaries of the estate or trust to treat all NII of the trust, solely for purposes of calculating the amount of the NII Tax, as if it had been distributed to its beneficiaries during a calendar year without the necessity of making any distributions.  This deemed distribution would only apply for purposes of calculating the NII Tax payable by the trust or estate.  The trustee or executor would determine the NII Tax payable after taking into account each such beneficiary’s other NII as well as such beneficiary’s threshold for purposes of the NII Tax.

There may be the concern that an estate, or more likely, a trust, might include "straw" beneficiaries whose lower NII might be used to artificially increase the amount of individual thresholds at which the NII Tax does not apply.  To avoid the use of straw beneficiaries, the election could be limited to avoid manipulation.  For example, the election could be limited to beneficiaries who are spouses and descendants of the grantor or decedent, and the amount deemed distributed to each could be fixed by formula in order to further limit manipulation.  A possible formula would attribute the deemed distribution first to the surviving spouse, if applicable, and then to descendants of the grantor or decedent on a per stirpesbasis.^[4] Alternatively, the election could be limited to only one beneficiary of the trust or estate, which could be the beneficiary who is in the highest income tax bracket.

Addressing these limitations, as well as determining the beneficiary’s other NII, may become administratively complex.  The trustee or executor would probably have to rely on the certification of the electing beneficiaries as to the amount and character of their incomes.  However, because the tax would remain the obligation of the trustee or executor, and only the computation of the tax would be affected, the administrative complexity might be manageable.

Alternatively, the trustee or executor could allocate the NII to the beneficiaries without distributing the net investment income to them.  Then each beneficiary would include such allocated NII in his or her own NII, and apply the beneficiary’s threshold for purposes of determining if the beneficiary has a NII Tax obligation.  The beneficiary would be paying the 3.8% NII tax on NII he or she did not receive (the deemed distributed NII from the trust or estate) but this would occur only if the beneficiary was a high income taxpayer with NII that exceeded the applicable individual thresholds.

[1]  Linked below are documents previously submitted by ACTEC addressing the problems under the regulations that arise when a trust is the designated beneficiary of a retirement plan.

[1]   A less common type of HEET is one that avoids a direct skip on funding, not by giving a non-skip person (such as a child of the grantor) an interest, but by providing that immediately after funding and for a period of time thereafter no person has an "interest in property held in trust" (as defined for GST tax purposes in section 2652(c)(l), which includes only present interests and excludes future interests) and also providing that distributions may or must be made in the future to non-skip persons.  An example would be a trust that directed that no distributions be made to anyone for five years and then authorized distributions to the grantor’s grandchildren for health and education, with the remainder at the time of the death of the last surviving beneficiary going to charity.  This trust would not be a "skip person trust," and thus there would not be a direct skip on funding.  Although the grantor also could avoid a direct skip simply by including a child as a discretionary beneficiary, this type of trust would avoid a potential taxable termination until the death of a grandchild (rather than the death of a child).  In this type Cheeto, GST tax will be payable when any person with an interest ceases to have an interest (e.g. dies) unless a non-skip person has an interest at that time, but the death of a grandchild is likely to occur later than the death of a child.  Under section 2652(c)(2) the interest of a non-skip person used to avoid GST tax when a beneficiary’s interest terminates can be disregarded.  However, section 2652(c)(2), as currently in effect, will not prevent the avoidance of the direct skip on funding because there is no interest to be disregarded.  Rather the future interests of the grandchildren would have to be treated as present interests to prevent avoidance of a direct skip on funding.

[1]  The proposal is described in terms of trusts with a "deemed owner" for income tax purposes.  For simplicity, this paper uses the terms "grantor" and "grantor trust," which should be read to include all deemed owners and all deemed owned trusts, as appropriate.

[1]  This paper is explicitly based on that belief.  If that belief is mistaken, some of the observations in this paper may not be fully applicable.

[2]  $11,950 is also the threshold amount of taxable income in 2013 at which trusts and estates are subject to the highest rate of tax.

[3]  If the beneficiaries of the trust or estate have NII that exceeds their threshold for the NII Tax, then the tax is going to be due, whether the trust pays it or distributes the NII and the beneficiaries pay the NII tax.  For middle income trusts and estates, however, the issue is that the tax can be avoided entirely by distributing the NII to the beneficiaries of the middle income trust or estate whose NII is below the applicable individual thresholds for the NII Tax.

[4]  There is not likely to be a need to attribute the income to a spouse of a living grantor because such a trust would be a "grantor trust" the income of which is taxable to the grantor (rather than to the trust) in almost all cases.