The BDIT and BDOT—Promising Structures, Unsettled Authority, and Practical Risk
In the first post in this series, we walked through the underlying income tax problem that trusts create and the three ways the tax law allows you to deal with it. That framework is helpful, but it’s only the starting point. Once you understand how Section 678 shifts the income tax burden from a trust to a beneficiary, the natural next question is how far that idea can be pushed in practice. Over the last several years, practitioners have developed more aggressive structures built on that same foundation, most notably the Beneficiary Defective Inheritor’s Trust (“BDIT”) and the Beneficiary Deemed Owner Trust (“BDOT”). Both are designed to capture the benefits of grantor trust treatment without the traditional limitations. And both raise legal and economic questions that deserve a more careful look than they often receive.
The BDIT: A Promising Structure With a Shaky Foundation
The BDIT starts the same way as the basic 678 Trust. A third party (a parent, sibling, or close friend, but not a spouse) creates an irrevocable trust for the benefit of the client and seeds it with an unreimbursed gift of $5,000. The client receives a short-term withdrawal right over that $5,000 contribution. When the withdrawal right lapses, the income tax law should continue to treat the client as the trust’s deemed owner under Section 678. The estate tax law does not.
So far, so good. But a trust with $5,000 in it is not particularly useful. The BDIT’s appeal lies in what comes next: the beneficiary sells assets to the trust in exchange for a promissory note, just as a grantor would in a traditional sale to an IDGT. The sale is ignored for income tax purposes (because the beneficiary and the trust are the same person for income tax purposes). For estate tax purposes, the sale should be respected because the beneficiary receives a note, the trust receives assets, and all future appreciation on those assets occurs outside the taxable estate. The beneficiary can serve as trustee, receive distributions for health, education, maintenance, and support, and exercise a special power of appointment over the trust assets.
The pitch is compelling: you get the benefits of an IDGT sale: frozen estate value, tax-free appreciation, income tax efficiency. All of this without consuming gift tax exemption on a seed gift beyond what the 5 and 5 exception covers. The client is both seller and beneficiary. The assets come back within reach.
The problem is that the BDIT’s continued grantor trust status after the withdrawal right lapses depends on a legal position that has never been authoritatively resolved.
The Lapse Problem
Section 678(a)(2) provides that a person continues to be treated as the deemed owner of a trust after they “partially release or otherwise modify” a general power of appointment, so long as they retain control that would make a grantor the owner under the regular grantor trust rules. The IRS has consistently taken the position, in private letter rulings, that allowing a withdrawal right to lapse is equivalent to a partial release or modification for this purpose.
Here is the difficulty: private letter rulings are binding only on the taxpayer who requested them. They are not precedent. There is no statute, no Treasury regulation, and no published revenue ruling that directly answers whether a lapse triggers Section 678(a)(2). The logic supporting the IRS’s position is reasonable: treating similarly situated taxpayers differently based solely on whether they formally released their withdrawal right versus simply let it expire seems arbitrary. But “reasonable” is not the same as “settled,” and planning built on unsettled authority carries legal risk that clients need to understand.
A further complication: if the BDIT fails to maintain grantor trust status after the lapse, the beneficiary would be liable for capital gains on the sale of assets to the trust and would owe tax on interest income from the promissory note. The transaction that was supposed to be invisible for income tax purposes becomes a taxable event. This is not a minor technical glitch; it is a structural failure.
The Economic Substance Problem
Even setting aside the lapse question, the BDIT raises a more fundamental concern: can a trust with $5,000 in equity legitimately purchase $500,000 — or $5 million — in assets?
A trust so leveraged that its equity represents less than 1% of its total assets is not a creditworthy buyer. If the sale is not respected as a bona fide arm’s length transaction, the IRS has powerful arguments available: the transfer could be recharacterized as a gift, with the promissory note treated as a retained interest that falls within the reach of Section 2036. Under Texas law, an overleveraged transfer to a self-benefiting trust risks being treated as a self-settled trust, with creditor protection stripped away. The statute of limitations on gift tax does not bar the IRS from asserting estate inclusion under Sections 2036 and 2038 after the beneficiary’s death, potentially decades after the transaction was structured. The executor of the estate may have to prove the bona fides of a transaction completed many years earlier, from documentation that may or may not have survived.
The standard mitigation is to have another trust guarantee a portion of the promissory note, providing the economic substance the sale needs. We’ve seen this work when the guarantor holds meaningful assets pledged as collateral.[1] But the guaranty fee paid to the guarantor must reflect what an unrelated third party would actually charge. This is not an academic requirement. A below-market guaranty fee is a gift from the BDIT to the guarantor. Industry convention has settled around 3% of the amount guaranteed as a rule of thumb, but that figure needs to be substantiated by a qualified appraisal, not assumed. Experienced practitioners who have pressed the question have found that when appraisers actually do the work, the supportable fee is often slightly below 3%, and in some cases materially so, depending on the nature of the assets pledged.
There is also the question of whether a willing guarantor can be found at all. As one commentary pointedly noted in the context of the Lindquist analysis, there is a reason Donald Trump reportedly could not find anyone willing to take illiquid real estate as security for a large appellate bond. The market for guarantors of illiquid, thinly capitalized trusts is thinner than the promotional literature suggests.
The BDOT: An Elegant Theory With a Fundamental Flaw
The Beneficiary Deemed Owner Trust (“BDOT”), developed and advocated by certain credentialed practitioners, pushes Section 678 in a different direction. Rather than relying on a $5,000 seed gift and the lapse of a withdrawal right, the BDOT argues that a beneficiary who holds only the right to withdraw the trust’s taxable income, not its corpus, should be treated as the deemed owner of the entire trust.
The statutory hook is Section 678(a)(1), which provides that a person is treated as the owner of any portion of a trust with respect to which they hold “a power exercisable solely by himself to vest the corpus or the income therefrom in himself.” The argument is that the disjunctive “or” means that a withdrawal power over income alone is sufficient to make the powerholder the deemed owner of the entire trust (including corpus the powerholder has no right to withdraw).
If this reading were correct, the BDOT would be a significant advance over the BDIT. It would avoid the lapse problem entirely, because Section 678(a)(1) doesn’t require a prior general power followed by a partial release. It would eliminate the capitalization problem, because the trust would not need a $5,000 seed gift or a leveraged sale to function. And it could, in theory, convert virtually any trust (including marital trusts) into a 678 trust simply by giving the income beneficiary the right to withdraw taxable income.
The problem is that this reading of Section 678(a)(1) is difficult to reconcile with how the grantor trust rules have been understood for seventy years.
The word “portion” in Section 678(a)(1) is doing important work. Under the grantor trust rules, a person is treated as the owner of the portion of a trust over which they hold the requisite power (not the entirety of the trust unless their power extends to the entirety). When the Clifford trust rules were codified in the 1950s, it was well established that a grantor who retained only the income interest in a trust was the deemed owner of the income slice of the trust, while the corpus belonged to someone else. There was no suggestion that holding the income interest made you the owner of the corpus as well.
Apply that logic to the BDOT. If the beneficiary can only withdraw taxable income (again, not corpus) then the beneficiary should be treated as the owner of the income portion of the trust, not the whole thing. The proponent acknowledges this creates anomalies: what happens if the trust has a net capital loss? It is not coherent to “withdraw” a loss. The proposed solution of the taxable income beneficiary probably gets the deduction requires additional legal steps beyond what the statute supports, and the proponent ultimately suggests practitioners simply “avoid” the problem. When a structural analysis generates a result that can’t be resolved, the better response is to reconsider the analysis.
There is currently no statute, no regulation, no published ruling, and no court decision that endorses the BDOT’s reading of Section 678(a)(1). The proponent’s argument is creative and carefully constructed, but it asks practitioners to take two steps beyond existing authority simultaneously: first, that a power over income alone is sufficient to make someone the deemed owner of the entire trust; and second, that this deemed ownership means the trust is entirely ignored for income tax purposes as to that person, even with respect to corpus the person has no right to reach. That is a significant ask.
The Bottom Line
The BDIT and the BDOT represent the ambitious edge of 678 Trust planning. The BDIT is a more established structure with a clearer path to execution, but it rests on unresolved legal authority regarding the lapse of a withdrawal right and requires rigorous economic substance: real capitalization, documented guaranty fees, and impeccable records that will hold up under IRS scrutiny years down the road. The BDOT is intellectually interesting and would be a genuine advance in the field if the statutory argument holds up, but it currently lacks any authoritative support, and the anomalies generated by the proponent’s own analysis suggest the reading of Section 678(a)(1) may be incorrect.
None of this means these structures should be dismissed. It means they should be approached with the same candor and care that any sophisticated planning tool deserves: with a clear-eyed assessment of what the law supports, what it doesn’t, and what the consequences of being wrong would be.
If you are thinking about whether either of these structures belongs in your estate plan, that is exactly the kind of conversation worth having with experienced tax and estate planning counsel.
And finally, while this is one potential solution, the biggest draw of a BDIT is control remaining with the Trustee, who is also a beneficiary, who is probably the client. It is attractive, but there are certainly other ways to obtain it. See, for example Private Client Case Study: Building A Legacy.
Read Part 1 of this series →
This post is for informational purposes only and does not constitute or contain tax or legal advice.
[1] Typically, 10 to 20 percent of the note amount
