Real, But Not Unlimited.

The benefits are real. They are also, in every direction that matters to planning, bounded. Understanding precisely where the boundaries sit is the difference between a structure that survives a client’s worst year and one that funds the estate.

The starting point is the widest possible net. On filing, the estate takes “all legal or equitable interests of the debtor in property as of the commencement of the case.” 11 U.S.C. § 541(a)(1). A beneficial interest in a trust is such an interest, and it comes in.

The escape hatch is narrow and textual. Section 541(c)(2) provides that “a restriction on the transfer of a beneficial interest of the debtor in a trust that is enforceable under applicable nonbankruptcy law is enforceable in a case under this title.” 11 U.S.C. § 541(c)(2). In other words, federal bankruptcy law borrows state (or other applicable) law: if the spendthrift restriction would be honoured outside bankruptcy, it is honoured inside it, and the interest never becomes estate property at all.

The Supreme Court settled the breadth of that borrowing in Patterson v. Shumate, 504 U.S. 753 (1992), holding that “applicable nonbankruptcy law” is not limited to state spendthrift law and includes federal law — there, the anti-alienation provision required of ERISA-qualified plans, which accordingly excluded the debtor’s pension interest from the estate. That decision is why a qualified retirement plan is, for most debtors, the single most reliable protected asset in the American system.

Every word of section 541(c)(2) does work, and each word is a place where structures fail:

• “A restriction on transfer.” The instrument must actually contain one.
• “Of the debtor.” The debtor must be a beneficiary, not the settlor of a self-settled arrangement dressed as a beneficiary interest.
• “Enforceable under applicable nonbankruptcy law.” The governing law must in fact enforce it — against these creditors, on these facts.

The oldest and most consistently applied limit is that a settlor cannot put his own property beyond the reach of his own creditors while continuing to benefit from it. California’s formulation is representative: a trust for the settlor’s own benefit does not carry an enforceable restraint against the settlor’s creditors. Cal. Prob. Code § 15304 (settlor may not create an enforceable restraint against the settlor’s own creditors); see also Cal. Prob. Code § 15306.5 (permitting a creditor to reach a portion of a beneficiary’s interest, subject to support limits). Most states say the same thing in some form, and where they do, section 541(c)(2) has nothing to bite on — the restriction is not “enforceable under applicable nonbankruptcy law,” so the interest is estate property.

A number of American jurisdictions have of course enacted domestic asset protection trust statutes that reverse this rule as a matter of local law. Whether a bankruptcy court sitting elsewhere will apply that local law to a debtor with limited connection to the enacting state is a live and heavily litigated question; In re Portnoy is the classic illustration of a court declining to apply a chosen foreign law to defeat the forum’s own public policy. In re Portnoy, 201 B.R. 685 (Bankr. S.D.N.Y. 1996); see also In re Lawrence, 227 B.R. 907 (Bankr. S.D. Fla. 1998). Any client relying on a self-settled structure should understand that the applicable-law question, not the statute’s text, is usually the battlefield.

Congress addressed self-settled structures directly in 2005. Section 548(e) of the Bankruptcy Code allows a trustee to avoid a transfer made within ten years before the petition, where the transfer was made to a self-settled trust or similar device, the debtor is a beneficiary, and the transfer was made with actual intent to hinder, delay, or defraud present or future creditors. 11 U.S.C. § 548(e), added by the Bankruptcy Abuse Prevention and Consumer Protection Act of 2005, Pub. L. No. 109-8.

Two features of that provision are frequently misread. It is a ten-year window, not a ten-year prohibition: a transfer inside the window is avoidable only if the intent element is proved. And the intent element covers future creditors, which is why the badges-of-fraud analysis — insider transfers, transfers after a threat of suit, retained control or enjoyment, concealment, transfer of substantially all assets — does so much work in these cases.

The practical consequence for planning is simple and unglamorous. A structure funded while the settlor was solvent, unthreatened, and had no reasonably foreseeable claim on the horizon has the strongest possible answer to section 548(e). A structure funded as trouble approached has a decade of exposure and a fact pattern that reads like the statute’s legislative history.

Even a textbook third-party spendthrift trust does not create a black hole. In Frealy v. Reynolds, the Ninth Circuit, applying the California Supreme Court’s answer in Carmack v. Reynolds, held that a bankruptcy estate is entitled to the full amount of spendthrift trust distributions due to be paid as of the petition date, subject to a carve-out for amounts the beneficiary needs for support or education where the instrument specifies that purpose; and that a creditor may reach a portion of expected future payments under the state statutory formula, again reduced by the beneficiary’s support needs. Frealy v. Reynolds (In re Reynolds), 867 F.3d 1119 (9th Cir. 2017), applying Carmack v. Reynolds, 2 Cal. 5th 844, 391 P.3d 625 (2017); see also Frealy v. Reynolds, 779 F.3d 1028 (9th Cir. 2015) (certifying the question).

The timing point is the one clients never anticipate. A beneficiary who files a petition a month before a scheduled distribution may find that the distribution belongs to the estate. Protection attaches to the interest that remains subject to the trustee’s discretion; it does not follow money the trust is already obliged to pay out.

This is the mechanical reason that discretionary language matters as much as spendthrift language. A mandatory income interest is an asset with a due date. A purely discretionary interest, in a trust administered by an independent fiduciary who may decide not to distribute at all, is much closer to an expectancy — and much harder for any creditor, in or out of bankruptcy, to value or seize.

If the beneficiary can compel distribution, the restraint is illusory. Where a beneficiary holds absolute and sole discretion to demand the trust property, the spendthrift provision fails. Kupetz, Spendthrift Trusts (“if a beneficiary under a spendthrift trust has absolute and sole discretion to compel distribution of the trust assets, the spendthrift provision must fail”). The same logic runs through the offshore cases: in Webb v Webb, the Privy Council held that a settlor whose retained powers allowed him at any time to secure the entire trust fund to himself held “a bundle of rights … indistinguishable from ownership,” so that no effective disposition had occurred at all. Webb v Webb (Cook Islands) [2020] UKPC 22 (3 August 2020), at [89].

And where control is retained but denied, bankruptcy courts have a remedy that does not require any trust analysis: civil contempt. In re Lawrence remains the standing example — a settlor who transferred roughly $7 million offshore shortly before a large arbitration award, insisted he was powerless to bring it back, and was held in contempt and jailed for years while the courts declined to credit an impossibility he had created himself. In re Lawrence, 238 B.R. 498 (Bankr. S.D. Fla. 1999), aff’d, 251 B.R. 630 (S.D. Fla. 2000), aff’d, 279 F.3d 1294 (11th Cir. 2002).

The ABI’s title is the right conclusion, and the qualifiers are where the work is:

• Third-party settled beats self-settled. A trust settled by a parent or grandparent for a beneficiary, with a proper spendthrift clause, is the strongest form of the protection in American law, and section 541(c)(2) keeps it out of the estate.
• Discretionary beats mandatory. What is due is reachable. What is discretionary is not.
• Independent administration beats family administration. A beneficiary who can compel a distribution has no restraint at all, and a trustee who has never refused anything invites the same argument.
• Seasoning beats everything. The ten-year reach-back of section 548(e) and the intent-based state statutes both measure from the transfer. Time is the one input that cannot be manufactured after the fact.
• Applicable law is a design decision, not a footnote. Section 541(c)(2) is a borrowing statute. Which law it borrows, and whether a court will honour that choice, is where self-settled structures are won or lost.

Bankruptcy strips a structure to its architecture. Anything ornamental falls away, and what remains is the answer to three questions: who settled it, who controls it, and when. Those questions have the same answers in Wilmington, Nassau and Charlestown — which is why the discipline, rather than the jurisdiction, is what does the protecting.

This note is general commentary for information only and is not legal advice. Bankruptcy and trust outcomes turn on the governing instrument, the applicable nonbankruptcy law, and the facts and timing of each case. The statutory and case law is stated as of September 2026.