His Own Personal Piggy Bank.

With that sentence, an irrevocable third-party trust settled in 1988 — thirty years before any of the claims that eventually mattered — ceased to be a spendthrift trust, became property of a bankruptcy estate, and was pierced as its beneficiary’s alter ego. Kokoszka v. James Samatas Discretionary Trust U-A-D 9/8/88 (In re Samatas), Adv. No. 21 A 00237, Bankr. No. 20 B 17355, Memorandum Opinion (Bankr. N.D. Ill. Mar. 2, 2026) (Baer, J.), Dkt. 179, at 54. The matter was tried over four days in June, July, and August 2025.

For anyone who has been told that asset protection is a matter of picking the right jurisdiction and signing the right document, Samatas is the necessary corrective. This trust had every structural advantage our field says to look for. It was settled by a third party, not the beneficiary. It contained a conventional spendthrift clause. It was drafted by a large Chicago firm, with a deliberate technique for keeping the beneficiary away from his own distributions. It had been in existence for more than three decades before the creditors arrived — seasoning of a kind almost no client ever achieves.

And it failed completely, because nobody ever actually administered it.

George Samatas established the Trust on 8 September 1988 for the benefit of his son James, one of three identical discretionary trusts he created for his three children. Id. at 4 & n.7. Article IX contained a customary spendthrift provision barring transfer, assignment, encumbrance, or creditor interference with income or principal prior to its receipt by the beneficiary. Id. at 4–5, 39. Neither party disputed that this was, on its face, a standard and effective anti-alienation clause; the trustee’s own expert conceded it was “pretty standard.”

The drafting went further than most. James Samatas served as a co-trustee alongside Craig Labus — a family friend and the family’s accountant of twenty-five years. But the instrument designated James a “restricted trustee,” and Article XIV § 3 provided that a trustee who is also a beneficiary “shall [not] have any voice, determination or vote relating to any discretionary distribution.” Id. at 5–7. All distribution decisions belonged to Labus alone. The defendants’ own expert explained the design at trial with admirable clarity: a beneficiary who can make discretionary distributions to himself “would defeat the whole purpose of that trust.” Id. at 7.

That is a sound instinct, and it is the instinct behind a great deal of modern trust drafting. It did not survive contact with the second half of the document.

James Samatas was not only beneficiary and restricted co-trustee. He was also the Trust’s investment advisor under Article VII § 1 — and under that Article, the trustee “shall sell, vote or take any other action with respect to the investment of the trust assets only upon the written instructions of said Investment Advisor,” with the trustee relieved of any duty to review those instructions. Id. at 6.

Bifurcating investment control from distribution control is, as the defendants’ expert testified, a common and legitimate technique. The court’s objection was not to bifurcation in the abstract. It was that in this instrument the investment-advisor power was broad enough to swallow the restriction it sat beside.

The debtor confirmed the point in his own testimony. He said the Trust let him, as investment advisor, direct a loan of trust funds to himself and then unilaterally forgive it; that he could “act almost . . . as if [he were] an unrestricted trustee”; and that Labus “did not have to approve the investment advisor thing. He had to know about it; he didn’t have to approve it.” Id. at 9, 52–53. On the strength of the instrument alone — before any evidence of conduct — the court found that the debtor “was able to reach the Trust assets at any time and had the power to make transfers from the Trust at will.” Id. at 53.

The defendants’ central legal argument was that none of the debtor’s conduct mattered. Under Illinois law, they said, a court determines whether a trust is a spendthrift trust by reading the trust.

Illinois courts apply a three-part test drawn from the Seventh Circuit’s decision in In re Perkins, 902 F.2d 1254, 1257 n.2 (7th Cir. 1990): whether the trust restricts alienation and creditor attachment; whether the beneficiary settled the trust and retained a power to revoke; and whether the beneficiary “has exclusive and effective dominion and control over the trust corpus, distribution of the trust corpus and termination of the trust.” Id. at 38. The federal frame is 11 U.S.C. § 541(c)(2), which excludes from the estate a restriction on transfer of a beneficial interest “enforceable under applicable nonbankruptcy law,” see Patterson v. Shumate, 504 U.S. 753, 759 (1992); Illinois’ statutory analogue is 735 ILCS 5/2-1403. Id. at 36–37, 40–41. The first factor was satisfied here and the second was not seriously in dispute — George Samatas, not his son, was the settlor. Everything turned on the third.

The court held that whether a beneficiary’s actual conduct may be considered under that third factor was a question no Illinois court, state or federal, and neither the Supreme Court nor the Seventh Circuit, had squarely decided. Id. at 45. It answered the question by adopting the reasoning of Richardson v. McCullough (In re McCullough), 259 B.R. 509, 518–19 (Bankr. D.R.I. 2001), a Rhode Island bankruptcy decision holding that a trust cannot be examined “in a vacuum” where the conduct of the parties discloses unfettered dominion over the assets. Id. at 46–47, 49.

The textual anchor is worth noting, because it will be cited for years. The first Perkins factor asks what “the trust” restricts — a document question. The third asks whether the beneficiary has “exclusive and effective dominion and control.” “Effective,” the court reasoned, means “existing in fact; actual” — which directs a court to “the practical reality of the beneficiary’s power.” Id. at 48–49, citing Lamar, Archer & Cofrin, LLP v. Appling, 584 U.S. 709, 715 (2018). A beneficiary who, “even as a co-trustee, can actually compel distributions or influence decisions” has effective dominion and control, whatever the instrument recites.

That is the holding planners should file away. In Illinois, and increasingly elsewhere, the question is not what the trust deed says the beneficiary may do. It is what the beneficiary has in fact been doing.

The factual record was, in the court’s words, one in which “seldom has the Court been presented with a matter in which a required element has been so compellingly established.” Id. at 54. The findings map almost item for item onto the things a trust officer is supposed to prevent:

• The debtor was the sole signatory on the Trust’s bank accounts. The checks bore his personal name at the top with no reference to the Trust at all. He paid personal living expenses of $5,000 to $10,000 a month from those accounts using a debit card — groceries, gasoline, Amazon, pet supplies. Id. at 18, 55.
• Disbursements to the debtor were booked as “loans” on an open account with no promissory notes and no ledgers; the balance exceeded $19 million before being “forgiven” in tax year 2023. A separate $10 million revolving line of credit had a repayment schedule that was never filled in. Id. at 11–12.
• The debtor sold Trust artwork and paperweights in his own name, through consignment agreements that never mentioned the Trust, and deposited the proceeds into his personal or revocable-trust accounts. In one instance he directed the buyer to make the check payable to his nephew. Id. at 16–17, 57.
• He caused the Trust to guarantee his personal legal fees and to pledge Trust-held corporate stock as security for that guaranty — a transaction the family’s own estate-planning counsel testified he had never been consulted about and would never have advised. Id. at 19.
• Labus, the trustee in whom every distribution decision was supposed to vest, was not a signatory on any account, never took an inventory of Trust property, did not know which assets the Trust owned, and learned of asset sales only if the debtor told him. The trustee’s own answer, when asked how he would know whether assets had been sold: “I have to be able to trust my beneficiary and my co-trustee.” The opposing expert concluded he “seems to have abdicated the role of trustee.” Id. at 15–16, 18, 57, 68.
• Almost none of it was documented. The court imposed an adverse inference that the written communications and loan ledgers the defendants testified about do not exist, because none were produced in response to subpoenas. Id. at 34–35.

The court found the spendthrift provision unenforceable and the Trust assets property of the bankruptcy estate. It then went further and found the Trust to be the debtor’s alter ego — a “mere façade” — permitting outside reverse veil piercing so that the bankruptcy trustee could reach the corpus directly on behalf of creditors. Id. at 61, 66–70, relying on Reid v. Wolf (In re Wolf), 644 B.R. 725 (N.D. Ill. 2022), aff’d, 2023 WL 6564882 (7th Cir. Oct. 10, 2023), and Goldstein v. Graft (In re Graft), 2025 WL 45085 (Bankr. N.D. Ill. Jan. 7, 2025).

One further piece of the decision deserves separate attention, because it involves a mistake clients make constantly.

On 2 July 2018, a California court entered judgment in the debtor’s long-running divorce, requiring him to pay his former spouse $1,740,000 in maintenance at $20,000 per month. The next day, he executed a bill of sale transferring roughly $3 million of artwork, paperweights, jewelry, and furniture into the Trust. The property never moved; it stayed in his house in Oak Brook, and he continued to sell pieces of it at will. Id. at 13–16, 79.

The bankruptcy trustee attacked that transfer on two theories. The constructive-fraud count under section 6(b) of the Illinois Uniform Fraudulent Transfer Act failed — the trustee simply did not put on evidence that the debtor was insolvent in July 2018, and insolvency is an element of that claim. Id. at 76–77 (740 ILCS 160/6(b)). The actual-intent count under section 5(a)(1) succeeded, and it did not need insolvency at all. The court found the badges of fraud stacked: transfer to an insider; possession and control retained after the transfer; the transaction undisclosed to anyone and omitted from the debtor’s financial statements; litigation already threatened and pending before the transfer, including a 5 June 2018 demand from a sibling’s trust and a law firm’s counterclaim seeking in excess of $4 million; the transfer made immediately after a substantial debt was incurred; and consideration that was not reasonably equivalent, the $3 million figure having been agreed by the debtor with himself on both sides of the paper. Id. at 77–83 (740 ILCS 160/5(a)(1), (b)). The transfer was avoided and its value made recoverable under 11 U.S.C. § 550(a). Id. at 84.

The point for clients is blunt. A trust that has stood since 1988 confers no protective glow on a transfer made in 2018 on the day after a judgment. Seasoning attaches to the transfer, not to the vehicle. Every funding event is separately vulnerable, and a decades-old trust used as a receptacle for eleventh-hour transfers simply hands the creditor a fraudulent-transfer claim wrapped in an alter-ego claim.

Samatas is not evidence that trusts do not work. It is evidence of the four things that make one work, tested to destruction.

1. Seasoning is necessary but not sufficient. Thirty years of history bought this debtor nothing, because the conduct that destroyed the trust was recent and continuous. Time protects a transfer; it does not protect a course of administration.

2. Independent administration has to be real, and it has to be documented. Labus was structurally the right answer — a non-beneficiary trustee holding sole distribution authority. He failed not because of who he was but because of what he did not do: no signature authority, no inventory, no minutes, no scrutiny, no records. A trustee who is loyal but passive is, for creditor purposes, indistinguishable from no trustee at all. An institutional or professional trustee that keeps books, holds the accounts in the trust’s name, and says no from time to time is not an administrative expense. It is the protection.

3. Retained powers must be read for what they let the client actually do. The “restricted trustee” device in this instrument was thoughtful drafting. It was defeated by an investment-advisor power that let the same person direct loans to himself and forgive them. When reviewing an existing structure, the right question is not “what does the document call this role?” but “what is the sum of every power this person holds, and could he reach the corpus by combining them?”

4. Formalities are not paperwork; they are the evidence. Separate accounts in the trust’s name. No personal debit cards. Written instructions where the deed requires written instructions. Loans that look like loans — notes, schedules, interest, repayment. Assets inventoried, insured in the trust’s name, and sold in the trust’s name with proceeds going to the trust. Each of these is trivial to do at the time and impossible to reconstruct afterwards. This debtor’s problem at trial was not that the answers were bad. It was that there were no records to answer from, and the court drew the inference that there never had been any.

Whether a given structure holds up will always turn on its own facts, on the governing instrument, and on the law of the jurisdictions involved — Samatas is an Illinois case decided under Illinois spendthrift law and the Illinois UFTA, and other jurisdictions draw these lines in their own places. But the direction of the reasoning is not parochial. Courts are increasingly willing to look past the deed to the conduct, and conduct is the one variable entirely within a client’s control.

The most uncomfortable fact about Samatas is that on the day it was signed in 1988, this trust looked better than most of what clients bring us. Third-party settlor. Irrevocable. Discretionary. Spendthrift clause. A restriction preventing the beneficiary from voting on his own distributions. Thirty years of runway.

It was administered into nothing — one debit-card charge, one undocumented “loan,” one consignment agreement in the wrong name at a time — and when the creditors finally came, the court found that the separate personalities of the trust and the man “no longer exist.” Id. at 69–70.

Protection is not an event that happens when a document is executed. It is a practice, sustained over years, of treating the trust as what it claims to be: someone else’s property, held by someone else, for purposes the client does not control. Clients who find that discipline inconvenient should understand what they are trading it for.

This note is general commentary on a published decision and is not legal advice for any particular person or matter. How any of it applies depends on the specific instrument, the facts, and the jurisdictions involved, and should be assessed with qualified counsel.