A Position Is Not a Structure.

The ruling is PLR 202630002, issued April 28, 2026 out of the Office of Associate Chief Counsel (Financial Institutions and Products), under index number 72.00-00 (I.R.S. Priv. Ltr. Rul. 202630002 (PLR-104852-26), released July 24, 2026). Its entire operative content is a reversal. It revokes an earlier revocation, and it does so with a sentence that deserves to be read slowly by anyone whose planning rests on the current disposition of a federal agency:

Twelve years, three letters, one taxpayer, and a position that has now been held, abandoned, and held again. What follows is what that sequence actually says about where durable protection comes from — and where it does not.

The facts are worth stating carefully, because the sequence is the lesson.

In 2014, in PLR 201424014 (PLR-138374-13), the Service issued two rulings to a life insurance company that proposed to offer a “new term certain annuity option with variable payments” — the letters call it the “New Annuity Option” — in connection with non-qualified deferred variable annuity contracts. The second of those rulings was the valuable one. It held that “[o]n and after the date an [o]wner elects the New Annuity Option, no amount will be includible in gross income before it is actually paid under the New Annuity Option.” In plain terms: electing the payout option did not itself trigger tax on the whole account value. The doctrine of constructive receipt did not bite.

In 2024, the Service changed its mind. PLR 202426001 (PLR-103975-24), issued April 3, 2024 and released June 28, 2024, revoked that second ruling on the stated ground that it “is not in accord with the current views of the Service.” Critically, that revocation “applied prospectively only to contracts with applications signed after” a specified date — the operative date is redacted in the published letter. The taxpayer kept the benefit of the old position for business already written; the door closed going forward.

In 2026, the Service changed its mind back. PLR 202630002 revokes the 2024 revocation, states plainly that “[t]his revocation applies retroactively,” and reinstates the 2014 holding. The intervening two years are, for this taxpayer, treated as though the reversal had not happened.

Practitioner commentary on the release has supplied a good deal of doctrinal architecture for the result — the argument that § 72 by its terms requires that amounts be received; that the TEFRA legislative history assumed deferral until annuity payments are received; that §§ 72(e)(4)(A), 72(u), and 264(a)(3) would be structurally redundant if inside build-up were already taxable under constructive receipt principles.

Those are respectable arguments. But they are not in this letter. PLR 202630002 gives no reasoning whatsoever. It states that the Service reconsidered and that the earlier position was correct, and it stops. A lawyer reading this ruling should be careful not to import a rationale the Service declined to write down, because a rationale that was never articulated cannot be relied upon, distinguished, or defended later. The Service reserved to itself the conclusion without the reasoning — which is precisely the posture that makes a position easy to revisit a third time.

This is not a criticism of the outcome. Deferral until actual payment is a sensible reading of a statute built around amounts “received.” It is an observation about the character of the authority. And that character is the whole subject of this note.

Every private letter ruling carries the same closing paragraph, and PLR 202630002 is no exception:

That is not boilerplate to be skimmed. It is the operative limit. One insurance company asked, paid, disclosed its facts, and received an answer binding as between itself and the Service. No other taxpayer acquired a right in that answer — not in 2014, not in 2024, and not now. A financial adviser who tells a client that “the IRS has ruled that this payout option is tax-deferred” is describing something that is true of one company’s contracts and is, as to everyone else, a well-informed prediction.

The asymmetry runs deeper. Under § 7805(b), the Secretary has discretion to prescribe the extent to which a ruling is applied without retroactive effect, and the annual letter-ruling procedure sets out when a taxpayer may ask for that relief — broadly, where there was no misstatement or omission of material facts, no material change in the facts or the applicable law, the ruling addressed a prospective transaction, and the taxpayer relied in good faith to its detriment (I.R.C. § 7805(b); see Rev. Proc. 2026-1, 2026-1 I.R.B.). Note what that framework means in practice. The 2024 revocation ran prospectively; the 2026 revocation runs retroactively. Both directions were favorable to this taxpayer. But the direction of the ratchet was the Service’s to choose, not the taxpayer’s — and a taxpayer who has to request relief from retroactivity has already lost control of the timeline.

Readers of this series know the recurring themes: protection must be seasoned, set in place years before any claim is on the horizon; it must be irrevocable and discretionary, so that the settlor holds no switch to flip when trouble comes; and it must be independently administered, by a trustee who is genuinely not the settlor wearing a second hat. Improvised, post-claim shielding does not create protection — it creates evidence.

PLR 202630002 adds a distinction to that framework that is easy to miss, because it operates on a different axis entirely.

Seasoning is a defense against a factual attack. When a transfer is made years before any claim exists, made while solvent, disclosed, and never undone, the badges of fraud have nothing to attach to. Time works for the client, and it works in one direction only: every year that a properly built structure stands undisturbed, the case against it gets weaker. Statutes of limitation run. Witnesses’ memories of a benign motive become the record. Under Florida’s fraudulent-asset-conversion statute, for example, a cause of action “is extinguished unless an action is brought within 4 years after the fraudulent asset conversion was made” (Fla. Stat. § 222.30). That is a clock that expires.

An administrative position has no such clock. There is no period after which the Service loses the ability to reconsider. The taxpayer in this saga did nothing wrong in 2024 and nothing right in 2026; the facts never changed. What changed was an internal view, twice, and the taxpayer was a passenger on both trips. No amount of seasoning would have prevented either.

The planning conclusion follows directly: an advantage that rests on an agency’s current view is not protected by the passage of time, and should never be described to a client as though it were.

None of this makes annuity contracts poor planning instruments. It makes precision about which of their advantages is durable essential — because in asset protection, annuities do double duty, and the two duties rest on entirely different footings.

The tax deferral is the contingent part. It depends on statutory construction, regulatory interpretation, and — at the margins where product design gets creative, as with a term-certain option offering payout flexibility — on the Service’s current view of constructive receipt. That view has now moved twice.

The creditor exemption is the durable part, where it exists at all. Several states place annuity contracts beyond ordinary creditor process by statute. Florida’s is among the broadest and has stood, in substance, for a century:

That is a legislative judgment, enacted in 1925 and amended since, not an administrative posture. It changes when a legislature changes it — publicly, prospectively, with notice. This is the kind of authority a structure can be built on. It is also, emphatically, state law: the analysis is different in Texas, different again in New York, and different for a client who moves. Whether any particular contract qualifies turns on its terms, the owner’s and annuitant’s residency, and the facts.

Here the two axes converge, and the familiar lesson reasserts itself.

A statutory exemption is durable, but moving assets into it is an act with a date on it — and that date can be examined. Florida’s § 222.30 addresses exactly this. It defines a “conversion” to reach every mode, direct or indirect, of changing an asset such that its proceeds become exempt from creditors while remaining the debtor’s property, and it makes such a conversion fraudulent where the debtor made it “with the intent to hinder, delay, or defraud the creditor,” giving the creditor avoidance, attachment, injunctive relief, and execution against the converted asset or its proceeds (Fla. Stat. § 222.30).

The pattern that statute was written for is a familiar one: a claim appears, and a debtor liquidates reachable assets to buy an exempt annuity. The exemption itself is not the problem; the timing is. The standard is actual intent, and actual intent is proved the way it always is — circumstantially, from the calendar. A purchase made in a year with no dispute in sight looks like retirement planning. The same purchase made three weeks after a demand letter looks like what it is.

So the annuity ends up illustrating both halves of the point at once. Its tax treatment is exposed to a risk that seasoning cannot cure. Its exemption is exposed to a risk that seasoning cures completely — provided the seasoning actually happened.

The design implication is not that annuities should be avoided, nor that tax positions should be ignored. It is that the load-bearing walls of a plan should be the parts that do not move.

A properly built structure — an irrevocable, discretionary trust, independently administered, funded on a clear day and fully disclosed — does not depend on any letter ruling for its integrity. Its protection comes from the fact that the settlor genuinely parted with the assets, at a time when no creditor was in view, into the hands of a trustee who is not him. If a contract inside that trust also enjoys favorable tax deferral, that is a benefit worth having, and worth revisiting when the authority underlying it moves. But the plan should not fall down if it moves.

Ask, of any advantage a plan is counting on: is this a position or is it a structure? If a two-page letter could undo it, it is a position — and positions are not what clients think they are buying. As always, whether a given arrangement is one or the other turns on the specific facts, the contract terms, and the jurisdictions involved.

PLR 202630002 will be filed away as good news for one insurance company, and it is. But its more lasting service is as a small, unglamorous demonstration of a distinction that matters enormously and is almost never drawn: the difference between an advantage that time makes stronger and an advantage that time does nothing for at all.

Seasoning is powerful — but it is powerful against a particular kind of attack. It defeats the argument that a transfer was made to hinder a creditor. It does not, and cannot, defeat the sentence “the Service has reconsidered its position.”

Build on the things that season. Enjoy the rest while they last. Private letter rulings bind only the taxpayers who obtain them, and exemption and fraudulent-transfer analysis is jurisdiction-specific and fact-specific; this note is offered for general planning discussion, not as legal or tax advice for any particular matter, and not as a prediction about the treatment of any particular contract.