When the Regulation Falls.

That backdrop is gone. On July 2, 2026, the United States Court of Federal Claims held that a Treasury regulation policing the boundaries of the GILTI regime — Treas. Reg. § 1.951A-2(c)(5) — was invalid because Treasury lacked the statutory authority to issue it (Keysight Technologies, Inc. & Subsidiaries v. United States, No. 25-137 (Fed. Cl. July 2, 2026) (Tapp, J.)). The case is worth the attention of anyone whose planning rests, even in part, on the assumption that a Treasury regulation will hold. Increasingly, it may not. The ground under aggressive regulatory positions has begun to move, and Keysight is a marker of how far.

To see why Keysight matters beyond its own facts, it helps to understand the narrow mechanism the court struck down.

The 2017 Tax Cuts and Jobs Act built the modern regime for taxing the foreign earnings of U.S. multinationals — chiefly the Global Intangible Low-Taxed Income (“GILTI”) rules of I.R.C. § 951A, which subject a U.S. shareholder’s controlled foreign corporations (“CFCs”) to current U.S. tax on much of their income, whether or not it is repatriated. But the statute’s effective dates did not line up perfectly. Because GILTI applies to CFC tax years beginning after December 31, 2017, CFCs on a fiscal year phased in on a different calendar than CFCs on a calendar year. That mismatch opened a window — Treasury called it the “disqualified period” — during which assets could be transferred from a fiscal-year CFC to a calendar-year CFC, producing a stepped-up basis without a corresponding current U.S. tax cost.

Treasury viewed that outcome as abuse and legislated against it by regulation. Treas. Reg. § 1.951A-2(c)(5) allocated the amortization and depreciation deductions attributable to those stepped-up, transferred assets solely to a CFC’s residual (non-GILTI) income — so that the deductions could not be used to reduce the taxpayer’s GILTI, and thus its current U.S. tax. Keysight, having made such transfers, claimed the deductions against GILTI for tax years 2020, 2021, and 2022, and sued for refund when the regulation stood in the way. By the company’s own account the stakes, projected through 2033, ran to the hundreds of millions of dollars — “upwards possibly of $500 million” (Current Federal Tax Developments, July 6, 2026).

The court did not decide whether Treasury’s policy was wise. It decided whether Treasury had the authority. That is the shift Loper Bright wrought.

In Loper Bright Enterprises v. Raimondo, 603 U.S. 369 (2024), the Supreme Court overruled Chevron and returned to the courts the job of saying what a statute means. Agency interpretations, the Court held, “are not entitled to deference”; they may earn respect under the older Skidmore standard, but only to the extent of “the thoroughness evident in [the agency’s] consideration, the validity of its reasoning,” and the like. A regulation no longer wins simply because the statute is ambiguous and the agency’s reading is reasonable. It must be authorized — and it must be right.

Judge Tapp’s opinion applied that framework and found the regulation wanting on both counts.

First, authority. Treasury leaned on § 7805(a), its general grant of power to prescribe “all needful rules and regulations.” The court held that this general authority, “standing alone, is insufficient to supply the requisite authority to the Secretary.” To read § 7805(a) as a roving license to fill perceived gaps, the court observed, would “invalidate Loper Bright itself” — because every substantive regulation could then be justified by the same open-ended grant. Nor did the specific GILTI provisions help Treasury: § 951A(c)(2)(A)(ii), which defines GILTI by reference to deductions “properly allocable” to gross tested income, “does not grant the Treasury specific authority, neither expressly nor impliedly,” to redefine which deductions qualify.

Second, meaning. Even setting authority aside, the regulation collided with the statute’s ordinary sense. “Properly allocable,” the court held, carries an established meaning drawn from the § 861-8 allocation principles — deductions “factually related” to the income they offset. The transferred assets’ deductions were factually related to the CFC’s tested income; the regulation’s contrary allocation “solely to residual” income was a policy choice dressed as interpretation.

And on the label of “abuse,” the court was blunt: “If a taxpayer acts within the literal bounds of the statute as Congress chose to enact it, characterizing Keysight’s … conduct as ‘abusive’ is simply incorrect.” The gap was Congress’s to close. Treasury could not close it by fiat and call the taxpayer who walked through it a wrongdoer.

The court granted Keysight’s partial motion for summary judgment and denied the government’s cross-motion, invalidating the regulation.

Keysight would be easy to file away as a one-off victory for a well-advised multinational. It is better read as one data point in a line.

Nearly two years earlier, in Varian Medical Systems, Inc. v. Commissioner, 163 T.C. No. 4 (Aug. 26, 2024), the Tax Court applied the same post-Loper Bright logic to a different TCJA effective-date mismatch — the gap between the § 245A dividends-received deduction and the § 78 “gross-up” — and invalidated the portion of Treas. Reg. § 1.78-1 that tried to legislate the gap shut. Where the statute’s text was clear, the court held, a regulation contradicting it fell “outside the boundaries of any authority” Treasury had been delegated.

The pattern is consistent, and it is instructive. In both cases Congress, drafting in haste, left a timing seam in a major international provision. In both cases Treasury tried to sew the seam closed by regulation, invoking anti-abuse purposes. And in both cases the courts, freed by Loper Bright from the obligation to defer, asked the antecedent question — did Congress give you this power? — and answered no. A planner watching this line should draw the general lesson, not merely the specific one: regulations that stretch beyond their statutory tether are now materially more vulnerable than they were three years ago, and “anti-abuse” is not a magic word that supplies the missing authority.

For Lighthouse clients, most of whom are individuals and families rather than multinationals, the direct holding of Keysight will rarely be on point. GILTI, disqualified periods, and § 861-8 allocation are the concerns of the corporate international group. But the meta-lesson reaches every planner, and it cuts in a direction our clients should welcome — provided they understand it correctly.

First, durable planning is built on the statute, not on a regulatory gloss that may not survive. The lesson of Keysight and Varian is not “ignore the regulations.” Regulations remain in force until a court sets them aside, penalties attach to positions taken against them without adequate disclosure, and the ordinary taxpayer who bets against a reg and loses pays interest and worse. The lesson is subtler: a position whose only support is that a court will deny Treasury deference is a fragile position, because deference now cuts both ways. The soundest structures are those that would stand whatever a court decided about a contested regulation — because they rest on the plain terms of the statute and on transactions that are what they purport to be.

Second, aggressive regulatory positions are a poor foundation for a plan meant to last decades. Trust and estate planning is generational by design. An irrevocable structure settled today may be administered for forty years and tested by a claim or an audit that no one can foresee. To hang such a structure on the current, contestable reading of a single regulation — one that a future Keysight might vacate, or that a future Treasury might rewrite, or that a future Congress might override — is to build a long-lived thing on a short-lived footing. The virtues we return to in these notes apply with equal force to tax exposure as to creditor exposure: a plan should be seasoned, its transactions real and completed rather than paper, and its posture transparent rather than dependent on a favorable turn of legal weather. A structure whose validity depends on concealment, on retained control, or on a single regulatory interpretation holding is not seasoned — it is exposed.

Third, the shift favors the taxpayer only until it doesn’t. It is worth saying plainly that Loper Bright is not uniformly a taxpayer’s friend. The same principle that let Keysight defeat an unauthorized anti-abuse regulation can, in another case, let the government defeat a taxpayer-favorable regulation the taxpayer had relied upon. Deference is symmetrical; its removal is symmetrical too. A regulation that today shelters a planning position is, tomorrow, just as susceptible to challenge as the one Keysight defeated. That is precisely why the durable answer is to anchor planning in statutory text and genuine substance, and to treat any position that lives or dies on a single regulation as provisional — to be revisited as the case law develops, not set and forgotten.

Keysight is, on its surface, a corporate international tax case about amortization deductions and a drafting seam in the GILTI rules. Read for its lesson, it is something larger: a clear signal that the era in which a Treasury regulation was, for planning purposes, indistinguishable from the statute has ended. The courts are again reading the statute for themselves, and regulations that reach past their statutory grant — however sensible the policy behind them — are falling.

For those of us who build structures meant to endure, the counsel is old and unglamorous. Plan on the ground that does not move: the words Congress actually enacted, transactions that are genuinely what they say they are, and a posture that would survive daylight. Regulations will come and go, and after Loper Bright they will go more often than they used to. Wealth planned to last should not be moored to any one of them.