Who Needs to Leave Behind a Garrison

Source: https://bryantmcgill.com/article-who-needs-to-leave-behind-a-garrison

There is a kind of defeat that leaves no ruins. No fleet appears on the horizon. No landing is contested, no capital occupied, no instrument of surrender signed on the deck of anything. The flag does not come down. The legislature continues to sit. The currency continues to circulate, the courts continue to rule, the armed forces remain the most formidable ever assembled by a political community, and the citizen wakes each morning inside a country that describes itself, accurately by every conventional measure, as the most powerful state in the history of the species.

And yet somewhere else — in a registry office on a nine-square-mile island, in a laboratory accreditation schedule, in a technical committee meeting in the south of France, in a statistical footnote quietly retired in February 2016 — decisions are being made that determine what the American state can borrow, what the American consumer may lawfully purchase, what the American manufacturer must build, and what the American government is permitted to see about its own finances. None of those decisions is made by an American. None requires anyone’s consent. None can be reversed by any weapon in the American arsenal.

This is the war that never ended, conducted with instruments that were never recognized as weapons.

The premise of this entire sequence — traced through the cable war, the standards war, the prestige apparatus, and the doctrine of hemispheric completion — has been that 1776 resolved the question of allegiance and left the question of dependency entirely untouched. The colonies severed the political bond. They did not sever the financial bond, the legal bond, the insurance bond, the maritime bond, the standards bond, the certification bond, or the epistemic bond, and in the two and a half centuries since, those bonds have not withered. They have migrated downward, into substrate, where they are harder to see and vastly more consequential. Empire is not reducible to redcoats and governors. Empire is the retained capacity of one political center to determine the conditions under which another exercises sovereignty — and that capacity survives the withdrawal of every soldier.

So conduct the thought experiment honestly. Suppose you were charged with defeating the United States from within, without war, without sabotage, without a single act that any court could name as hostile, and with the requirement that the operation remain permanently deniable and permanently legal. What would you build?

You would build exactly two things.

You would arrange for the marginal financier of the American state — the buyer who absorbs new debt at the margin and therefore sets the price of American fiscal capacity — to be domiciled in a jurisdiction whose law is not American, whose regulator Washington cannot reach, whose beneficial ownership Washington cannot compel, and whose position Washington’s own statistical apparatus cannot measure. You would ensure that this position was maximally leveraged and funded overnight, so that its unwinding would be violent and its continuation would be a permanent hostage.

And you would arrange for the gate through which every electronic device enters American commercial life — the body that decides what may lawfully be sold, installed in a hospital, mounted on a wall in a school, plugged in at a military installation — to be owned outside American jurisdiction, accredited outside American jurisdiction, and operating against technical scopes authored outside American jurisdiction, with the examination deliberately bounded so that the most consequential capabilities inside the device pass through unexamined and undisclosed.

Both of these things exist. Both are documented in the public record of the United States government. Neither is a secret.

And here is the part that ought to disturb a policymaker more than any allegation of design: nobody built them. They accreted, incrementally, largely through American choices made for reasons of efficiency, market liberalization, and administrative convenience, over decades in which the jurisdictions in question were understood as friendly and the delegations in question were understood as technical. That is precisely why they have never been recognized as what they are. A conspiracy can be exposed. An architecture that emerged from a thousand individually rational decisions has no author to indict, no meeting to subpoena, no document to declassify — and no natural constituency for its repair.

What follows is an accounting of both columns in the fullest detail the public record permits.

Before the evidence, the concept — because without it the evidence reads as a collection of grievances rather than a single structure.

Every state performs a small number of functions that are not merely administrative but constitutive. They do not regulate activity that would otherwise exist; they determine whether a thing exists at all in the eyes of law. Whether a security is owned, and by whom. Whether a company is a person. Whether a risk is insured. Whether a device may be sold. Whether a model may be deployed. These are adjudication functions, and they are the load of sovereignty in the same sense that a footing is the load of a building: invisible, unglamorous, and the thing that fails first when it fails.

Territory once determined adjudication, because adjudication required physical presence — a courthouse, a customs officer, a registrar with a ledger. That coupling has dissolved. Adjudication is now performed through registries, accreditations, statistical instruments, and technical scopes, none of which requires occupation of ground and all of which can be domiciled anywhere. Territory now matters chiefly because of what can be instantiated upon it, and adjudication no longer needs to be instantiated on the territory it governs.

The consequence is exact and, once seen, difficult to un-see. A state that holds every acre of its territory, commands every approach to its coastline, and fields every carrier group it can afford may nonetheless have surrendered the determination of what is owned within it, what may be sold within it, and what its own government is able to observe about itself. It will not feel like a loss of sovereignty. It will feel like efficiency. And it will be defended, when questioned, in exactly those terms.

The two columns that follow are both instances of this single mechanism. That is why they belong in one document rather than two.

Begin where a policymaker would begin, with the Treasury’s own Major Foreign Holders of Treasury Securities table, December 2025, in billions of dollars.

Japan appears first at 1,185.5. The United Kingdom is second at 863.1. Mainland China is third at 684.4. Belgium follows at 477.3, Canada at 468.3, Luxembourg at 434.0, and the Cayman Islands at 421.6. Further down, Bermuda appears at 101.4. The grand total of all foreign holdings is 9,269.5.

Read as published, the table tells a story of diversified international demand: an Asian creditor, a European financial center, a strategic competitor, and a scattering of custodial jurisdictions. It is the story every commentary on American debt has told for twenty years.

Now perform an arithmetic operation the table’s own structure discourages. The United Kingdom, the Cayman Islands, and Bermuda are three separate line items. They are also three jurisdictions whose ultimate constitutional authority is the same, whose final court of appeal is the same body sitting in London, and whose governors — in the case of the latter two — are appointed by the Crown. Summed, they are $1,386.1 billion.

That figure exceeds Japan. Using nothing but the Treasury’s published numbers, unadjusted, the largest concentration of foreign-held United States sovereign debt sits inside British constitutional space — and it has never been presented that way, because the table does not present it that way.

It once did. Which brings us to the footnote.

At the bottom of the same table, in the notes, sit two discontinued categories.

Footnote 2 records that the United Kingdom line formerly included the Channel Islands and the Isle of Man. Footnote 4 records a category called Caribbean Banking Centers, comprising the Bahamas, Bermuda, the Cayman Islands, the Netherlands Antilles and Panama, and — from a new series beginning June 2006 — the British Virgin Islands.

Both aggregates were discontinued after data for February 2016.

The two statistical categories that would have rendered the offshore and British-jurisdiction concentration legible as a single number were retired a decade ago. Every analyst, every legislative staffer, every journalist writing about foreign ownership of American debt since 2016 has read a table in which Cayman, Bermuda, Jersey, Guernsey, the Isle of Man and the British Virgin Islands appear either as unrelated line items or nowhere at all, submerged in an “All Other” residual that stood at $710.8 billion in December 2025.

No claim of intent is necessary. The disaggregation is the finding, and it is published on Treasury’s own server. A state that cannot see a pattern in its own data will not act on it, and the categories that made the pattern visible no longer exist.

In October 2025, Federal Reserve Board staff — Daniel Barth, Daniel Beltran, Matthew Hoops, R. Jay Kahn, Emily Liu and Maria Perozek — published a FEDS Note titled The Cross-Border Trail of the Treasury Basis Trade. Its conclusion, stated without ornament, is that the Treasury International Capital system undercounts Cayman-domiciled hedge fund holdings of United States Treasuries by approximately $1.4 trillion as of the end of 2024, with regulatory data from Form PF and the IMF’s portfolio survey both confirming a rise the official statistics do not show.

Adjusted, the Cayman position is not $421.6 billion. It is somewhere between $1.8 and $2.0 trillion — larger than Japan, larger than the United Kingdom, larger than China. Adjusted, total foreign holdings are not $9.27 trillion but closer to $10.7 trillion. Adjusted, the British-jurisdiction bloc identified above approaches $2.79 trillion, or roughly twenty-six percent of all foreign-held United States sovereign debt — before adding Jersey, Guernsey, the Isle of Man and the British Virgin Islands, which remain invisible inside the residual.

But the detonating figure is not the stock. It is the flow, and it appears in the paper’s fifth chart, covering absorption of net issuance of Treasury notes and bonds from January 2022 through December 2024.

Over that window, foreign official holders — the central banks and sovereign reserve managers who dominate every public discussion of American debt — absorbed $32 billion.

Cayman-domiciled hedge funds absorbed $1,225 billion.

That is roughly thirty-seven percent of net note and bond issuance, an amount equal to every other foreign investor in the world combined, during a period when the Federal Reserve was running off $1,208 billion of its own holdings. The Fed’s authors state the implication directly: Cayman hedge funds are increasingly the marginal foreign buyer of American sovereign debt.

The sovereign-creditor frame is obsolete. The recurring anxiety about Beijing dumping Treasuries describes a mechanism that has been superseded. The marginal financier of the United States government is not a foreign state. It is a levered arbitrage position domiciled in a British Overseas Territory of ninety thousand residents.

Here intellectual honesty is not merely a virtue but a strategic necessity, because the obvious reading is wrong and a policymaker will catch it in ten seconds.

Those Cayman funds are overwhelmingly American-managed. Greenwich, Manhattan, Chicago. They are American entities registered in George Town with negligible physical presence, structured there for tax neutrality, investor-mix reasons, and regulatory flexibility — a characterization stated plainly by Vítor Constâncio, the former Vice-President of the European Central Bank, in his own reading of the Fed’s finding. Nobody in London is deciding to buy American debt. If this document claimed that Britain owns America’s borrowing, it would deserve to be dismissed.

The accurate formulation is considerably worse.

American capital finances the American state through a legal jurisdiction that is not American — where the applicable company law, trust law, and insolvency regime are not American; where the supervisory authority is not American; where beneficial ownership cannot be compelled by American process; where the final appellate authority is the Judicial Committee of the Privy Council; and where, until eight months ago, the position was invisible to the United States Treasury’s own measurement system.

The consequence is not that a rival controls the position. It is that no one who answers to the American electorate can observe it, supervise it, stress it, or wind it down. In a crisis, the Treasury Secretary would be attempting to manage the disorderly unwinding of the largest single concentration of foreign-domiciled demand for American debt while possessing neither jurisdiction over the funds, nor authority over their regulator, nor — until the Fed’s own researchers reconstructed it — an accurate measurement of its size.

The position is not merely offshore. It is structurally brittle in ways that are documented across the Federal Reserve system, the Office of Financial Research, the Financial Stability Board, and the IMF.

The trade itself is simple: buy the cash Treasury, sell the corresponding future, finance the bond in repurchase agreements, and harvest the price differential. Its profitability depends entirely on leverage, and the leverage depends on haircuts. As of the last comprehensive figures, 73.8 percent of hedge fund repurchase-agreement borrowing occurred at zero or negative haircuts. Fed officials have cited leverage ratios above 18:1 for the largest funds; a 2023 Board note estimated effective leverage in Treasury repo at 56x in the absence of a haircut floor, falling to roughly 25x if a two-hundred-basis-point minimum were imposed.

The aggregate has grown accordingly. Hedge fund net repo borrowing reached approximately $1.8 trillion by the end of 2025 — about six percent of all marketable notes and bonds outstanding. Board staff estimated aggregate cash-futures basis trade volumes at roughly $830 billion in September 2025, close to double the early-2020 peak. Research from the Dallas Fed has found the expansion of this activity associated with ten to twenty basis points of widening between secured and unsecured funding rates, meaning the position is now large enough to interfere with the transmission of monetary policy itself.

The precedent for its failure is not hypothetical. The March 2020 Treasury market dysfunction — the episode that required the most aggressive central bank intervention in the instrument’s history — has been attributed in substantial part to the rapid unwinding of exactly these positions, at roughly half their present scale. Governor Lisa Cook named leveraged hedge fund Treasury positions a systemic vulnerability on 20 November 2025.

Assemble the picture. The marginal buyer of American sovereign debt is levered somewhere between eighteen and fifty-six times, funded overnight, at haircuts approaching zero, in a volume around six percent of the outstanding note and bond market, in a jurisdiction the United States does not supervise, in a position the United States could not measure until October 2025, with a demonstrated capacity to seize the market when it unwinds. A hostile actor asked to design a hostage arrangement could not improve on this, and no hostile actor designed it.

The instinct at this point is to conclude that the leverage runs to London — that the Privy Council, the appointed Governor, and Westminster’s constitutional supremacy amount to a hand on the valve. That instinct is half right, and the half that is wrong matters enormously.

The formal authority is real. Parliament is constitutionally supreme over the Overseas Territories with no limits on its ability to legislate for them, and the Privy Council can legislate directly by Order in Council, altering local law and even local constitutions.

Now observe what happened when Westminster tried to use it.

Section 51 of the Sanctions and Anti-Money Laundering Act 2018 required a Secretary of State to prepare, no later than 31 December 2020, a draft Order in Council compelling any Overseas Territory without a publicly accessible register of beneficial ownership to introduce one. That Order has still not been made. Subsequent amendments seeking to force the timetable were defeated in 2022 and 2023.

What the Territories produced instead was a slow, technically sophisticated retreat. The Cayman Islands launched a legitimate interest register in February 2025 — not a public one — and its 2026 regulations attach a CI$250 annual access fee and a demonstrated-interest test administered locally. The British Virgin Islands slipped to April 2026 and was placed under increased monitoring by the Financial Action Task Force. The Falkland Islands deferred to July 2026 on capacity grounds. And the Territories have found a constitutional shield in European jurisprudence: the Court of Justice of the European Union’s judgment in Sovim SA v Luxembourg Business Registers, holding that fully public beneficial ownership registers constitute a disproportionate interference with privacy rights — a ruling not binding on any of them, invoked by all of them, and now the stated basis on which further UK pressure may be resisted in local courts.

Eight years. Formal constitutional supremacy. Repeated ministerial pressure and a statutory mandate. And the world’s first, second and fourth most damaging corporate tax havens have conceded a paywalled register with a discretionary access test.

This is the finding that reorganizes the column. The Americas do not contain a British-controlled financial appendage. They contain a jurisdictional void — formally British, operationally autonomous, substantively serving global and largely American capital — that neither London nor Washington governs. An adversary-controlled chokepoint is a negotiation. An ungoverned one is not, because there is no counterparty with standing to close it. The uncontrolled version is strictly worse than the controlled version, and it is the version that exists.

Sovereign debt is the largest exposure but not the only one.

The British Virgin Islands hosts on the order of 360,000 to 400,000 active companies against a resident population near 32,000 and ranks first on the Corporate Tax Haven Index. Bermuda is the world’s principal venue for the cession of United States peak-zone catastrophe risk — Florida hurricane, California earthquake. Bermuda reinsurers account for approximately thirty-six percent of the global reinsurance market measured by property and casualty net premiums earned; the market carried $219.87 billion of gross loss exposure in 2024, paid nearly $1.1 trillion in claims between 2016 and 2024, and supplies roughly a quarter of the reinsurance behind the Texas Windstorm Pool. Notably, seventy-six percent of gross premium written there is written by United States SEC registrants — the same pattern as Cayman, and the same conclusion: American risk, American capital, non-American law.

The Tax Justice Network’s current ranking places the British Virgin Islands first, the Cayman Islands second, Switzerland third, Bermuda fourth, and Jersey eighth. Eight of the ten jurisdictions with the highest haven scores belong to the United Kingdom’s network, which the index credits with roughly thirty-three percent of all global corporate tax abuse risk and an estimated $84 billion in annual revenue losses to other countries. The United Kingdom itself ranks well down the list — a structure the Network characterizes as one set of rules at home and another for the satellites.

The pricing of American catastrophe risk, the incorporation of a substantial fraction of the world’s shell entities, and the financing of the American state all pass through the same constitutional space. Not through occupation. Not through garrison. Through registry, trust law, regulatory arbitrage, and a statistical instrument that could not see $1.4 trillion of its own subject matter.

The second column has been argued badly, including by this author, and the bad version must be dismantled before the strong version can stand.

It is sometimes claimed that FCC certification verifies the presence of CALEA-compliant lawful-intercept capability in consumer devices. This is false. The Communications Assistance for Law Enforcement Act binds telecommunications carriers and the manufacturers of carrier-plane equipment — switches, transmission systems, network infrastructure — and is enforced through an entirely separate process. FCC equipment authorization under Part 2, Subpart J concerns spectrum use, radio frequency emissions, and electromagnetic compatibility. A Telecommunications Certification Body does not test for interception capability, is not asked to, and possesses no protocol by which to do so.

It is likewise claimed that a chain runs from ETSI’s Technical Committee on Lawful Interception through EN 303 645 to the certifier and thence to signals collection. It does not. Those are separate technical committees producing non-interoperating deliverables, and certification against the consumer IoT baseline touches no lawful-interception specification whatsoever.

Both claims are discarded here. They are unnecessary, and they are weaker than the truth.

The actual structure is this. The certification surface is definitionally limited to declared parameters, and everything strategically significant operates beneath the declaration threshold.

Part 15 measures declared emissions. EN 303 645 evaluates declared interfaces, declared update mechanisms, and declared data-protection provisions. Safety listing evaluates thermal, electrical and mechanical hazard. None of these examines the Intel Converged Security and Management Engine or AMD’s Platform Security Processor — autonomous subsystems present across essentially the entire x86 installed base, running independent processors with their own memory and cryptographic engines, possessing full access to system memory without the owner-controlled cores being aware, holding an independent network stack that bypasses the operating system’s firewall, and remaining active while the host system is powered off but connected.

These are not vulnerabilities. They are documented, intended product architecture, described in the manufacturers’ own datasheets and security white papers, with manufacturing fuses set by the chip vendor before shipment and field-programmable fuses set subsequently by the OEM — a two-stage cryptographic provisioning that occurs before the device reaches any purchaser. Nor does the certification examine firmware provenance, supply-chain custody, or side-channel emission behavior below the declared threshold.

The device therefore passes through the gate with its most consequential capabilities unexamined and undisclosed, and it does so entirely lawfully, because the certifier is executing the published scope faithfully. The proper object of scrutiny is not the certifier’s integrity. It is the scope — and the scope is authored in Geneva, Sophia Antipolis, and Brussels.

The assertion that one can hardly obtain anything in America that is not Intertek-certified is true as intuition and unprovable as market share. The provable claim is stronger.

There is no unlisted path to the American market.

OSHA requires that thirty-eight categories of product used in the workplace be certified by a Nationally Recognized Testing Laboratory. Since 2015 the FCC has required most equipment certification applications to be filed through a Telecommunications Certification Body rather than with the Commission itself — meaning the TCB does not merely test but issues the legal authorization to market. Retailers impose listing requirements independently. Insurance underwriters impose them. Authorities having jurisdiction impose them at the point of installation. A manufacturer that declines certification does not face higher costs; it faces no route to the shelf.

Now count the exits. OSHA recognizes roughly twenty-one NRTLs. UL Solutions is the only American operator of scale. The remainder resolves to London, Toronto, Cologne, Munich, Geneva, Paris, Luxembourg and Oslo — Intertek, CSA Group, TÜV Rheinland, TÜV SÜD, DEKRA, SGS, Bureau Veritas, Eurofins, Nemko, QPS, QAI.

The United States did not outsource market-access adjudication to Britain. It outsourced it, essentially in full, to whoever would perform it — and Britain bought the largest position. That is the topology, and the topology is the vulnerability. Intertek is the largest foreign node, not the anomaly.

Its scale is nonetheless the appropriate object of attention: more than one thousand laboratories and offices across more than one hundred countries, some forty-five thousand employees, more than four hundred thousand clients, 2025 revenue of £3,431.6 million and first-half 2026 revenue of £1,771 million. Its consumer-facing mark, ETL, descends from Thomas Edison’s Lamp Testing Bureau of 1896 by way of the Electrical Testing Laboratories — an American testing apparatus absorbed into British corporate structure through the Inchcape holding company, itself descended from a Calcutta merchanting partnership of 1847 that carried British military post during the Indian Mutiny. The American consumer sees Edison’s mark and infers domestic governance. The governance has been London since 1988.

And as of this writing, it is ceasing to be London, in a manner that proves the argument better than the British framing ever could.

On 18 June 2026, Intertek’s board accepted a recommended final cash offer from Isotope Bidco Limited, a newly formed vehicle indirectly owned by EQT X EUR SCSp and EQT X USD SCSp, each acting through its manager EQT Fund Management S.à r.l. — Luxembourg. Minority co-investors include Luxinva S.A. and Platinum Ivy, vehicles associated with the Abu Dhabi Investment Authority, and ATIC Second International Investment Company LLC, associated with Mubadala. The terms are £60.00 in cash plus the retained 107.7 pence final dividend: £61.077 per share, an equity value near £9.5 billion and an enterprise value near £10.9 billion, at a premium of approximately sixty-two percent to the undisturbed price. EQT bid four times, beginning at £51.50 in April.

Shareholders approved on 6 August 2026 with 98.7 percent of votes cast. On completion — expected in the fourth quarter of 2026 or the first quarter of 2027, subject to court sanction and regulatory conditions — Intertek’s shares are cancelled from trading on the London Stock Exchange and the company is re-registered as a private company.

Consider precisely what is changing hands. The purchased entity holds FCC Telecommunications Certification Body authority — delegated United States federal regulatory power. It holds OSHA NRTL recognition. Through Acumen Security it holds accreditation inside the NIAP Common Criteria scheme, the evaluation program for products destined for United States national security systems, managed and operated by the National Security Agency. Through EWA-Canada it holds accreditation under the Cryptographic Module Validation Program, run jointly by NIST and Canada’s Communications Security Establishment. Through Intertek NTA it holds NCSC CHECK status at the highest accreditation level, in a scheme operated by a body that is by public declaration part of GCHQ, with staff cleared to United Kingdom SECRET.

That portfolio of delegated sovereign authorities was priced by a private equity auction, sold to a Luxembourg fund structure with Gulf sovereign wealth in the syndicate, and will shortly stop producing public financial disclosure altogether. No American authority set the price, and it is not publicly established that any American authority reviewed the transaction.

Whether a CFIUS filing was made — over Intertek Testing Services NA, over the Houston headquarters entity, over Acumen Security in particular, which is a plausible candidate for treatment as a critical-technology United States business — is not visible in the reporting available. The scheme is expressly conditional on regulatory clearances including foreign-investment approvals across multiple jurisdictions, and the answer would be found in the conditions schedule of the Scheme Document. This is the single most consequential unresolved question in this document, and it is answerable from a public filing.

Everything above could be dismissed as structural anxiety were it not for what the Federal Communications Commission has done in the last fifteen months, in its own voice, on its own reasoning.

On 22 May 2025 the Commission adopted the “Bad Labs” Order, prohibiting FCC recognition of any TCB, test laboratory, or laboratory accreditation body owned by, controlled by, or subject to the direction of a prohibited entity — and prohibiting such bodies from participating in the equipment authorization program at all, including through the Supplier’s Declaration of Conformity route. The prohibition applies regardless of where the laboratory is located. The rules took effect 8 September 2025. Enforcement was immediate: the first withdrawals landed that month, and by April 2026 the Commission had withdrawn or denied recognition to twenty-three laboratories.

The reasoning is the part that matters. The Commission’s finding is that laboratories obtain, in the ordinary course of certification, unrestricted access to device specifications, circuit designs, software code, and operational parameters — and that such access constitutes an intelligence surface in itself, independent of any tampering. The Commission further found that a number of recognized laboratories had deep ties to Chinese state-owned enterprises, to military-civil fusion structures, and in some cases were state actors outright, and that these laboratories had tested thousands of devices bound for the United States.

Then the Commission went further. On 29 January 2026 it adopted attestation and disclosure requirements for licensees owned by, controlled by, or subject to the jurisdiction of a foreign adversary. On 23 March 2026 it updated the Covered List to include all foreign-made routers. On 30 April 2026 it adopted a Second Report and Order and proposed, in an accompanying notice, to prohibit recognition of laboratories, TCBs and accreditation bodies located in, or conducting testing and accreditation in, countries that are not Reciprocal Economies — a jurisdictional rather than ownership test.

And in the further notice accompanying the original order, at paragraph 146, the Commission asked whether the structure of its own rules — which incorporate ISO/IEC 17025 and ISO/IEC 17065 as the standards against which accreditation bodies assess laboratories and certification bodies — raises questions about the integrity of the equipment authorization program or the impartiality of TCBs and test labs.

Read that again, because it is the hinge of this entire column. The United States government has formally found that conformity assessment is a national security function; that the ownership, control and jurisdiction of the certifier are material to that function; that the certification process itself transfers sensitive design information to the certifier; and that the international accreditation framework underneath the whole arrangement may be structurally inadequate to the security question.

The premise is jurisdiction-neutral. It contains no term restricting it to any particular country. It has been applied to one.

This document does not argue that London is Beijing. It argues something narrower and much harder to answer: America wrote the test, found the principle sound, and has declined to run it on anyone else — while the largest non-American holder of its delegated certification authority is sold to a Luxembourg fund and taken private.

The final piece of the machine column is the mechanism by which foreign technical specifications become American device architecture, and it is the least conspiratorial and most inexorable process in this entire document.

The instruments are public and dated. Article 3.3(d), (e) and (f) of the Radio Equipment Directive, activated by Delegated Regulation (EU) 2022/30, made cybersecurity a condition of market access for connected radio equipment in the European Union, with the EN 18031-1, -2 and -3 series becoming mandatory on 1 August 2025. The United Kingdom’s Product Security and Telecommunications Infrastructure Act runs in parallel and converges on the same technical baselines, ETSI EN 303 645 among them.

Above these sits the Cyber Resilience Act, Regulation (EU) 2024/2847, in force since 10 December 2024. Its framework for the notification of conformity assessment bodies applies from 11 June 2026. Mandatory reporting of actively exploited vulnerabilities and severe incidents to ENISA — within twenty-four hours, with follow-up at seventy-two hours and fourteen days — applies from 11 September 2026, and applies to products already on the market. Full product obligations apply from 11 December 2027, with penalties calculated against worldwide turnover and consequences extending to recall and loss of market access. Member States are directed to ensure sufficient notified body capacity by 11 December 2026.

CEN, CENELEC and ETSI accepted a standardization request on 3 April 2025 to develop forty-one harmonized standards — fifteen horizontal, twenty-five vertical — in support of the Act. As of mid-2026, none had been published in the Official Journal. The practical consequence is perverse and important: without published harmonized standards, manufacturers cannot claim presumption of conformity and must instead engage notified bodies for individual assessment. A standards vacuum has been converted into a durable expansion of discretionary authority for precisely the certification firms under discussion, at precisely the moment one of the largest is being taken private.

Now the mechanism itself. No manufacturer of scale builds two versions of a product. The tooling, the firmware branch, the certification cost, the inventory complexity and the warranty exposure make architectural bifurcation irrational at any volume. The architecture required to sell in Europe therefore becomes the architecture shipped into American homes, hospitals, classrooms, utility substations and military installations — not by treaty, not by capture, not by anyone’s consent, and not by any decision an American voter or legislator was ever offered.

Submission to transatlantic standards was never signed. It was amortized.

The same machinery is now moving to the cognitive layer. ISO/IEC 42001, the first international management system standard for artificial intelligence, is delivered by the same certification firms under the same accreditation architecture, timed against the European AI Act’s high-risk obligations arriving in August 2026. As of March 2026 the ITU reported 203 AI-related ITU-T standards in force with 199 more under development, coordinated through the AI Standards Exchange under the World Standards Cooperation, reinforced by the Global Digital Compact and WTSA-24 Resolution 101, with ISO, IEC and ITU adopting a joint statement in Seoul in December 2025. That layer deserves its own treatment and will receive it elsewhere. It is noted here only to establish the trajectory: the same delegation, the same accreditation chain, and the same absence of American authorship, migrating from the device to the intelligence that will govern the device.

Now the reframe on which the entire second column turns, and the reason a policymaker should read this section twice.

The standard objection to everything above is that the capabilities in question are legitimate. Remote management subsystems exist because enterprise fleets must be administered. Certified surveillance equipment exists because buildings must be secured, borders monitored, infrastructure protected, and crimes solved. TÜV Rheinland and its peers certify enormous quantities of entirely lawful security apparatus for entirely lawful purposes. All of this is true, and none of it is a rebuttal.

Because the correct analytical object is neither intent nor capability. It is the elasticity of authorization.

A subsystem that reads all system memory, communicates independently of the operating system, and persists through power-off is a fleet-management tool when the operator is an enterprise IT department and a collection surface when the operator is a service holding legal process over it. Nothing physical distinguishes the two cases. The distinguishing variable is a legal relationship — who may compel, under what authority, with what notification, subject to what review. And legal relationships are not physical constants. They are the most volatile parameters in the entire system.

Under conditions of alliance the distinction is academic, which is exactly why the installed base was permitted to accumulate without scrutiny. The conformity regime that admitted it was constructed during a period in which jurisdictional co-location with allied states was read as a safety feature rather than an exposure — and the reading was, at the time, defensible.

Under conditions of competition, the same installed base changes character with zero physical modification and zero marginal cost. No device is touched. No factory is compromised. No shipment is intercepted. No conspiracy is required at any point, past or present, because the architecture that would be exploited is the architecture that was lawfully certified, lawfully sold, and lawfully installed.

That is what a quiet struggle looks like between parties who remain formally allied and materially rivalrous — a contest over the completion of independence conducted entirely inside instruments that no treaty covers and no defense budget addresses.

And this document does not need to allege that any such exercise has occurred or will occur. It is sufficient, and considerably more powerful, to establish three facts that are not in dispute:

The capability is pre-positioned. It is present, by design and by disclosure, across essentially the entire American computing base.

Its activation would be unilateral. It requires no American participation, no American authorization, and no notification to any American authority.

It is not reciprocated. The American conformity stack does not occupy the equivalent position inside anyone else’s homes, hospitals and installations, because America did not build one.

To which must be added the finding from the previous section: its exercise would be undetectable by any certification process now in force, because the certification process examines declared parameters and the relevant capabilities are below the declaration threshold — and because the United States government has already recognized this exact chain of reasoning, in writing, when the counterparty was China.

The money and the machine appear unrelated until the mechanism from Section I is applied to both.

In each case, the United States delegated an adjudication function — a function that determines legal existence rather than merely regulating conduct. Whether a security is counted as owned. Whether a device may lawfully be sold. In each case the delegation ran to parties outside American jurisdiction. In each case the delegation was justified as technical, efficient and apolitical. In each case the arrangement was invisible to the electorate, cheap for the holder, and entirely legal. And in each case no military instrument can address it, because there is nothing to strike: the exposure is a registry entry, an accreditation schedule, a technical scope, and a discontinued statistical category.

That is the answer to the question this document is named for. Who needs to leave behind a garrison? Nobody does. A garrison is expensive, conspicuous, politically radioactive, and vulnerable to precisely the kind of force the United States is optimized to project. It must be manned, supplied, defended and eventually withdrawn, and its withdrawal is photographed. A registry costs nothing, provokes nothing, is never photographed, and cannot be liberated by a landing.

Empire did not withdraw from the Americas. It changed its mounting.

Because the argument’s strength depends on refusing to overstate, every material claim is tiered.

Established. The December 2025 Treasury holdings figures and the discontinuation of the Caribbean Banking Centers and expanded United Kingdom aggregates after February 2016, all from Treasury’s published table. The Federal Reserve’s finding of a $1.4 trillion undercount and the absorption figures of $32 billion foreign official against $1,225 billion Cayman hedge fund across 2022–2024. The leverage, haircut and repo aggregates as published by Board staff, the Dallas Fed, the OFR and the IMF. The failure of Section 51 of the 2018 Act to produce an Order in Council, the Cayman legitimate-interest register and its access fee, and the FATF increased monitoring of the British Virgin Islands. The Corporate Tax Haven Index rankings and the thirty-three percent network share. Bermuda’s reinsurance market share and loss exposure. The OSHA NRTL roster and its national composition. The FCC Bad Labs Order, its effective date, the twenty-three laboratory actions, the January, March and April 2026 measures, and the Commission’s stated reasoning including its question about ISO/IEC 17025 and 17065. The RED delegated regulation, EN 18031 mandatory date, and the Cyber Resilience Act timetable. The terms, ownership structure, shareholder approval and delisting consequence of the EQT acquisition of Intertek. The documented architecture of Intel CSME and AMD PSP as published by their manufacturers.

Strongly indicated. That the British-jurisdiction share of foreign-held United States sovereign debt exceeds a quarter once the Cayman adjustment is applied and the Crown Dependencies and British Virgin Islands are accounted for — the direction is certain, the precise figure depends on residual composition. That the absence of published harmonized standards under the Cyber Resilience Act materially expands notified body discretion. That architectural bifurcation between European and American product variants is economically irrational at volume and therefore rare.

Plausible, unverified. Reports that a large majority of electronics testing for United States–bound devices has been performed in Chinese-controlled laboratories. Estimates placing the Cayman adjustment nearer $1.85 trillion than $1.4 trillion.

Unresolved. Whether any CFIUS filing was made or review conducted in connection with the acquisition of Intertek Group plc by Isotope Bidco Limited, and what the regulatory conditions schedule of the Scheme Document specifies. Whether the FCC intends its Reciprocal Economies proposal to reach beyond the People’s Republic of China in practice. Whether the retirement of the Treasury aggregates in February 2016 followed any deliberation, and where that deliberation is recorded.

Explicitly rejected. That FCC equipment authorization verifies lawful-intercept capability in consumer devices. That certification against EN 303 645 connects to ETSI’s lawful interception specifications. That the Cayman Treasury position is beneficially owned by British nationals or directed from London. That any of the arrangements described required coordination, conspiracy, or hostile intent by any named party.

The forward-leaning conclusion is available because none of this was seized. It was vacated — and vacated positions can be reoccupied by the party that vacated them.

The cheapest measure costs nothing at all. Treasury can restore the discontinued aggregates. Republishing British-jurisdiction and offshore-center totals alongside the country lines would render visible, in a single number on an existing page, a concentration that has been legible only to those willing to perform arithmetic the table discourages. A state that cannot see itself cannot govern itself, and this particular blindness is a formatting decision.

The second measure requires only consistency. Extend the Bad Labs principle jurisdiction-neutrally. The Commission has already found that the ownership, control and domicile of certification bodies are national security variables and that the accreditation architecture beneath them may be inadequate. Applying a standard the United States authored to every foreign holder of delegated American certification authority — with graduated treatment reflecting genuine differences among partners rather than a binary — is not escalation. It is the elimination of an exception that has no principled basis.

The third measure is jurisdictional. Bring holders of delegated American regulatory authority explicitly within foreign-investment review. An entity that issues legal authorizations to market on behalf of the FCC, certifies to OSHA-recognized scopes, and operates evaluation laboratories inside NSA-managed and NIST-administered programs is not an ordinary commercial acquisition target, and the question of whether its transfer to a foreign fund structure is reviewable should not have to be inferred from a scheme document.

The fourth is industrial. Rebuild domestic conformity assessment capacity as a strategic sector, on the same logic that produced the semiconductor and critical minerals programs. One American operator of scale in a market of twenty-one is a single point of failure, and the market will not correct it, because certification revenue accrues to incumbents and incumbency here is an accreditation asset rather than a technological one.

The fifth is already underway and should be finished. Central clearing of cash Treasuries in late 2026 and of Treasury repo in mid-2027, together with meaningful haircut floors of the kind the Financial Stability Board has recommended, converts an opaque bilateral leverage stack into an observable, margined one — which is the difference between a hostage and a counterparty.

And the sixth is the largest. Author the conformity grammar rather than consuming it. The elements exist: the Center for AI Standards and Innovation at NIST, the American AI Exports Program implemented on 21 October 2025 around full-stack export packages, and the executive actions of December 2025 and June 2026 treating advanced systems as national security infrastructure. The strategic logic is identical to the semiconductor case and the cable case. A civilization that exports its technology stack but imports the standards by which that stack is judged has won the market and lost the adjudication, and adjudication is the layer that compounds.

Which returns the argument, as everything in this sequence does, to 1823 and to John Quincy Adams.

Canning offered a joint declaration. It would have been costless, immediately advantageous, and materially indistinguishable in its effects, since the Royal Navy was going to hold the water either way. Adams refused — not because Britain was wrong about recolonization, but because a co-signature is a co-authorship, and co-authorship, once conceded, does not expire. He accepted the substrate dependency he could not yet escape and refused to ratify it into a permanent legal fact.

Two centuries later the objects have changed completely and the question has not moved an inch. Who counts what is owned. Who decides what may be sold. Who writes the scope of examination. Who authors the grammar by which legitimacy is conferred.

The United States has an army that no power on earth can face, a navy that has held every ocean for eighty years, and an economy that finances the world. And it cannot see who holds a quarter of its foreign-held sovereign debt, cannot supervise the leveraged position that absorbs a third of its net issuance, cannot compel the beneficial ownership records of the jurisdiction where that position sits, does not own the gate through which its own citizens’ devices enter their homes, and does not write the standard against which those devices are judged.

No garrison ever had to be left behind. That is the entire point. The war with empire was never a war of positions, and the independence declared in 1776 was never finished — because finishing it was never a matter of removing anyone from the ground. It is a matter of taking back the ledgers, the registries, the scopes, and the right to say what is real.

That work has not begun. It could begin this fiscal year, and most of it is free.

A reader who reaches this point will ask the obvious question, and it deserves an answer rather than a silence.

Did anyone know? And beyond that: was the murky trans-jurisdictional system built — deliberately, by someone, for a purpose — rather than merely accreted?

The first question has a documented answer, and the answer is yes. Nobody was fooled. Hamilton knew: by 1795, roughly forty percent of United States government securities were held by foreign investors, principally British and Dutch houses, as a direct and intended consequence of assumption — the argument developed at length in How Hamilton Became America’s Most Sophisticated Cultural Trojan Horse. Anchoring European capital to American fiscal capacity was not a leak in the design. It was the design.

The second question also has a documented answer, and it is the more consequential one — because the answer is yes, the system was built, and it was not built by Washington.

Begin with what the United States did, because it is the half most often mistaken for the whole.

Regulation Q capped what American banks could pay depositors while market rates climbed past the ceiling. The Interest Equalization Tax, effective on purchases after 18 July 1963, taxed American acquisition of foreign securities to defend the balance of payments. The Voluntary Foreign Credit Restraint program followed in 1965; the Foreign Direct Investment Program became mandatory in 1968, which contemporaneous analysis identified as the market’s real take-off point. The Securities and Exchange Commission’s own historical society records the outcome without euphemism: these controls essentially halted the internationalization of the American securities markets, closed American capital markets as a practical matter to Europeans seeking dollars, and made London once again the center of international finance.

Then comes the fact that disposes of any reading in which Washington laid the trap. On 3 December 1981, the Federal Reserve Board amended Regulations D and Q to authorize International Banking Facilities — a segregated onshore booking regime free of reserve requirements, interest ceilings and deposit insurance assessments, with nine states adding tax relief. The stated purpose, in the contemporaneous legal literature, was expressly to repatriate international banking business then being conducted in offshore financial centers: to let American banks and their customers reach the Euromarket inside American political and legal jurisdiction, in an American time zone.

A state does not spend a decade constructing an escape route and then build a formal mechanism to close it. Washington created the pressure that pushed the capital out, recognized the loss, attempted a correction in 1981, and recovered a slice of wholesale bank booking while the entity layer stayed gone. Forty-five years later the marginal financier of the United States government is a fund registered in George Town.

Now the reversal, and the reason this note exists rather than a shrug.

The terrain those American controls pushed capital onto was not neutral ground that happened to be lying there. It was constructed, in British sovereign space, by a British regulator, through a deliberate decision not to regulate — and the decision is documented.

In June 1955, Bank of England staff noticed that Midland Bank was taking United States dollar deposits unconnected to any commercial transaction and paying interest above what American regulation permitted. The Bank looked at it, considered it, and concluded internally that it would be wise not to press Midland any further. In 1957, defending sterling after Suez, Britain raised rates and restricted overseas lending in sterling — and London’s banks simply moved their international lending from sterling into dollars, into the market Midland had opened.

Then came the constitutive move. The Bank of England declared that Euromarket books, though kept in London by banks in London, were not in London for regulatory purposes — they were elsewhere. That word is the Bank’s own, and it is the entire doctrine. Where elsewhere was, the Bank did not specify, because it did not matter: the banks kept two sets of accounts, and one set was held to describe activity occurring in a place that was never named. What the elsewhere designation produced was therefore not a transfer of jurisdiction from one authority to another but a void — a legal space in which activity is deemed to occur where it does not occur, and consequently falls under no supervisor at all. What mattered was the consequence: no regulatory authority on earth had jurisdiction over the market. Not the Bank of England, which had disclaimed it. Not the Federal Reserve, which could not reach it. A legal space was created in which activity occurring in one jurisdiction is deemed to occur nowhere, and it was created on purpose, by an institution acting within its authority, in the decade in which territorial empire was being wound down.

That is the point worth stating plainly, because it is the answer to the question this note began with. The rules of engagement in that arena are not the rules of American finance, and that is not an accident of drift — it is the product specification. No disclosure regime. No reserve requirement. No lender of last resort. No supervisor with standing. An arena in which a contest can run indefinitely without any referee possessing jurisdiction is not a loophole in a market. It is a purpose-built venue, and the party that builds the venue writes the rules of the venue, collects the rent, and never has to win a battle inside it.

The venue was then extended outward through exactly the territories decolonization was otherwise emptying — the Crown Dependencies as the inner ring, the Overseas Territories beyond them, each offering what the metropole could not offer at home: tax neutrality, flexible trust and company law, confidentiality, and, decisively, English common law with Privy Council appeal, which is the actual selling point. Tax Justice Network’s finding that this network carries roughly a third of global corporate tax abuse risk while the United Kingdom itself ranks well down the same index is that asymmetry rendered as a number. So is the Bank’s own recorded position on the satellites: no objection to their providing boltholes for non-residents, provided United Kingdom capital did not thereby escape United Kingdom rules. Other nations’ capital was welcome to leave other nations’ jurisdiction. British capital was not.

Eurodollar lending, negligible in 1958, reached roughly twenty-five billion dollars by 1968 and one hundred thirty billion by 1973, with London holding near eighty percent; five hundred billion by 1980; and by 1998 the substantial majority of international lending was being transacted offshore. The City’s recovery from post-Suez decline into the position it still occupies is directly traceable to this sequence. That is what the record shows, and it is why “we drew them into a swamp” cannot be sustained: the swamp was theirs, they dug it, and it restored them.

Three qualifications keep this from becoming a theory of omnipotence, and each is established elsewhere in this document.

The builder lost control of the product. Section II sets out the record: eight years after Section 51 of the Sanctions and Anti-Money Laundering Act 2018 required a draft Order in Council compelling public beneficial ownership registers, no Order has been made; Cayman has conceded a paywalled register with a discretionary access test; the British Virgin Islands sits under increased FATF monitoring; and the Territories now invoke European privacy jurisprudence as a constitutional shield against Westminster itself. A venue whose builder can no longer set its rules is no longer an instrument of the builder. It has become weather.

The tenant got the larger prize. The Euromarket made dollars available outside American regulation, which is precisely what made the dollar unassailable — it could be held, borrowed and settled without submitting to Washington. That was not designed by anyone, but it was recognized and then defended, and it is among the deepest reasons the dollar survived the collapse of Bretton Woods. Britain hosted the mechanism that entrenched American monetary primacy and took the fee income. This is a landlord relationship in which each party mistakes itself for the one holding leverage — a structure more unstable than either party being genuinely in control, and structurally identical to the certification column above.

Design at the origin does not imply direction at the destination. That British institutions built the arena in the 1950s and 1960s is established. That American capital entered it to escape American regulation is established. That anyone, in London or New York, is steering the $2.79 trillion now inside it is supported by nothing in the record, and Section II argues affirmatively against it.

One speculative reading survives scrutiny, and it concerns Bermuda rather than Cayman: the premium is written overwhelmingly by United States registrants, the risk is American hurricane and earthquake, and the constitutional counterparty in a catastrophic failure is a Crown-appointed Governor and the Privy Council rather than a state legislature. Whether anyone intended to lodge American catastrophe exposure beneath a British constitutional umbrella, the effect is that it sits there — and this document establishes that the umbrella could not be opened even if asked.

Which returns to the only claim the argument actually requires. Designed at the origin, entered by choice, and ungoverned at the destination — it is measurable, and it is currently not measured. That is why the cheapest recovery listed above is also the first: restore the aggregates Treasury retired in February 2016. A state that cannot see itself cannot determine whether it was outmaneuvered, merely exposed, or both, and it is not entitled to the flattering answer until it looks.

Bryant McGill is a Wall Street Journal and USA Today Best-Selling Author, founder of Simple Reminders, and architect of the Polyphonic Cognitive Ecosystem. A Congressionally Recognized Ambassador of Goodwill and United Nations appointed Global Champion, his work spans naval intelligence systems, computational linguistics, and civilizational governance architecture.

Who Needs to Leave Behind a Garrison (August 2026). This document. The two columns of American jurisdictional subservience — sovereign debt domiciled beyond supervisory reach, and the conformity-assessment gate that determines what may lawfully exist — read as a single delegated-adjudication failure. Read it for the arithmetic a policymaker cannot dismiss.

Indicators That the American War With Empire Never Ended (2026). The evidentiary register underlying the entire sequence — documented instruments and analytic propositions tiered across force, coercion, doctrine, institutional secession, substrate repatriation, the money firewall, and the cognitive war. Read it for the raw record from which this document’s two columns are drawn.

The Fifth State: Why We Are at War and What the War Is Actually For (April 2026). The five-civilizational taxonomy and the conversion mechanism through which the Fourth State is authored into the Fifth. The carrying grammar for the entire geopolitical corpus.

The Monroe Doctrine, the British Substrate, and the Unfinished Decolonization of the Americas (August 2026). The doctrinal companion to this document: the Trump Corollary read not as revival but as completion, and the hemispheric residue relocated from garrison to clearing layer.

From Telegraph to Waterworth: The Cable War the UK Already Lost (April 2026). The physical layer beneath both columns — Victorian submarine infrastructure, its regulatory chokepoints, and the American rerouting now underway.

The British Are Coming. Again? Not by Sea, but by Standard. (April 2026). The standards war beneath the cable war, and the argument this document extends into conformity assessment and market-access law.

Intertek and the Future of AI-Mediated Surveillance Distribution (April 2026). The certification chokepoint in detail, including the imperial genealogy from Caleb Brett and Edison’s Lamp Testing Bureau through Inchcape. The CALEA and TC LI chains asserted there are corrected in Section III above.

Prestige Networks: Transatlantic Blame from the Civil War to Modern America (January 2026). The centuries-long pipeline through which legitimacy has been rationed, and the human layer of the adjudication problem.

How Hamilton Became America’s Most Sophisticated Cultural Trojan Horse (June 2025). Debt assumption as the foundational instrument of financial continuity after political separation — the eighteenth-century ancestor of Column One.

Manufacturing Sovereignty (Abridged) (June 2025). The industrial base as sovereignty precondition, and reshoring as substrate repatriation rather than protectionism.

Allies Are Not Friends (January 2026). Alliance as utility-bounded rather than ontological — the analytic vocabulary required by Section IV.

Why U.S. Business May Tilt Toward Russia Over the UK (April 2026). Strategic realignment under Fifth State logic, and why present transatlantic friction is structural rather than personal — the geopolitical premise underlying Section IV.

Manufacturing Sovereignty — European Edition (June 2025). The European disposition of the same substrate argument, and the standards environment into which American manufacturers now build.

The Spider’s Web: Britain’s Second Empire (Michael Oswald and John Christensen, 2017). The documentary account of the second empire thesis: London’s transformation into an offshore jurisdiction from the mid-1950s, the extension of the arrangement through the Crown Dependencies and Overseas Territories during decolonization, and Christensen’s first-hand account of Jersey’s conversion into a satellite of the City. Read it for the architecture of Column One in its historical form.

The Spider’s Web — full transcript. The elsewhere designation stated directly: banks kept two sets of accounts, and the Bank of England declared that the London Euromarket accounts were not in London but elsewhere, and therefore not its responsibility to regulate — a legal space in which activity is deemed to occur where it does not occur, and consequently under no regulator’s jurisdiction at all.

Film review: The Spider’s Web (New Thinking for the British Economy). Includes the Bank of England’s recorded position on the satellite jurisdictions — no objection to their providing boltholes for non-residents, provided the opportunity was not thereby created for United Kingdom capital to leave United Kingdom rules — and the growth series from the 1960s to the near-total offshore share of international lending by 1998.

The Men Who Stole the World: An Interview with Nicholas Shaxson (openDemocracy). Shaxson on London becoming an offshore jurisdiction from the mid-1950s — precisely the moment of decolonization — through a deregulated space created by designation rather than by legislation. The author of Treasure Islands, from which the Midland Bank episode of June 1955 and the Bank’s decision not to press the matter further are drawn.

The Imperial SEC? Foreign Policy and the Internationalization of the Securities Markets, 1934–1990 (Securities and Exchange Commission Historical Society). The Interest Equalization Tax, the 1968 controls, and the Commission’s own account of how American capital controls drove the market to London.

The Imperial SEC? Foreign Policy and the Internationalization of the Securities Markets, 1934–1990 (Securities and Exchange Commission Historical Society). The Interest Equalization Tax, the 1968 controls, and the Commission’s own account of how American capital controls drove the market to London.

Concept of Offshore Financial Centers (IMF Working Paper 07/87). The Interest Equalization Tax, Voluntary Foreign Credit Restraint and Foreign Direct Investment Program as the regulatory drivers of Eurocurrency growth after 1966.

The Internationalization of the Securities Markets (Bodolus, 1974). Contemporaneous analysis identifying the mandatory 1968 foreign direct investment controls as the market’s real take-off point.

International Banking Facilities (Key and Terrell, Federal Reserve Board, September 1988). The history of the IBF proposal, the December 1981 amendments to Regulations D and Q, and the state tax treatment that accompanied them.

Banking (University of Miami Inter-American Law Review, 1981). Contemporaneous statement of the IBF regime’s express purpose: to repatriate international banking business then conducted in offshore financial centers.

The Cross-Border Trail of the Treasury Basis Trade (FEDS Notes, 15 October 2025). Barth, Beltran, Hoops, Kahn, Liu and Perozek. The $1.4 trillion TIC undercount, the absorption decomposition, and the marginal-buyer finding.

Major Foreign Holders of Treasury Securities (U.S. Department of the Treasury / Federal Reserve Board). The primary series, including footnotes 2 and 4 recording the aggregates discontinued after February 2016.

Decomposing Hedge Funds’ U.S. Treasury Exposures (FEDS Notes, 22 June 2026). Basis trade volume estimates and the financial stability framing.

Measuring Cross-Border Securities Positions (FEDS Notes, 13 February 2026). Judson and Kim on custodial bias and the systematic asymmetries between TIC and IMF portfolio data.

Hedge Fund Treasury Exposures, Repo, and Margining (FEDS Notes, September 2023). Banegas and Monin on effective leverage and the arithmetic of a haircut floor.

Rising Hedge Fund Leverage Affects Monetary Policy Implementation (Federal Reserve Bank of Dallas, 28 May 2026). Net repo borrowing at approximately $1.8 trillion and the spread effects on policy transmission.

Treasury Market Disruptions and Policy Options (Congressional Research Service). Repo haircut policy options and the Financial Stability Board recommendation on non-bank leverage.

Foreign Holdings of Federal Debt (Congressional Research Service). The country-level series and the standard framing this document argues has been superseded.

Public Registers of Beneficial Ownership in the Overseas Territories and Crown Dependencies (House of Commons Library). Parliamentary supremacy, the Order in Council mechanism, and the implementation record.

Beneficial Ownership Registers: Overseas Territories (Hansard, 3 July 2025). The ministerial statement recording territory-by-territory slippage and the FATF action against the British Virgin Islands.

Beneficial Ownership Transparency in the Cayman Islands (Oxford Business Law Blog, April 2026). The access fee, the legitimate-interest test, and the Sovim judgment as constitutional shield.

Corporate Tax Haven Index (Tax Justice Network). Current rankings and the United Kingdom network’s share of global corporate tax abuse risk.

ABIR Fact Sheet 2026 (Association of Bermuda Insurers and Reinsurers). Global reinsurance market share, gross loss exposure, claims paid, and the SEC-registrant composition of gross premium written.

FCC Bans “Bad Labs” From U.S. Equipment Authorization Process (Federal Communications Commission, Report and Order and FNPRM, adopted 22 May 2025). The Commission’s findings on certification-body ownership, laboratory access to design information, and program integrity.

Promoting the Integrity and Security of Telecommunications Certification Bodies, Measurement Facilities, and the Equipment Authorization Program (Federal Register, 7 August 2025). The codified rule text and the 8 September 2025 effective date.

FCC Acts in Campaign Against ‘Bad’ Equipment Test Labs (Cooley, November 2025). Early enforcement actions and the prohibited-entity designation architecture.

FCC Targets Equipment Testing and Telecom Licenses (Hogan Lovells, May 2026). The Second Report and Order, the Reciprocal Economies proposal, the January 2026 attestation rules, and the March 2026 Covered List expansion to foreign-made routers.

OSHA’s Nationally Recognized Testing Laboratory Program (Occupational Safety and Health Administration). The recognition framework, the scope-of-standards structure, and the current roster.

Recommended Final Cash Acquisition of Intertek Group plc (EQT, 18 June 2026). Terms, structure, and fund vehicles.

Offer for Intertek (Intertek Group plc). The bidder identification, including EQT X EUR SCSp, EQT X USD SCSp, Luxinva S.A. and ATIC Second International Investment Company LLC, and the scheme documentation.

Intertek 2025 Full Year Results Announcement (3 March 2026). Revenue, margin, laboratory network scale, and 2026 acquisition activity.

The Cyber Resilience Act: Summary of the Legislative Text (European Commission). Entry into force, the notified body framework from 11 June 2026, reporting obligations from 11 September 2026, and full application from 11 December 2027.

EU Cyber Resilience Act: Key 2026 Milestones (Hogan Lovells, January 2026). The staged compliance calendar and the early-applying building blocks.

The Cyber Resilience Act: Why Manufacturers Must Act Now (UL Solutions). The 3 April 2025 standardization request and the forty-one harmonized standards mandate.

EN 18031 and the EU Cyber Resilience Act (2026). Delegated Regulation (EU) 2022/30, the 1 August 2025 mandatory date, and the parallel operation of the RED and CRA regimes.

Intel CSME Security White Paper (Intel Corporation). The manufacturer’s own description of the autonomous subsystem, its independent processor and network capability, and the two-stage fuse provisioning model.

AI Standards Exchange Database (International Telecommunication Union). The World Standards Cooperation gateway, WTSA-24 Resolution 101, and the coordinated AI standardization program.

Winning the Race: America’s AI Action Plan (The White House, 23 July 2025, and subsequent executive actions). The Center for AI Standards and Innovation, the full-stack export framing, and the December 2025 and June 2026 directives.

The Department of Commerce Announces American AI Exports Program Implementation (International Trade Administration, 21 October 2025). The export package architecture and its standards implications.