Ratified
by
Expiry.

This volume is a list. Everything else in it is scaffolding for the list.

There is a way of telling this story that requires a conspiracy, and it is the wrong way, because it is both unprovable and unnecessary. Nobody had to meet. What happened instead is duller and much harder to reverse: a relay, run in public, over fifty-four years, by named people holding dated pens.

Each runner does roughly the same three things. Opens a door that was previously subject to review. Announces that this is the last door, or the necessary one, or the end of a trend. Hands the file to a successor who opens the next one and says the same thing.

Nobody in the chain is doing anything secret. Most of them held a press conference about it. The transfer is not in any single handoff; it is in the fact that the handoffs kept going, in the same direction, through changes of party, changes of government and changes of continent, and that almost none of them were ever put to a vote.

So here is the baton, from Nixon to a Queensland premier in Washington last week. Forty-five entries. Every one of them dated, public, and attributable to a person who could have declined and did not.

What was handed over
was never the trade.
It was the review.

Go back to the third row of the list, because everything downstream of it is shaped by it.

When Carter restored China's most-favoured-nation tariff status in 1980, he did not restore it permanently. Under Jackson–Vanik it came up for renewal every single year, and every year Congress held a debate and a vote on whether to grant it. Tiananmen was argued in that vote. Prison labour was argued in that vote. Currency policy, proliferation and religious persecution were argued in that vote.

It was, in practice, the only recurring leverage the United States held, and its recurrence was the whole of its value. Clinton drew its teeth in 1994 by delinking human rights from it. Six years later he removed the tooth socket.

PNTR did not open a door that was closed. It removed the hinge, so the door could not be closed again.

An asset can be bought back. A mine can be re-permitted, a port lease can be bought out at a price, a company can be renationalised. Britain bought CGN out of Sizewell. Australia voided Victoria's memorandum. Italy let its own lapse. Those were expensive and slow and they were all possible.

What is not recoverable at any price is a standing occasion to say no — the annual vote, the sitting-day debate, the review with a live decision at the end of it. And the list above is, from 1980 onward, mostly a record of those being retired one at a time: abolished in Washington in 2000, allowed to expire in Ottawa in 2012, and allowed to expire again in Ottawa in 2026 on a five-and-a-half-billion-dollar mining bid.

Which brings the relay to the one runner who wrote a date on it.

The text is public.
The case is not.

Now to the corrected version of “hidden contracts.” The contract is not hidden. What is hidden is what happens when somebody uses it.

Article 28 governs public access to the arbitration. One sentence of it is generous and the rest of it is not:

• The award must be published, subject to redaction of confidential information.
• Every other document submitted to or issued by the tribunal is public only if the disputing Contracting Party determines it to be in the public interest.
• Hearings may be opened to the public on the same discretionary basis.

Read the second bullet with the parties named. When a Chinese investor brings a claim against Canada, the disputing Contracting Party is Canada. The government being sued decides whether the country it governs is allowed to see the case against it.

That is a defensible drafting choice in a vacuum. It is harder to defend against Canada's own subsequent practice. Two years after FIPA came into force, Canada agreed in CETA that the UNCITRAL Rules on Transparency would govern investor–state proceedings, that hearings shall be open to the public, and that a broader class of documents than the UNCITRAL rules themselves require would be released. Transparency there is mandatory. Under FIPA it is optional, and the option belongs to the party with the most to lose from exercising it.

Canada knew how to write an open arbitration clause. In this one agreement, with this one counterparty, it wrote a closed one.

Eight years after entry into force, Canada did the thing the treaty was written to make expensive. The three divestiture orders of 2 November 2022 hit completed investments held by Chinese investors — on any ordinary reading, covered investments under a treaty in force, and forced divestiture is the fact pattern investor–state arbitration exists to address. If FIPA has teeth, that was the bite.

No claim has been filed. Not in 2022, not since. No published award, no registered case, no tribunal constituted. That is a real point against the alarmed reading and it deserves saying without qualification.

It also has to sit next to the shape of Article 28. Only an award must be published, so a claim settled quietly or withdrawn after consultations produces no document anyone is obliged to release. That is a structural gap, not evidence that anything fell through it. The honest sentence is narrower than either side would like: we can confirm no award, and we cannot confirm no claim. The drafting is what puts those two in different categories.

The policy turned.
The treaty never
had to move.

On 16 January 2026, Prime Minister Carney stood beside Xi Jinping in Beijing and announced a new strategic partnership. The Prime Minister's Office released a Canada–China Economic and Trade Cooperation Roadmap the same day, with energy and minerals named among the areas of cooperation.

Thirteen years and forty days after “the end of a trend.” No legislation was required to get there, because none had been required to establish the restriction in the first place. That is what it means for a constraint to be policy.

Then came the test case, and it went the way the mechanism predicts.

Zijin Gold International, a subsidiary of China's largest gold miner, agreed to acquire Toronto-listed Allied Gold for C$5.5 billion — C$44 a share in cash, for a company running about 375,000 ounces a year out of Mali and Côte d'Ivoire from a head office of roughly sixty-five people in Toronto. It would have been the largest Chinese acquisition of a Canadian-listed miner since Nexen.

Ottawa's national-security review of that transaction concluded in early April 2026. It did not conclude with a decision. The forty-five-day window passed, no order was made, and the deal cleared by operation of the clock. The net-benefit review followed and was granted in late May.

And then the ending nobody drafting this story would have written.

On 29 July 2026 the takeover collapsed — not in Ottawa, but in Beijing. Zijin could not obtain Chinese regulatory approval, and terminated. It settled instead for a minority position: 12.8 million shares at C$32.55, C$416.6 million, about 9.2% of Allied Gold, closed on 10 August 2026.

Follow the reviews in that sequence and count which ones bound anything.

Canada's security review: expired.

Canada's net-benefit review: granted.

China's regulator: refused.

In 2026, on a Canadian company, on Canadian soil, under a Canadian statute, the only review in the chain that actually stopped anything was the one held in Beijing. That is not an accusation. It is the sequence of events, and it is the cleanest single illustration this volume has of what a periodic review is worth when it still has a decision at the end of it — and what it is worth when it does not.

Every runner
announced they
were the last.

Go back through the list one more time, reading only the quotations.

Harper, December 2012: not the beginning of a trend, but rather the end of a trend.

Osborne, September 2015: let’s create a golden decade.

Morrison, April 2021: not in the national interest, and cancelled.

Sunak, November 2022: the golden era is over.

Meloni, December 2023: not renewed.

Carney, January 2026: a new strategic partnership.

Four of those are closures and two are openings, and they are not sequenced the way a policy argument would sequence them. They alternate. That is what a relay looks like from inside: each runner sincerely believes their leg is the decisive one, and each is followed by somebody who runs the other way and is equally sincere.

Which is why the argument about intent is a dead end. It does not matter whether Harper meant it, and there is no reason to think he didn’t. His restriction was policy, and policy is exactly as durable as the next cabinet’s appetite. Thirteen years and forty days, as it turned out.

Measured that way the list sorts itself, and it sorts against intuition. The loud instruments turn out to be the reversible ones. Victoria’s Belt and Road memorandum caused a national argument and was voided in thirty months. Italy’s caused a G7 incident and expired by letter. Britain removed a Chinese state company from a nuclear plant it had already been invited into. The US–Australia minerals framework got a White House signing ceremony and a dollar figure, and creates no rights or obligations under any law.

The quiet ones are the ones still standing. A ninety-nine-year lease on the Port of Darwin, signed by a territory chief minister in 2015, has survived two federal governments that promised to unwind it. A treaty tabled in Ottawa for twenty-one sitting days in the autumn of 2012, which nobody debated and nobody voted on, binds the Crown until 2045.

Nothing in that requires anyone to have been bought, or briefed, or coordinated. It requires only that the reviews kept being retired, and that each runner handed the baton forward.