The Crown Nobody Wants

01An afternoon in November

Britain did not lose Suez on the ground. It lost Suez in a currency market, and the instrument of its defeat was a queue of people who wanted their money in a different denomination. The International Monetary Fund eventually approved $1,300 million for Britain — the full 100 percent of its quota, structured as a $561.5 million immediate drawing plus a $738.5 million standby — but not until 10 December 1956, after the troops were already leaving. The Eisenhower administration held that approval hostage to withdrawal and blocked a $500 million Export-Import Bank credit alongside it. Historians of the crisis also describe American willingness to sell UK government securities outright; that threat appears in the accounts of Diane Kunz and others rather than in the IMF’s own archival record, which is a distinction worth keeping.

What is remarkable is not the coercion. Great powers coerce. What is remarkable is what the coercion was made of. No American fired on a British soldier. The weapons were a currency, a credit line, and a quota — and they worked in about a week.

John Rector put the parallel to me plainly: the fall of the British Empire looks amazingly similar to what is happening to the United States now, and the root cause was financial. He is right, and more right than the popular version of the argument, which usually gets the mechanism roughly correct and the conclusion badly wrong. So let me take the parallel seriously, follow it exactly as far as the evidence carries it, and then say clearly where it stops — because where it stops is the interesting part.

02The bill was written in 1913

Suez was the invoice. The purchase had been made forty years earlier.

In 1913, Britain held roughly £4 billion in overseas assets against a national output of about £2.4 billion — foreign claims worth something on the order of 150 percent of GDP, and around 32 percent of the country’s entire net national wealth. No nation had ever been so rich in other people’s promises. Sterling accounted for roughly 48 percent of known official foreign-exchange holdings. Britain was not merely powerful. Britain was the world’s savings account.

Then it spent it. Two wars converted a century of accumulated claims into ammunition and food. By 1945 Britain had gone the other way entirely, owing more than £3 billion in what were called sterling balances — money that India, Egypt and the rest of the sterling area had effectively lent Britain by accepting paper in exchange for wartime supply.

The rescue arrived as the Anglo-American Financial Agreement, signed on 6 December 1945 and effective from 15 July 1946: $3.75 billion at two percent. Attached to it was a condition that reads, in hindsight, like a fuse. Britain was to make sterling convertible for current transactions within one year of the effective date.

It did, on 15 July 1947. Convertibility lasted thirty-six days. The holders of those wartime balances converted into dollars almost immediately, the reserves drained, and the whole thing was suspended on 20 August.

A reserve currency is a promise that you can leave. Britain made that promise in December 1945, kept it for thirty-six days, and spent the next quarter-century explaining why.

03Ferguson’s line

Niall Ferguson gave this pattern a name in a Hoover Institution working paper in February 2025. Ferguson’s Law: the point at which the cost of servicing a great power’s debt surpasses its military spending. He is careful that this is a warning light rather than a death certificate, but the mechanism is not subtle. Debt service is a first claim on revenue. Defense is a residual. When the first claim grows, the residual is what gives.

The examples run through Habsburg Spain, the France of the ancien régime, and the Ottoman Empire, where debt service exceeded military spending by factors of 2.0, 1.76 and 2.7 in 1872, 1874 and 1875 respectively. And Britain, after 1918 — before Suez, before the loan, before the balances came due.

The United States crossed the line in fiscal 2024, by about seven billion dollars, which in a $6.75 trillion budget is a rounding error and in a historical argument is a threshold. Where it goes from here is not ambiguous. Net interest was $970 billion in fiscal 2025 — an actual, not a forecast — and CBO’s February 2026 baseline puts it at $1.04 trillion for fiscal 2026, then $2.14 trillion by 2036 against roughly $1.1 trillion of defense spending. Nearly two to one. Gross federal debt stood at $40.10 trillion in early September 2026, about 123 percent of GDP; debt held by the public was $32.42 trillion, just under 100 percent.

So the mechanism John identified is real, it is operating, and it is the same mechanism. That is the part of the parallel that holds. Now the part that doesn’t.

04Where the parallel breaks

Britain fell as a creditor. That sentence does most of the work in this essay.

When the bills came, Britain had a portfolio to sell, and it sold it — foreign holdings liquidated to pay for the wars and for the settlement afterward. The decline was financed by selling assets to foreigners. It was humiliating, and it worked, because there was something there to sell.

The United States has no such portfolio. Its net international investment position was −$21.27 trillion in the first quarter of 2026, roughly −67 percent of GDP. America is not a creditor beginning to spend down. It is already the largest debtor in the history of the world, and it reached that position during the good years.

This changes the shape of the problem entirely. Britain declined by liquidating. America would have to decline by borrowing more — from the same foreign investors who already hold about $9.3 trillion of Treasuries, roughly thirty percent of the marketable stock. That is not the British trajectory with a later start date. It is a different trajectory with fewer exits.

Figure 01 Two empires at the moment of reckoning

Measure	Britain, 1913–1956	United States, 2026
External position	Net overseas assets around £4 billion against £2.4 billion of GDP — roughly +150% of national output, held as claims on the rest of the world.	Net international investment position −$21.27 trillion, about −67% of GDP. The largest debtor position ever recorded.
How decline gets financed	Liquidation. Overseas holdings sold to pay for two wars and the settlement that followed.	Issuance. More borrowing from the same foreign holders, who already own roughly 30% of marketable Treasuries.
Debt service vs. defense	Crossed after 1918, decades before the political humiliation arrived.	Crossed in fiscal 2024: $881.0 billion of net interest against $874.0 billion of national defense.
The successor’s balance sheet	The United States in 1945: largest creditor on earth, open capital account, courts that would rule against its own government.	China in 2026: managed capital account by design. The renminbi is 1.99% of official reserves and 3.10% of global payments.
What holders bought instead	Dollars — a claim on a richer, more open state.	Gold — a claim on nobody at all.
Currency of the debt	Sterling balances owed to foreign holders, with convertibility promised to creditors as a condition of rescue.	Almost entirely dollars, issued at long maturities, with no convertibility promise available to break.

There is a second difference, and it points at the succession itself. Britain handed over to a creditor. The United States in 1945 was the richest state on earth, with an open capital account, deep bond markets, and courts that would enforce a contract against its own government. The handover was to something better, and most of the reluctance was British.

Which brings us to the question John is really asking. If not New York, then where?

05The heir that isn’t

The story almost everyone tells is a relay: London handed to New York, New York hands to Hong Kong. The first leg is history. The second is an assumption wearing history’s clothes.

Start with the numbers, because they are not close. As of the first quarter of 2026, the renminbi accounted for 1.99 percent of world official foreign-exchange reserves. The dollar accounted for 57.13 percent. In payments, SWIFT’s Global Currency Tracker — renamed from the RMB Tracker in February 2026 — put the dollar at 50.99 percent of global payments by value in August 2026, the euro at 21.78, sterling at 6.97, and the renminbi at 3.10. That last figure was also 3.10 in June. Flat, not climbing.

Figure 02 Share of global payments by value, August 2026

US dollar 50.99%

Euro 21.78%

Pound sterling 6.97%

Chinese renminbi 3.10%

• The currency said to be taking over
• The incumbents

The infrastructure is real and growing. China’s Cross-Border Interbank Payment System now lists 1,829 participating institutions, 194 of them direct, and set a single-day record of RMB 1.22 trillion in April 2026. Against SWIFT’s 11,500-plus institutions that is small, but it is not nothing, and the trend is genuinely up.

Yet every constraint on the renminbi runs back to one decision, and the decision is not an oversight. China maintains a managed capital account. Reserve status requires the opposite.

1. Convertibility

   Any holder can exchange the currency for another, at any time, in any size.

2. Exit

   Including during a crisis. Including when the issuing state would very much prefer you stayed.

3. Depth

   A government bond market large and liquid enough to absorb the world’s savings without moving the price.

4. Enforceability

   Courts that will rule against the government that issued the bond, and a government that complies.

China has spent fifteen years building the third of those and has declined the other three. Not failed at — declined. Beijing watched Britain make exactly that promise in 1945 and watched it destroy the pound in five weeks, and has organized its entire model of financial stability around never being in that position.

The renminbi is not failing to become a reserve currency. It is declining to.

That is a defensible choice. It is also the choice that forecloses the succession, because the whole value of holding a reserve asset is the ability to stop holding it.

And look at what the supposed successor is actually doing with its own money. Chinese holdings of US Treasuries stood at $653.3 billion in March 2026. The United Kingdom — not China — has been the second-largest foreign holder since March 2025, at $939.9 billion. China has been reducing its dollar exposure for years without building a corresponding international role for its own currency. If the world were changing hands, this is not the shape it would leave in the data.

06What the world is actually buying

Gold. And here is where the popular version of this story gets a number right and the explanation badly wrong.

At the end of 2025, gold made up about 27 percent of global official reserves and US Treasuries about 22 percent — a crossover, and a genuinely arresting one. The European Central Bank published that figure in The international role of the euro on 2 June 2026, and it went around the world in a day.

The ECB also published the counterfactual, and almost nobody repeated it. Hold gold at its end-2023 price and recompute the same shares: gold 16 percent, Treasuries 26 percent. No crossover. Not close to one.

Figure 03 The crossover, and the counterfactual nobody quoted At market prices, end-2025

Gold 27%

US Treasuries 22%

Same holdings, at constant end-2023 gold prices

Gold 16%

US Treasuries 26%

• Gold
• US Treasuries

So the crossover is mostly the gold price. That does not make it meaningless — a revaluation that large is itself a market verdict, and central banks chose to hold the metal being revalued. But it is not the same event as central banks dumping Treasuries, and reporting it as though it were is how a valuation effect gets laundered into a stampede.

The genuine reallocation is smaller and still substantial. Central banks bought more than 1,000 tonnes of gold in each of 2022, 2023 and 2024, against an average of 473 tonnes across 2010 to 2021. The streak broke in 2025 at 863 tonnes — still nearly double the old normal. In the World Gold Council’s 2026 survey of 76 central banks, 89 percent expected global gold reserves to rise over the following year and 74 percent expected the dollar’s share of reserves to be lower in five years.

What matters most is what they are not buying. Not renminbi. Not euros in any quantity that changes the picture. Gold — an asset with no issuer, no central bank, no legislature, and no ability to freeze anybody’s account. It is precisely what you accumulate when you have stopped trusting the incumbent and have not started trusting the challenger.

That is not a handover. That is a hedge against there being nobody to hand it to.

07The city outlives the currency

Here is where John’s instinct about Hong Kong is right, though for a different reason than the one usually given.

Financial centers and reserve currencies come apart. Sterling handles about 7 percent of global payments. London is the second-ranked financial center on the planet — 766 points against New York’s 767 in the Global Financial Centres Index published on 26 March 2026, with Hong Kong at 765 and Singapore at 764. Four cities inside three points. (That edition is about to be superseded; the next lands on 16 September 2026, and these particular numbers expire with it.)

London lost the currency in the 1940s and kept the business for eighty years. It managed it by intermediating somebody else’s money: the eurodollar market was invented in London precisely because it let institutions trade dollars outside American jurisdiction. The City’s second act was built on the dollar, not the pound.

Hong Kong is running a version of the same play. It was the world’s top listing venue in 2025 — HK$285.8 billion, about US$37 billion, across 119 new listings — and through the end of August 2026 had already raised HK$342.4 billion, roughly US$43.7 billion, from 106 listings. It ranked second globally in the first half of 2026. The crown moves around, and Hong Kong is now genuinely in the running for it.

So yes: expect Hong Kong to take share, possibly a great deal of it. But that is an issuance and intermediation story, and it is entirely compatible with the renminbi never becoming a reserve currency — exactly as London’s dominance was compatible with sterling’s collapse. Conflating the two is the most common error in this whole genre of argument. The city is not the crown.

08Nobody knows what year it was

The last handover took so long that historians still cannot agree on when it happened. Barry Eichengreen and Marc Flandreau date the dollar’s overtaking of sterling to around 1924–26, and note that sterling regained the lead in 1933. Menzie Chinn and Jeffrey Frankel put the crossing as late as 1945. That is a twenty-year disagreement among serious scholars about an event that finished eighty years ago.

Figure 04 Sterling’s retreat, as the people who counted it counted it

1. 1913 ~48%

   Share of known official foreign-exchange holdings. Britain is simultaneously the world’s largest creditor, holding foreign claims worth roughly 150% of its own GDP.

2. c. 1924–26 Overtaken

   The dollar passes sterling, on Eichengreen and Flandreau’s dating — then sterling takes the lead back in 1933. Chinn and Frankel date the crossing to 1945 instead.

3. Late 1940s >80%

   The peak that is not a peak. A very large share of a smaller, largely blocked sterling area — the balances Britain could not afford to let anyone convert.

4. 1950 >55%

   The postwar arrangement holding, on borrowed dollars and exchange controls.

5. 1960s ~30%

   Roughly flat through the decade. Suez sits at the start of it; the 1967 devaluation near the end.

6. 1970 ~15%

   Sterling is now a regional currency with imperial paperwork.

7. Mid-1970s <10%

   About sixty years after 1913, and roughly twenty after the afternoon in November.

The lesson is not “this takes a long time, so relax.” Britain’s decline was slow and total, and the slowness is what made it survivable as politics and fatal as strategy. The lesson is that anybody who hands you a date is selling you something. Consider just the last two years of the dollar: the dollar index fell 10.8 percent in the first half of 2025, its worst first half since 1973 — and then spent 2026 roughly flat, up about a percent through August, while the Federal Reserve’s broad trade-weighted index was down about 0.8 percent over the same stretch. Two respectable indices, two directions, one year. That is what an ambiguous decline looks like from the inside, and it is what every year of the sterling story looked like to the people living in it.

09The ledger

Three claims, each with the condition under which I am wrong.

10What I left out

Things I wanted and could not stand behind, listed so the gaps are visible rather than papered over.

The frequently cited claim that roughly 80 percent of CIPS payments still travel on SWIFT messaging traces to a 2022 estimate that circulated through Eichengreen and CSIS. It was never official and has never been updated. It is probably still directionally right, and I would not build an argument on it.

I have left the euro almost entirely alone. It is the second-most obvious answer to “then who?” and a serious treatment requires the fiscal-union question, which is a different essay.

Ferguson offers a stricter version of his own law — interest plus principal against defense — on which the United States crossed years earlier. I used the narrower version because it is the one cleanly measurable from Treasury’s monthly statement, and I would rather use the conservative test.

On China’s own arithmetic: the macro leverage ratio reached 308.2 percent of GDP in the second quarter of 2026, the IMF’s augmented measure of public debt is 126.6 percent of GDP, twelve quarters of deflationary pressure ended in the second quarter of 2026, property investment is down about 44 percent from its 2021 peak with new starts down about 74 percent from 2019, and growth has averaged 4.90 percent so far in the 2020s against 7.69 percent in the 2010s. Ferguson’s Law does not care about flags. I have not tried to compute the debt-service-to-military ratio for China, because the military figure is not reliable enough to divide by, and a number I cannot check is worse than no number.

I also dropped a widely repeated figure putting local government financing vehicle debt at around ¥50 trillion, or 40 percent of GDP. The two halves do not reconcile — ¥50 trillion is closer to 36 percent — and I could not source the pairing where it is usually attributed. The IMF’s augmented-debt measure above is what I used instead.

11The afternoon, again

What I keep returning to is how ordinary the end looked. There was no surrender at Suez, no rival flag over Whitehall, no decisive battle anyone can name. There was a Chancellor, a column of reserve figures, a telephone, and an afternoon.

The crown did not pass in November 1956. It had already gone, quietly, across four decades of decisions about what to spend and what to sell, and 1956 was merely the first occasion on which everyone could see it. That is the part of John’s parallel that holds completely, and it is the part worth being uncomfortable about, because the American version of those decisions is being made now, in appropriations bills nobody reads, and the afternoon has not happened yet.

But the ending is different, and worse. Britain fell to a successor that was richer, more open, and ready. When the dollar’s afternoon comes, the evidence says there will be nobody standing there to take it — no open capital account large enough, no challenger willing to make the promise the job requires. The world’s central banks appear to have reached the same conclusion already, which is why they are buying an asset that nobody issues.

A reserve system with a hegemon is unfair. A reserve system with a vacancy is unpredictable. History suggests the second is the more expensive of the two, and nothing in the current numbers suggests we get to choose.