Ausländer besaßen die Hälfte der US-Schulden. Jetzt besitzen sie ein Drittel.

On March 31, 2026, US debt held by the public crossed 100% of the economy. The Committee for a Responsible Federal Budget (CRFB) marked the milestone the next month: $31.27 trillion in publicly held debt against $31.22 trillion of GDP, a ratio of 100.2%. “We’ve heard plenty of alarm bells in the past few years about our fiscal path,” wrote CRFB president Maya MacGuineas, “but this one rings especially loudly.”

Outside a brief COVID-era crossing in 2020, the only previous time American debt exceeded the size of its economy was the immediate aftermath of World War II, when the ratio peaked at 106% in 1946. The country had just defeated Nazi Germany and Imperial Japan. This time, the only thing the country defeated was the discipline to stop borrowing.

The headline number is the part everyone will quote. The story underneath is the one that matters: the people buying that debt are not the people you think they are, and the swap happened a decade ago while no one was watching.

Half Becomes a Third

The traditional model of US sovereign borrowing rests on a single assumption: foreign central banks, oil exporters, and Asian export economies recycle their dollar surpluses back into Treasuries. That assumption is breaking.

In September 2013, foreign holders owned roughly 49% of all US Treasuries outstanding. By December 2025, that share had fallen to about 31%, a multi-decade low. The drop is not a denominator illusion alone. Foreign holdings did rise in absolute terms, reaching roughly $9.2 trillion by December 2025, but the total stock of US debt grew far faster. The marginal Treasury issued in 2026 finds a different home than the marginal Treasury issued in 2013.

The retreat is asymmetric. Allies are holding firm. Rivals are leaving. As of December 2025, Japan held $1.2 trillion and the United Kingdom held $0.9 trillion. China held $0.7 trillion, down from a November 2013 peak of $1.32 trillion, a decline of roughly 47% over twelve years. The dollar’s share of global reserve assets fell to a 31-year low in early 2026 as central banks diversified into gold and other currencies.

The marginal Treasury buyer is no longer a patient People’s Bank of China desk recycling export receipts. It is something else entirely.

Who Replaced the Sovereigns

Two new buyer classes absorbed the difference. One is boring. The other is dangerous.

The boring half: money market funds. Total money market fund assets stood at roughly $7.63 trillion as of late April 2026, near record levels reached earlier in the year. The bulk of that capital sits in government-focused funds that hold Treasury bills and short-duration Treasury collateral as their core assets. As foreign demand for long-duration coupons softened, the Treasury Department leaned harder on bill issuance, which the money market complex absorbed without complaint. This is real demand. It is also short-duration demand, which means the Treasury must roll the entire stock at whatever yield the market demands every few months.

The dangerous half: borrowed-up hedge funds running the cash-futures basis trade. A basis trade is a relative-value arbitrage. The fund buys a cash Treasury bond, simultaneously sells a Treasury futures contract against it, and finances the long position by borrowing cheaply in the overnight repo market at very high gearing ratios. The position earns a tiny spread on a very large notional book.

How large is the book? The Federal Reserve’s own measurement, published in March 2024, estimated that hedge funds had amassed at least $317 billion in Treasury holdings tied to basis trades since the first quarter of 2022, with hedge-fund short futures positions reaching $991 billion by January 2024. The Fed described the position as “typically highly leveraged” and warned that this borrowing “can increase Treasury market fragility.” Investment funds, the bucket that contains hedge funds and money managers, took down more than 70% of note, bond, and TIPS auction awards by late 2025.

A patient sovereign holder funds the United States by sitting still. A borrowed-up hedge fund funds the United States as long as the repo market keeps lending and the basis spread keeps existing. Those are very different financial products wearing the same Treasury label.

The 1946 Mirage

The standard counterargument to debt panic is the postwar drawdown. America carried 106% debt-to-GDP in 1946, the textbooks say, and the country simply grew out of it. The implied lesson: relax, the economy will outrun the borrowing.

That story is wrong. A 2023 NBER working paper deconstructed the postwar decline and found that economic growth alone explains a small share of the recovery. The dominant mechanisms were primary budget surpluses (the government ran one in all but four years from 1947 to 1970), surprise inflation that quietly devalued the principal, and explicit financial repression: regulatory caps that forced banks and insurers to hold low-yielding Treasuries regardless of whether they wanted to. The combined fiscal surplus alone equaled 43% of the 1946 debt stock by the time it was retired.

None of those tools are loaded in 2026.

The federal deficit already runs above 6% of GDP, and the CBO’s February 2026 outlook projects the debt-to-GDP ratio rising from 101% this year to 120% by 2036, surpassing the 1946 peak without a war to justify it. The Peter G. Peterson Foundation reports cumulative federal interest payments of $16.2 trillion over the next decade, citing CBO projections. Surprise inflation as a quiet default mechanism would require an independent Fed willing to tolerate it. As covered in the DOJ’s Trojan Horse on rate pressure, the administration is actively eroding that independence.

The financial repression playbook is reassembling, but the regulatory cage that made it work in 1946 (capital controls, Reg Q deposit caps, captive savings) was demolished decades ago. What remains is the pressure on the Fed without the architecture that historically channeled it into orderly debt reduction.

The Fragility Quotient

The buyer-composition shift would matter less if the new bid were robust. It is not.

The basis trade has a known failure mode. In March 2020, a sudden surge in volatility forced hedge funds to unwind positions, with margin calls hitting both legs of the trade simultaneously. Multiple Federal Reserve and BIS analyses found that this forced deleveraging was an important contributor (alongside foreign central bank selling and dealer balance-sheet constraints) to the broader Treasury market dislocation that followed, with the Federal Reserve responding by purchasing roughly $1.6 trillion of Treasuries between early March and the end of May to restore market function. The Dallas Fed’s 2025 paper examined how sensitive the basis trade is to funding-condition shifts and documented the rapid-deleveraging mechanism.

The math is unforgiving. Treasury auctions for 2-, 5-, and 7-year notes drew weak demand in early April 2026, and Bloomberg reported that traders were scouring every auction print for evidence of a foreign bid strike. The CBO’s February 2026 outlook projects cumulative deficits of $23.1 trillion from 2026 to 2035, all of which has to clear at some price. If foreign sovereigns are absent and the borrowed-up bid wobbles, that price is a higher yield. A higher yield directly compounds the interest cost that already exceeds the entire defense budget, as documented in the Imperial Overdraft analysis of Q1 fiscal year 2026 spending.

What Comes Next

The crossing itself is not the cliff. The buyer base is.

Three signals to watch in the next two quarters:

1. Auction tail size. The gap between the yield expected and the actual stop-out yield at Treasury auctions is the cleanest measure of demand stress. Sustained tails above two basis points on long-duration paper would mark the moment the borrowed-up bid stops absorbing supply.
2. Repo market spreads. The basis trade dies when overnight financing gets expensive. Watch the Secured Overnight Financing Rate (SOFR) versus interest on reserve balances; a sustained widening signals borrowing is being forced down.
3. Fed political pressure. The administration’s clash with Chair Powell, examined in the DOJ rate-pressure analysis, is the precondition for financial repression. A capitulating Fed buying Treasuries during sticky inflation would mean the 1946 playbook has been forced back into use, this time as a silent default rather than an orderly drawdown.

The watchdogs are right that crossing the threshold is a warning. They have the wrong warning. The country is not borrowing more than it can pay; the country is borrowing more than its old creditors will fund. The Office of Financial Research’s most recent Hedge Fund Monitor measured repo borrowing for the top ten hedge funds at roughly eighteen-to-one as of the third quarter of 2024. The new creditors, money market funds with three-month memories and that levered Treasury complex, will keep the auctions clearing as long as nothing breaks. They have broken before. They will break again. The only question is what the Federal Reserve is allowed to do when they do.

The blank check is not blank. The signature on it changed.