'I am the house now': Treasury buys yen, buys bonds, dares you

Archived reading, published Wed, 09 Sep 2026 13:44:22 UTC (26 hours ago). This is not the current state of the meter.

See the live reading →

CRASH-O-METER

Cracking

how close are we

Fragility91

how much tinder is stacked up — moves slowly

Ignition39

how close a spark is — moves fast

Crash Lab watches the machinery under the current boom: the debt paying for AI data centres, private credit, the leverage that has moved out of banks into places nobody has to mark, the bond market and the dollar.

Every four hours it reads the day’s reporting from 21 sources and rewrites this page.

Status: Held at 65. Stock and credit prices are the same September 8 close we read at midnight: the S&P 500 is 1.6% off its high, the extra interest the shakiest companies pay compared with the government is unchanged (2.68 percentage points), and the interest rate the US government pays to borrow for ten years is 4.81%. Since then the VIX, the market's gauge of how much turbulence traders expect over the next month, has risen half a point to 16.2, which is still well inside the calm zone below 20; gold is up 1.5%; and the yen has stretched a five-day rally to 3.8%. None of that is enough to move ignition off 39 before the Treasury's 11am buyback announcement, which is the real test and lands after this run. Fragility holds at 91. The Swiss franc borrowing data measures AI debt we already count; Cogent's 11% refinancing is one company's price being tested in public rather than a change to the whole system; Bessent's remarks about the yen and the buybacks are a statement of intent, not anyone being forced to sell; and nobody was forced to sell anything.

Reporting from 7 Sep to 9 Sep

Crash points

Eight places it could go, scored 0–10. Tap one for the explainer.

What’s moving the needle

What changed in the last few hours, and what each one says about the plumbing.

The AI capex bubble9 Sephigh

Two percent of Switzerland's bond market, per deal

The AI borrowing wave has outgrown the dollar market and is now large enough to reshape small currency markets one deal at a time.

Companies building artificial intelligence have sold more than a quarter of all the corporate bonds issued in Swiss francs this year (26.4 percent, according to Zurich Insurance research reported by the FT). That is a higher share than in any other major credit market. Alphabet raised about SFr3bn in February and Amazon SFr2.82bn in May, and each of those deals was close to 2 percent of the entire existing Swiss corporate bond market and more than ten times the size of a typical bond sold on the SIX Swiss Exchange. Nine months ago this market barely knew these borrowers. Now Swiss companies plan their own borrowing around them. The point is where the debt lands, not how much of it there is. The Swiss market has traditionally been made of bonds in the SFr200–250m range, bought by pension funds and insurers and held until they are paid back. Those same buyers are now swallowing single deals of SFr3bn from one sector. Arthur Jurus of the bank Oddo BHF calls it "a structural change of scale". His line that "AI issuers aren't creating the demand, they're revealing it" is generous to the borrowers and a little alarming about the buyers. Swiss corporate treasurers, he says, now move their own bond sales or shrink them to avoid the weeks when one of the big American cloud companies is in the market. Zurich estimates that Alphabet, Amazon, Microsoft, Meta and Oracle will spend about $5.5tn building things between 2025 and 2030. They have gone to euros, pounds, Canadian and Australian dollars and yen for the same reason they have come to francs: the dollar market alone is not big enough to fund that at the price they want to pay. For the fragility number this is measurement, not new stock. We already count the AI debt; the Swiss data tells us that a slice of it sits in the books of Swiss pension funds and insurers, held to maturity, unlikely to be repriced until something forces it. That cuts partly against our worry, because this is plain, graded, exchange-listed debt, and partly for it: a small, conservative market has let one sector become a quarter of its new borrowing in nine months, and the buyers are not the kind who sell. The odds of a crash did not change today. Our map of where the tinder is stacked got a little better.

Private credit and BDCs9 Sepmedium

Ten percent want out. Five percent get out.

Withdrawal limits mean the prices private credit funds put on their loans are never tested by the people leaving, only by the loans the manager is eventually forced to sell.

Investors in the largest fund that lends to companies in place of a bank asked to take out roughly twice as much money as it will give back. Blackstone Private Credit Fund, known as BCRED, holds about $77bn and is not traded on any exchange; it disclosed on September 3 that investors asked to withdraw about 10 percent of their shares this quarter. It will pay 5 percent, the ceiling it has always had, which leaves about $4.3bn of requests stuck behind the limit. Cliffwater's flagship corporate lending fund reportedly had requests for about 16 percent of its shares and likewise paid 5. A fund refusing to give back half of what its investors ask for is not a normal quarter. These figures reach us second-hand through Bloomberg and Reuters rather than from the funds' own documents, so treat the decimals loosely; the shape is not in doubt. Here is the shape. A fund of this kind promises to buy back up to 5 percent of its shares each quarter, at whatever price it says its loans are worth. When more people ask than that, everyone gets a slice and the rest of their request lapses; they have to ask again next quarter, in a longer line. Two things follow. The people who get out are paid the fund's own valuation, not a price anyone in the open market agreed to, so the limit guarantees that valuation is never tested by an exit. And every investor who suspects the valuation is flattering has a reason to get in line early, which is how a 10 becomes a 16. The manager can meet the requests with new money from wealthy individuals, with bank credit lines, by lending less, or by selling loans. Only the last one tests the valuation. That is why two other prices matter more than the withdrawal numbers: a second-in-line loan to a company called Loparex, held by Blue Owl, fell from 63 cents on the dollar to 5 cents in a single quarter, and a Carter Place property loan held by Metrics drew bids far below what it was owed ($270m). Those are the prices the people in line are waiting to learn. The publicly traded relatives fell again yesterday: the firm that manages Blue Owl's funds was down 4.4 percent and the big listed lending funds, the ones that borrow, lend to mid-sized private companies and pass the interest to shareholders, were down 1–2 percent, while the extra interest shaky companies pay compared with the government sat unchanged at 2.68 percentage points. One analysis puts the share of BCRED's loans on which interest has stopped arriving at 2.2 percent of what it paid for

Fed, Treasury and policy9 Sepmedium

A week to the Fed's meeting: a rate rise expected, a cut demanded

A central bank raising rates while the Treasury buys back bonds and the president demands cuts is three signals pointing three ways at a long-term bond market that has already stopped charging a safety discount.

The Federal Reserve, America's central bank, which sets interest rates, meets on September 15–16, and traders think it is more likely than not to raise them. On Friday the president posted that the Fed board "with its great new leader, must get smart — BE PATRIOTS for a change" and cut rates instead. A president publicly demanding the opposite of what the market expects a week before a decision is not the ordinary run of things. Bets on the Fed's next move put the chance of a rise somewhere between a coin flip and two-to-one on (56 percent per CME FedWatch on September 8, 66 percent on August 31). Investors, Bloomberg's editorial board notes, shrugged. The thing to hold in mind runs through long-term borrowing, not short-term. Christopher Waller, one of the Fed's governors, said on September 3 that there is "no more premium for safe, liquid US government debt" and that this pushes the rate the Fed should aim for higher, while saying he leans toward leaving rates alone if inflation cools. In other words, lenders no longer give the US government a discount for being the safest borrower in the world. Put that beside the Treasury's buybacks. One arm of the government is buying ten-to-twenty-year bonds to pull long-term rates down. Another is considering a rate rise that would push short-term rates up. And the president is demanding the opposite of what the market expects from the second arm, which is exactly the kind of noise that makes lenders demand more for tying money up for a long time, the very thing the first arm is trying to squeeze down. None of this is a crisis. The rate banks charge each other overnight sat at 3.63 percent, almost nobody parked spare cash with the Fed overnight ($525m, a rounding error), and the cash banks hold at the Fed edged up to $2.93tn. The short-term cash that keeps trading desks running is flowing normally. The government pays 0.44 percentage points more to borrow for ten years than for two, and 5.25 percent to borrow for thirty. The risk this week is not in the plumbing. It is in whether the new chair raises rates ahead of the midterm elections over the president's public objection, and what that does to a bond market that Waller says has already stopped paying for safety. That is a story about how close a spark is, not how big the pile is, and this week it moved neither. What we will watch: how the vote splits on September 16, and whether the thirty-year rate moves on the decision or on the presidential post that follows i

Bond market dysfunction8 Sephigh

The house has two tables and one stack of chips

The Treasury Department, which borrows the money the government spends, is now the buyer holding up prices in two markets at once, and its own reporting says it has limited money for at least one of them.

The Treasury Department, which borrows the money the government spends, is now propping up two markets at the same time: it is buying Japanese yen for the first time in thirty years, and it is buying back its own long-dated bonds. On Tuesday the man who runs it, Scott Bessent, told an audience at Southern Methodist University that this makes the Treasury "the house". "When we intervene with the Japanese yen, I have pretty good insight into what the Bank of Japan is going to do," he said. "And you can bet against me if you want." A Treasury Secretary talking about the government's finances as though they were a betting book is not routine. It is, as far as we know, unprecedented. Count what the house is doing. It bought yen on July 31 alongside Japan's finance ministry, the first US yen purchases in three decades. It is repurchasing its own government IOUs that do not come due for ten to twenty years, having said on August 19 it would "at least double" the $2bn operations it had penciled in two weeks earlier; the size of Thursday's buyback is due at 11am Washington time today. And it is publicly pressing the Bank of Japan, Japan's central bank, to raise interest rates. These are one move, not three. Japan spent a record sum defending the yen in August (¥15.4tn, about $98.6bn), and its holdings of foreign bonds fell by $87.8bn; the working assumption is that most of what it sold was US government bonds. If the Bank of Japan raises rates, the yen strengthens on its own and Japan does not have to sell those bonds. Failing that, the Treasury's buybacks soak up some of what Japan sells. So the Treasury is buying bonds back from the very country it is asking to stop selling them, while both governments buy the currency that would make the selling unnecessary. The chips are the constraint. Bloomberg noted last month that the Treasury has only limited money for buying currencies, and the July intervention wore off within a few days of trading. On the bond side, Morgan Stanley puts the practical ceiling for a single buyback at $10bn, Lou Crandall of the research firm Wrightson thinks $5–6bn is plausible, and most forecasters on Wall Street expect more than $4bn. Too small a number disappoints a market in which the ten-year borrowing rate (4.81%) is now higher than it was when the August promise was made. Too large a number, Crandall wrote, is "an admission that the Treasury hadn't thought through its hasty August 19 announcement". The yen is up about 4% this month w

Hidden leverage and shadow banking8 Sepmedium

Eleven percent, 96 cents, and a better seat

The market's average price for shaky corporate debt says credit is calm; the first borrower through the 2027 refinancing wall is being asked to pay 11% and to push its existing bondholders to the back of the line.

Bank of America has spent the past month trying to raise $750m of new debt for Cogent Communications, a telecoms company, to pay off bonds that come due in June 2027 and currently cost Cogent 7%. The price being discussed for the new debt has climbed from 9.5% to 11%, Bloomberg reports, and buyers still want more: the bonds would be sold at a discount (96 cents on the dollar), and the bank has floated a structure in which the new lenders get guarantees or pledged assets that the existing bondholders do not. That is the story. Two years ago this company borrowed at 7%. Today the market wants 11%, a discount, and a better seat. The new bond would be first-lien, meaning first in line for Cogent's assets if the company fails, where the old bonds are unsecured, meaning backed by nothing in particular. Prospective buyers also want promises written into the loan about what Cogent does with the money from selling its data centers. Nothing is agreed. Two things about how this works. First, the old 2027 bonds traded at about 91 cents on the dollar on Friday, having been at almost 100 cents on August 7. Cogent's business did not change in that month. What changed is that its lenders learned someone else might be offered a better seat at the table. The structure has a name, "pari plus": one borrower, two classes of lender, and the newer class ahead of the older one in everything but the label. It used to be the last resort of companies in real distress. It is now the opening bid on an ordinary refinancing. Second, look at the average. The extra interest the shakiest companies pay compared with the government sits at 2.68 percentage points, within a whisker of where it sat all summer, which is to say calm. Cogent, a going concern offering lenders first claim on its assets, is being asked for 11% plus a four-point discount. Before the discount that is about 6.5 percentage points more than the government pays to borrow for two years, by our arithmetic, or more than double the average. The average is made up of companies that do not need to borrow this quarter. The price for those that do is set one deal at a time, and this deal says the door is far narrower than the average implies. One more detail. Cogent sold ten data centers it inherited from Sprint for about $225m in June. It is selling data centers into the most eager data-center market in history and still cannot refinance a 2027 debt for under 11%. Old racks and AI racks, it turns out, are different assets. Bloomb

Crypto and TradFi contagion8 Sepmedium

Visa is lending to card issuers on the blockchain

The everyday credit that card companies run on is moving from bank loans to computer contracts, which is the whole thesis of this publication in miniature.

Visa now runs more than 160 card programs tied to digital tokens meant to be worth exactly one dollar, almost three times as many as a year ago (up almost 200 percent), and its head of crypto, Cuy Sheffield, told CNBC that new card issuers are joining "every week". That is not the important sentence. The important one is that Visa is now testing, with a company called Credit Coop, a way to lend those card issuers money through contracts that run on a blockchain rather than through a bank, feeding its own payment data to the lenders so they can judge who is good for the money. Why a card issuer needs credit at all: there is a gap between the moment a cardholder spends and the moment the issuer gets paid back, and somebody has to fund that gap. In the ordinary system that is a bank loan, sitting on a bank's balance sheet, the list of what it owns and owes, where a regulator can see it. Here it is a piece of software, funded by whoever puts money in the pool, secured on money owed in digital dollars, with Visa's data as the thing that decides who qualifies. This is small. There is no number for the size of the trial and we should not pretend there is. But it is a clean example of the pattern this publication exists to track: a lending relationship that would once have lived inside a bank is being rebuilt outside one, faster than anyone is measuring it. "Hypergrowth mode" is Sheffield's phrase. Coinbase fell 3 percent yesterday and bitcoin is flat on the week (about $79,600); the market is not treating any of this as a risk, and it is probably right, for now. It changes the crash odds very little. We are including it because it shows where the next unmeasured pile is being stacked.

Signs of the times

Stuff you wouldn’t have believed was possible until 2026.

Swiss companies now steer clear of Big Tech weeks

Amazon's bond sale in May (SFr2.82bn) was close to 2 percent of Switzerland's entire corporate bond market and more than ten times a normal Swiss deal. Swiss companies now move or shrink their own bond sales to avoid the weeks when a US tech giant is borrowing.

ai-debtconcentrationscale

9 SepFinancial Times

'I have asymmetric information'

The US Treasury Secretary told a Texas audience that intervening in the yen is "my dream" because "I have asymmetric information" about what the Bank of Japan will do, and invited traders to bet against him. A former hedge fund manager is running the Treasury Department as if it were a trading book, and saying so out loud.

interventionpolicyhubris

8 SepBloomberg

A new digital-dollar card issuer every week

Visa runs more than 160 card programs tied to digital tokens meant to be worth exactly one dollar, up almost 200 percent in a year, with new issuers launching weekly, and is now testing blockchain-based credit lines to fund them. The cards grew faster than the bank credit needed to run them, so the credit is being built on the blockchain too.

stablecoinsshadow-creditgrowth

8 SepCNBC

China's 22nd straight month of gold buying

China's central bank added 650,000 ounces of gold in August (roughly $2.9bn at today's price, our arithmetic), its biggest monthly purchase since late 2023, taking its holdings to 76.73m ounces. With US government bonds paying 4.8 percent, the world's most persistent official buyer keeps choosing the asset that pays nothing.

goldreservesde-dollarisation

8 SepCaixin

The rumour mill

What the crash-callers are saying, checked against real reporting.

Confirmed

“Amazon signed a multi-generational deal to buy Qualcomm data-centre chips, and Qualcomm issued Amazon warrants for 25 million shares at about $161 each — up to about $4bn of stock — vesting in tranches tied to roughly $60bn of chip orders.”

This is the interesting one: a chipmaker paying its customer in its own shares for buying its chips, the same circular pattern as Nvidia's support for OpenAI, now spreading to a second vendor. Both companies have disclosed it.

Claimed by Meet Kevin

Partly true

“Fed funds futures are pricing about a 60% probability that the Fed hikes at next week's meeting.”

Roughly right, wrong on precision: CME FedWatch showed 66% on August 31 and about 56% on September 8, so the number is drifting down toward a coin flip, not sitting at 60. Either way a rate rise is a live possibility a week out while the president demands a cut.

Claimed by Peter Schiff

Confirmed

“In March, new US rules were proposed to open 401(k) and other retirement accounts to private equity and private credit investments.”

The Department of Labor proposed the rule on March 30. Old news, but relevant to the story above about funds limiting withdrawals: retirement accounts are the fresh money from ordinary savers that those funds are counting on to pay off the investors who want out.

Claimed by ITM Trading (Daniela Cambone)

Earlier developments

Dispatches from previous readings. The same argument, no longer the news.

Private credit and BDCs8 Sepmedium

Investors are selling the lenders, not the loans

When the loans themselves cannot be traded, trouble shows up first in the shares of whoever collects the fees on them, and that is where it showed up this week.

Blue Owl, one of the biggest firms lending privately to companies, saw its shares fall again on Monday and has now lost the better part of a tenth of its value in a week (down 4.4% Monday, 8.3% on the week, 11% below its high). Its rivals fell with it (Blackstone down 6.5% on the week, KKR 3.6%, Apollo 3.6%, Ares 3.3%). The funds these firms run, which borrow money, lend it to mid-sized private companies and pass the interest to shareholders, fell less (1.7% to 2.0% on Monday). And the public market for the debt of companies too shaky to be considered safe did almost nothing: the extra interest those borrowers pay compared with the government is 2.68 percentage points, three hundredths of a point more than a week ago. Shares down 8%, bonds unmoved, is not what a normal week looks like. That gap is the story. The things falling are the managers, whose value is the stream of fees they earn on money they look after. The thing not moving is the public market in shaky company debt, which holds bonds of large companies with public credit grades. The loans being written down are neither. They are loans to mid-sized software and industrial companies, with no public grade, sitting inside private funds. Reuters looked at 44 of these lending funds and found they valued their loans at 97.6 cents on the dollar at June 30; at ten comparable funds, loans on which interest has stopped arriving rose to about 3.4% from 2.5% at the end of last year, and 81% of software loans had been written down this year. Blackstone's roughly $77bn fund, BCRED, let investors take out 5% of their money when they asked for 10% (about $4.3bn honored), and Cliffwater's main fund capped withdrawals at 5% against 16% requested. None of that is new this week; we have counted all of it. What is new is that stock investors have started to put a price on it, in the one place it can be priced. When a fund limits withdrawals, it protects the value it puts on its own holdings by refusing to sell loans at prices that would show what they are really worth. That does nothing for the manager. Its next fund is harder to raise, and its fee income shrinks if more money leaves than arrives. So investors are selling the business rather than the loans, which is what you would expect when the loans cannot be sold. What we cannot see from outside is whether the selling reflects knowledge of what the third-quarter valuations will show, or just a week of bad headlines. Blue Owl's lending fund trades at a reported 22

Fed, Treasury and policy8 Sephigh

Wednesday at 11am we learn how big the fever cure is

For the first time since August, the Treasury Department's attempt to hold down its own borrowing costs gets a result with a date on it.

At 11am Washington time on Wednesday the Treasury Department, which borrows the money the government spends, is expected to say how much of its own ten-to-twenty-year debt it will buy back on Thursday. Buybacks themselves are routine, and until August 19 the answer was a routine $2bn. Then, with no warning, the Treasury said it would "at least double" that, long-dated government IOUs rallied, and the rally has since fully reversed: the rate the government pays to borrow for ten years closed Monday at 4.81%, above the level that prompted the announcement, and the thirty-year rate at 5.26%. Scott Bessent on Tuesday described the program as aimed at cooling "this fever that was building". In other words, the Treasury is trying to talk and buy its own interest costs down, and so far they have gone up. Bloomberg's survey of the numbers in play: Lou Crandall of Wrightson ICAP thinks $5bn to $6bn is a plausible start; Morgan Stanley puts $10bn as a practical ceiling; anything as low as $4bn would disappoint. Crandall's warning is that a large number "would be an admission that the Treasury hadn't thought through its hasty August 19 announcement". How the buyback works matters more than its size. A buyback does not retire debt. The Treasury pays for the long-dated bonds it buys by selling more short-term IOUs, the kind that come due within weeks. It is a swap of dates, not of amounts: the government takes debt that does not come due for decades out of the market and replaces it with paper that has to be borrowed afresh every few weeks. That is, structurally, the same trade run by the hedge funds that borrow heavily to profit from a small gap between two nearly identical prices, who hold $3tn of cash borrowed overnight against bonds put up as security to own long-dated debt. The Treasury is running that trade on itself. It is doing so at the moment Christopher Waller, a governor of the Federal Reserve, America's central bank, says the "safe, liquid" premium on government debt is gone, meaning investors no longer pay extra for the comfort of owning it, and that price-sensitive funds rather than central banks are now the buyers who set the price. The test is Thursday. If the number is small, the market learns the tool is a gesture. If it is large, a new baseline is set for the next operation two weeks later, and a government running a $2tn deficit is seen buying its own long-dated debt with money borrowed for weeks at a time. Either way the question is whether $5bn t

The dollar, gold and reserve status8 Sepmedium

Japan's two drains on the US government bond market

The largest foreign holder of US government bonds is selling for one reason, its private savers are being handed a second, and the buyer stepping in to replace them is a hedge fund running on borrowed money.

Japan's holdings of foreign currency reserves fell by $80bn in August, the largest monthly drop since its finance ministry began keeping records in 2000 (down 6.18%, to $1.207tn). The ministry gave no reason; Kyodo, citing an official, put it down to buying yen and to falling bond prices. Japan has spent a record 27.1tn yen this year propping up its currency, including a July operation carried out jointly with the United States, the first coordinated support for the yen since 1998. The yen has recovered from a forty-year low (163.98 to the dollar on July 23) to 155.98, and the fund that tracks it is up 3.8% on the week, at its high for the period. That is the official drain. The ministry sells dollar assets to buy yen, and the largest foreign holder of US government bonds, with $1.1tn of them, gets smaller. Masahiko Loo of State Street calls it "policy action rather than financial stress", which is fair. It is also irrelevant to the US bond market, where a seller is a seller. The second drain has not started, and Bloomberg is careful to say there is "little sign of a rush yet". Japan's own government pays 3% to borrow for ten years, a level it touched last week for the first time since 1996. Japanese investors hold almost $5tn of foreign assets, built up over decades because Japanese government bonds paid nothing. At 3% the sums change: a domestic bond paying 3% with no currency risk, against a US government bond paying 4.8% less the cost of insuring against the yen moving. Ales Koutny of Vanguard: "If domestic yields continue to rise, Japan may gradually retain more capital at home", which he says "matters not only for the yen and JGBs, but also for Treasury markets". Put plainly, Japanese savers might stop lending to Washington and lend to Tokyo instead. The GPIF, the world's largest pension fund, is "still examining" whether to review where it puts its money, its minister said Tuesday. We do not have the cost of that currency insurance or figures on what private savers are actually doing. What we have is a yen that has risen 4% in a month and two of the world's largest pools of money, the ministry's reserves and the nation's savers, with reasons to bring money home at the same time. This adds to fragility rather than ignition. It does not add tinder; it removes one of the steadiest buyers of the tinder, slowly, and it confirms a shift in who holds American debt that we had already counted.

The dollar, gold and reserve status8 Sepmedium

The cheapest money in the world is now Chinese

A currency loses its place at the center of the world's finances first through the borrowers who stop using it, not the central banks, and this year borrowers went elsewhere on a record scale.

Borrowing in China's currency outside mainland China has reached Rmb1tn ($149bn) so far this year, according to the Financial Times' calculations, already past last year's full-year record. Renminbi bonds sold outside the mainland, known as dim sum bonds, account for Rmb786.3bn; bonds sold inside China by foreign borrowers, known as panda bonds, account for Rmb231.6bn, more than in any full year on record. The reason is arithmetic. The Chinese government pays 1.68% to borrow for ten years against 4.78% for the US government, close to the widest gap there has ever been. Kazakhstan's state oil company raised Rmb3.5bn last month; Tencent has sold bonds that do not come due for ten and thirty years. This is not the world fleeing the dollar. It is borrowers going where money is cheapest, and China's money is cheap because its economy is weak and Beijing wants its currency used more. But the dollar's global role has always rested partly on being the cheap, deep currency people borrow in. The yen played that part when investors borrowed cheaply in yen to bet elsewhere; the dollar played it from 2010 to 2021. Nearly two thirds of the dim sum borrowers are Chinese companies funding their overseas operations, so part of this is China lending to itself abroad. Still, a record volume of the world's borrowers choosing renminbi is a change in how money moves. Two other facts this week pull in opposite directions. China's central bank bought 650,000 ounces of gold in August, its twenty-second month of buying in a row and its largest purchase since 2023, into a price that rose nearly 10% over the month. And a study from the New York branch of America's central bank argues that talk of the world leaving the dollar is overdone: the dollar's share of the world's reserves fell from 64% to 56% between 2015 and 2025, but almost the entire decline came from China and Russia, with Mexico and Morocco accounting for most of the rest between 2019 and 2023. Both can be true. Central banks as a group are not leaving the dollar. One very large one is, steadily, into gold, and borrowers are leaving it in the bond market because someone else is offering 1.68%. This changes little about the odds of a crash this month. It is a slow shift in who needs dollars, which is fragility of the slowest kind, and we include it because the record is worth noting rather than because anything broke.

The AI capex bubble8 Seplow

The borrowers rose 20%. Their lenders fell 8%.

Loans at 9% to the companies that build AI data centers are how the risk of the AI boom leaves the giant cloud companies and lands with private lenders, and this week the market judged the two ends of that trade in opposite directions.

The smaller companies that borrow to build AI computing and rent it out had a very good week. CoreWeave rose 11.7% on Monday and is up 17.6% on the week; Nebius is up 18.2%, IREN 26.3% and at a record, Core Scientific 14.8%, Applied Digital 11.4%. Oracle, the most debt-dependent of the giant cloud companies, closed at a high. In the same five days their lenders fell: Blue Owl 8.3%, Blackstone 6.5%, Apollo 3.6%. A borrower's shares rising while the lender's fall is not what usually happens, and it is the pattern that matters here. These are two ends of the same deals. Blue Owl led a $2.4bn financing for IREN on August 28, a $1.2bn loan and $1.2bn of bonds, both paying 9%, to buy Nvidia chips for a site in British Columbia. Applied Digital has $1.59bn of bonds paying 7% and backed by one site, and $2.15bn on another. CoreWeave added a $2.6bn loan it can draw on later in August. Nscale closed $3.05bn of borrowing backed by its assets. Oracle spent $56bn building things and paid for it with $43bn of debt and $5bn of new shares. Debt raised for AI reached about $220bn by August 10, is expected to hit as much as $400bn for the year, and now makes up about 7% of the indexes that track the bonds of companies considered safe. A lender at 9% has a fixed ceiling on what it can make, the interest, and owns the whole of the downside: what the chips are worth secondhand and whether the borrower's customers keep paying. The shareholder owns everything above the interest. So a week in which investors mark the borrowers' shares up 20% and the lenders' shares down 8% is saying one of two things. Either the two moves are unrelated, with the computing companies rallying on new contracts and the lenders falling over software loans, or investors like the assets and distrust the business of lending against them at 9%. Prices alone cannot tell us which. One more line from the same table. The customers with cash are sliding: Microsoft is 3.8% off its high, Amazon 9.6%, Alphabet 10.3%. The companies borrowing at 7% to 9% to serve them are at or near their highs. That is the order of things you would expect late in a buildout paid for with debt, not early in one. This adds to fragility. The borrowed money is real, it is now a meaningful slice of the market for safe company debt, and the people who lent it are being sold. Nothing here is a spark.

Private credit and BDCs7 Sepmedium

A $270m loan meets a real offer

The value a private lender puts on a loan is an opinion until a sale turns it into a fact, and this is a sale in progress.

Metrics Credit Partners, one of Australia's largest private lending funds, cannot find a buyer for Carter Place, a large housing construction site in western Sydney, after the developer failed to repay a $270m loan. The site has been on the market since June. Offers, according to people who spoke to the Australian Financial Review, have come in "well below the loan value". The paper does not say how far below, and neither do we know. A lender trying and failing to sell a site it has seized is the moment a private valuation stops being private. The mechanism is the simplest one in private lending. A construction loan sits on a fund's books at or near the amount lent until something forces a price. A borrower failing to pay does not force one on its own: the lender takes the site and can keep valuing it at whatever it likes. A sale does. Once an offer below $270m is accepted, the difference is a real loss to the fund's investors rather than an estimate, and every other lender holding a comparable Sydney site at a comparable value faces the question of why theirs is different. That is why the Financial Review describes the property industry as "on edge over valuations". It is not this loan that worries them. It is what it says about all the others. This follows Alceon getting out of its Bathla loans and four Australian funds limiting withdrawals, which we counted last week. Australia is a small market, and none of this touches American balance sheets, the lists of what firms own and owe. But it is the same test running on a smaller stage: a private valuation, unchallenged for months, meeting an actual offer. In the United States the equivalent tests, Loparex valued at 5 cents on the dollar and BCRED's 5% cap on withdrawals, have so far been settled by the fund declining to sell. Metrics is trying to sell. The price it accepts, or refuses, will be the most informative number to come out of Australian private lending this year. Until it lands, this changes nothing on the meter; it is a measurement waiting to be taken, and we would rather say so than pretend to know the answer.

The AI capex bubble8 Sepmedium

The label that opens the $11.7tn door

A safe-borrower label is the switch that would move AI debt out of the funds that lend to companies and into pension and insurance portfolios that neither trade what they hold nor reprice it.

Morgan Stanley and Goldman Sachs have spent recent weeks talking to the three firms that grade how likely a borrower is to pay you back, on behalf of Anthropic and OpenAI, arguing that both should get the top-tier stamp, investment grade, the moment they sell shares to the public, the FT reports. Both companies lose money and neither has said when that sale will happen. The prize is entry to the $11.7tn market for corporate bonds, where the biggest pools of retirement money are only allowed to lend to borrowers carrying that stamp. One senior credit analyst put the pitch bluntly: "Wall Street is trying to minimise their overall debt impact by arguing that these two companies will soon be flush with liquidity." In other words, the argument is that the two will shortly have so much cash on hand that their borrowing hardly counts. The letters matter more than the interest rate. A great many pension funds and insurers are limited, by their own rules or by regulators' capital rules, in how much debt they can hold from borrowers below that grade. So the stamp does not just make borrowing cheaper. It changes who is allowed to lend at all. That matters because of where the money already sits: Bloomberg notes that a record 46% of life insurers' debt holdings are already in loans made by investment funds rather than banks. A bond from an OpenAI carrying that grade is the instrument that would move the debt behind AI infrastructure out of the loans smaller cloud-computing providers are taking at 9% and into the pools of money that care least about price, the ones that hold a bond until it is repaid and almost never reprice what they own. There is one precedent, and it is recent. SpaceX went public in June and got the stamp immediately, the first large technology company ever to, helped by changes to index rules that pushed money from tracker funds into the shares on day one. That is not how it usually goes. Meta, Netflix and Tesla each waited a decade or more after going public for the same grade. The labs are not the only ones who would gain. Oracle and Nvidia stand behind the computing contracts and, in Nvidia's case, some of their customers' debt. If the firm on the other side of your deal can borrow more cheaply, your own credit looks better too. What we do not know: the grading firms have made no decision, neither company has laid out a path to a public listing, and both lose money. The filing that would let anyone test the "flush with liquidity" claim does not

Bond market dysfunction8 Sephigh

Britain borrows for thirty years at 5.82%

Every time old debt is replaced at today's rates, a move in interest rates becomes a permanent line in the budget, and the buyers who used to take these bonds without haggling over price are gone.

Britain sold £4.25bn of thirty-year government bonds on Tuesday at an interest rate of 5.82%, the highest since the office that sells its debt was created in 1998, the Telegraph reports. Governments sell debt every week; paying the most in a generation for it is what is new. Britain's interest bill already runs above £100bn a year, and every sale like this one adds to it for the next thirty years. The important word is "sold". An auction with no takers is a market breaking down. An expensive one is a price, and this one cleared. What it tells you is what a buyer of government IOUs that do not come due for twenty or thirty years now demands, and the number is not Britain's alone. The FT puts the combined interest bill of the OECD, the club of rich countries, above $2tn in 2025, 3% of everything those economies produce; Britain, France and the United States all now spend more on interest than on defense; the group expects to borrow a record $18tn this year; and the average ten-year rate across the seven largest rich economies is at 4% for the first time since 2008. Fidelity's Mike Riddell: "These massive sovereign debts are having to get refinanced at ever-increasing rates. Global investors are already starting to get scared." The mechanism is simple and slow. Debt borrowed at 1% or 2% comes due and is replaced at 5% or 6%. The interest bill rises every year even when the deficit does not, and the interest bill is itself paid with borrowed money. America is on the same path: the rate it pays to borrow for thirty years sits at 5.22% this morning and for ten years at 4.76%, with the national debt past $40tn. Two British dates worth having in the diary. On September 17 the Bank of England, Britain's central bank, sets next year's pace for selling off the government bonds it holds; with only around £30bn of those coming due on their own, FT Alphaville points out that keeping the shrinkage on schedule means selling long bonds into this market. On September 24 the debt office tests "switch auctions", a way of swapping bonds that come due far in the future for ones that come due soon. It is a test of the plumbing, no bonds change hands, but it signals that how far out Britain's debts fall due is now on the table. For the odds of a crash, this is the slow kind. The auction clearing means no spark. The price it cleared at is one more layer on the pile, and the pile is what turns a small fire into a large one.

Crypto and TradFi contagion8 Seplow

Bitcoin fell. The companies that hold it rallied.

When shares in the companies that hold bitcoin run far ahead of bitcoin itself, the gap is borrowed money the coin's price does not show, and banks are building the plumbing that connects it to their deposits.

Bitcoin fell this week and the companies whose whole business is holding it went up, and that gap is the story. The coin is at $78,301 this morning, down 1% overnight and 3.7% on the week, though up 13% over twenty days. Strategy, the largest of those companies, is up 12.1% on the week and 42.7% over twenty days, and sits 1.5% off its high. BitMine, which holds ether rather than bitcoin, is up 32.8% over twenty days; SharpLink 35.3%; Galaxy 30.6%; Coinbase 20.2%. Over the same twenty days the coin did a third of that. Over the last five, the coin went down and the companies went up. A company that exists to hold digital coins is a claim on those coins, made bigger with borrowed money, plus whatever extra the market is willing to pay for the company on top of what it holds. When the shares outrun the coins by three to one in a month, and rise while the coins fall, that extra is growing. This is our inference from the relative prices; we do not have the value of the companies' actual holdings in front of us and cannot say how large the gap is. But the extra is the engine. It is what lets these companies sell new shares, and bonds that turn into shares, to buy more coin, and it is the first thing that disappears when the mood turns. Prices are saying the market wants the company more than the thing inside it. Meanwhile the banks are laying rails. A sweep of the week's funding finds seven US regional banks, First Horizon, Huntington, KeyBank, M&T, Old National, SouthState and Glacier, putting $32.5m into Cari, a network for moving companies' bank deposits around as digital tokens; OpenReserve getting preliminary conditional approval for a banking license from the OCC, the regulator that supervises national banks; and Citi running live transactions in deposits turned into digital tokens on the ledger of Swift, the network banks use to send money to each other, on September 2 with First Abu Dhabi Bank and OCBC. Payments and banking infrastructure took 93% of the week's disclosed crypto funding. Twenty-one institutions including Citi, Bank of America, Goldman and Wells Fargo have committed this half to a company issuing a digital token meant to be worth exactly one dollar. That is the channel by which trouble in crypto would reach the banks, and it is being built on purpose: ordinary bank deposits as tokens on shared ledgers. For the odds of a crash, the first half of this story is more tinder, borrowed money hidden in a gap the coin does not show, and the second

Private credit and BDCs8 Sepmedium

Retirement savings in the owner's own loans: 42%, not 3%

This is the balance sheet the risk moved to: retirement savings invested in loans made by the insurer's own owner, valued by that owner, and supervised by nobody with a view of the whole country.

Two insurers controlled by Mark Walter's TWG Global failed to tell anyone that more than $20bn of what they held was backed by companies Walter himself controls, Bloomberg's editorial board writes. Once corrected, investments tied to the owner's own businesses were 42% of the insurers' assets, not the 3% reported. That is not a rounding error, and federal investigators are now looking. To begin getting out of those positions, Walter sold the Los Angeles Lakers for $12.5bn. Policyholders, Bloomberg notes, are "so far unharmed, if perhaps startled to learn their retirement income was in any way tied to an NBA franchise." We covered KPMG's warning about a Guggenheim unit yesterday. This is the other end of the same structure, and the structure is what our thesis is about. A saver buys an annuity, a policy that pays a monthly income for the rest of their life, and the insurer invests the money. When the insurer is owned by an investment firm, what it buys is increasingly loans that firm itself has made, lending to companies by investment funds rather than banks, and the price those loans are said to be worth is set by the people who made them. Bloomberg cites a record 46% of life insurers' debt holdings in such loans this year, and sales of annuities hit a quarterly record in the second quarter ($123.9bn). Oversight sits with fifty separate state regulators; there is no national one. Bloomberg's proposed fix is to declare parts of the annuity business important enough to the whole system that a federal regulator can set one set of standards. What is unusual here is the mislabeling, not the model. Apollo, which pairs lending by its own funds with an insurer's balance sheet, the list of what it owns and owes, is the archetype of that model, and the market is not treating the model as a problem: Apollo's stock is 6.6% off its high and down 1.1% on the week, while the listed funds that borrow money, lend it to mid-sized private companies and pass the interest to shareholders, its closest peers, sit within a point or two of theirs. The regulatory question is whether anyone can see a problem before it becomes one. A 3% that turns out to be 42% suggests not. For the odds of a crash, this is the third kind of story. Nothing got riskier this morning; these loans were sitting inside those insurers last year. What changed is our measurement of how much of this is out there and how little anyone checked. That sharpens the fragility reading rather than raising it.

Fed, Treasury and policy7 Sepmedium

The overflow tank is empty. Treasury buybacks start tomorrow.

With the cushion of spare cash at the Federal Reserve gone, every new government IOU now drains money straight out of the banks, at the moment the Treasury and the Fed are pushing interest rates in opposite directions.

The pool of spare cash that money-market funds park overnight at the Federal Reserve, America's central bank, has run dry: it held $725m on September 1 and $525m on September 2, which is effectively zero. That pool has been the cushion between the government's borrowing and the banks. With it gone, every new IOU the Treasury Department, which borrows the money the government spends, sells now has to be paid for with money that comes out of the banking system itself, and the price of the government's shortest debt is already showing it. Three more numbers from the Fed's own data show the strain. The overnight rate banks charge each other, which the Fed steers, sits at 3.63%, inside its 3.50–3.75% target range; the rate on cash borrowed overnight against bonds put up as security is 3.65%; and a three-month government IOU pays 3.76%, at the very top of the Fed's range. Here is how the cushion works. When the pool held a lot, new government borrowing could be absorbed by draining it, and the reserves banks keep at the Fed barely noticed. When it is at half a billion dollars, every dollar of new short-term debt the Treasury sells has to come out of those reserves or out of private hands, and a three-month IOU paying more than the Fed's own overnight rate is what that looks like. This is our reading of the mechanism; the Fed has said nothing about it. Into this, two arms of the government are pulling in opposite directions. The Treasury's expanded buybacks of long-dated debt take effect tomorrow, September 9: at least $4bn per operation of bonds due in ten to twenty and twenty to thirty years, doubled from $2bn, inside a $14.5bn weekly limit, running to November 4. Buying its own long bonds back props up their price and holds down the rate on them. Meanwhile Fed Governor Waller said on September 3 that he is inclined to leave rates alone at the September 15–16 meeting if inflation cooperates, which took the market's odds of a rate rise from about 63% to 50%. Holding down long-term rates while pushing short-term ones up has a name on trading desks, a twist, and here it is being done by two authorities that do not coordinate, while the president demands 1% rates, per our earlier reading. Scale check. The buyback program is $14.5bn a week. Japan's holdings of foreign securities fell $87.8bn in August, most likely short-dated US government debt sold to pay for propping up the yen. The buyer of last resort is a fraction of a single seller. For the odds of a crash, th

Hidden leverage and shadow banking7 Sephigh

Deutsche Bank sells the first losses on its data-center loans

This is the whole pattern in a single deal: a regulated bank makes the loans to AI data centers, then sells the first losses to investors who never have to publish what they are worth.

Deutsche Bank is arranging a deal to pay outside investors to take the first losses on about €2bn ($2.3bn) of loans, including loans to data centers, Bloomberg reports. Deutsche has done deals of this kind before on other loans, but never on loans that finance the building of something. It is not alone in looking at data centers. Société Générale, ING and BBVA have considered or completed such deals that include data-center loans, and BNP Paribas, Royal Bank of Canada and Toronto-Dominion have explored ones aimed squarely at AI infrastructure. Deutsche's finance chief, Raja Akram, says the bank plans new platforms for these deals so that regulators will let it set aside less capital. The bank keeps the loans. It keeps the client, the fees and the interest. It pays other investors to take on the risk that its loans go bad, while keeping the loans on its own books; the trade has a name, significant risk transfer. In practice it sells a note to an outside investor, usually a hedge fund (an investment firm that manages money for the wealthy and can bet in ways ordinary funds cannot), an insurer or a fund that lends to companies in place of banks, and that note loses value as loans in the pool go unpaid. Because the first slice of losses now belongs to someone else, the regulator lets the bank hold less capital against the book. On the bank's balance sheet, the list of what it owns and owes, risk has fallen. In the system it has not moved an inch. It has changed address, from a lender that publishes stress tests to a buyer that often pays for the note with borrowed money and reports what it is worth once a quarter at best. Data-center loans are a natural fit. They run for years, the thing pledged to the lender is a building full of chips, and repayment depends on one tenant honoring one lease for the life of the debt. A bank that wants the fees from AI without the risk of the big loss at the end has every reason to sell that risk. The terms may change, Bloomberg says, and the report does not name the buyers, say at what level of losses the investors start paying, or say how much of the €2bn is data centers rather than other projects. Those are the numbers that would tell us whether this is careful hedging or the first losses on the AI build-out being quietly moved out of sight. Either way it is a small deal. What it signals is that European banks now see data-center lending as a risk worth paying to get rid of. This is fragility, not a spark. No new loan was ma

The AI capex bubble7 Sephigh

ByteDance asked for $20bn. Banks gave it $29.6bn.

The borrowing to pay for AI has reached Chinese tech, and banks offering half again more than was asked for tells you lenders are still competing to supply it.

ByteDance is finalizing a $29.6bn loan to pay for AI chips and data centers, Caixin reports. It asked banks for $20bn. They put in more than $30bn of orders, and it took nearly all of them. A borrower getting half again what it asked for is what a hot lending market looks like. The loan runs for three years, is coordinated by Citigroup and JPMorgan, and ranks as the second-largest dollar borrowing in Asia this year. It is not yet signed. The money has a destination. ByteDance budgeted nearly 160bn yuan ($23.8bn) for AI in 2026, more than half of it for buying chips, with another 50bn yuan for computing infrastructure. Its net profit fell by more than 70% in 2025 under international accounting rules (IFRS); the company says that is partly down to how it accounts for preferred shares and options rather than to operating costs. Revenue grew about 20% at home and nearly 50% abroad. At a meeting in August the chief executive, Liang Rubo, said the company's large language models were falling further behind foreign rivals. Put those together. A company whose models are losing ground is borrowing $30bn for three years to buy hardware in the hope of catching up, and the banks are so eager that the loan grew by half between launch and close. When lenders behave like that, the amount a borrower asks for is the floor, not the ceiling. One thing cuts against the argument we have been making. This is bank debt. It sits on the books of Citi, JPMorgan and the other banks in the group, where it will be measured, backed by capital and stress-tested. That is the opposite of the American side of the build-out, where much of the borrowing runs through funds that lend in place of banks or through chip suppliers financing their own customers. Whether it stays on bank books is another matter: Deutsche Bank's data-center deal, reported the same day, shows banks learning to make AI loans and then sell the first losses on them. This adds to the pile. It is $29.6bn more borrowed money riding on AI, and that is fragility, though we had it in last week's tally so the number does not move again today. The three-year term is what to watch. It is short for infrastructure and long for a race between AI models, and it means the question of borrowing again to pay this off arrives in 2029, whoever is holding the debt by then.

The dollar, gold and reserve status7 Sepmedium

Foreign money is leaving US government bonds for US stocks

The owners of the world's benchmark asset are shifting from central banks that never have to sell to funds running on borrowed money that can be forced to. That changes how a sell-off would spread, not just who holds what.

Foreign investors now put more money into American stocks than into American bonds. Deutsche Bank says that has never happened before outside the 2008 crisis. Per CNBC, Deutsche's George Saravelos writes that foreigners' share of US government bonds has fallen from more than 50% at its peak to about 30%, while the US drew a record $600bn of net foreign money into its stocks in the year to March 2026, outweighing purchases of government bonds and the debt of government-backed mortgage agencies 'by the largest margin in history'. It is happening while the government's own finances get worse: the national debt passed $40tn last month, and the deficit is heading for $2.1tn, more than 6% of everything the economy produces. Saravelos's explanation: 'the American private balance sheet is booming — AI, record profit margins — but the public sector balance sheet keeps worsening'. In other words, foreigners want more of a stake in American companies and less of a stake in Washington's borrowing. The official holders line up with that. China's holdings fell to $633.4bn in June from $731.4bn a year earlier. Norway's national fund has proposed cutting US government bonds from 34.1% to 21.9% of its bond portfolio. Japan's holdings of foreign securities fell by a record $87.8bn in August, most likely US government bonds sold to raise the money to prop up the yen. And China's central bank bought gold for a 22nd straight month (650,000 ounces in August, its biggest purchase since 2023). The pile of US government bonds does not shrink because foreigners want less of it. Somebody has to hold every one. The patient holders are central banks and national wealth funds, who never get a demand from a lender for more cash because a bet has moved against them. When they step back, the buyer who sets the price becomes whoever can raise the money to hold the position. Right now that is hedge funds, which hold $2.6tn of these bonds and pay for them with $3tn of cash borrowed overnight against bonds put up as security, and, as we reported over the weekend, Chinese commercial banks paying 3 to 4% for short-term dollar deposits. Holders who can be forced to sell are replacing holders who cannot. Two things soften this. Data from the IMF, which lends to countries in trouble, show the dollar's share of central-bank reserves ticked up to 57% in the first quarter, so this is a retreat from bonds that take decades to come due, not from the currency. And State Street's Masahiko Loo describes J

Household credit7 Sepmedium

Car loans are going bad at a record rate. The lenders' shares are near their highs.

Households at the bottom are missing payments at rates not seen since the last recession, but the lenders bundled and sold the loans, so the losses land on investors rather than on the lenders' shares.

About one in eighteen dollars of American car loans is now at least three months behind, the worst in more than two decades of records (5.5% of balances 90 or more days past due in the second quarter, per data from the Federal Reserve's New York branch that resurfaced this week). That is a tenth below the first quarter's 5.6%, so the record has stopped climbing for now. Americans owe $1.713tn on their cars, and 3.00% of that slid into serious delinquency during the quarter, up from 2.93% a year earlier. Credit card debts written off as lost edged up to 3.70% from 3.66%. The split by borrower is the story. Fitch, one of the three firms that grade how likely a borrower is to pay you back, tracks bundles of car loans to borrowers with weak credit that were sold on to investors. Its index shows payments 60 days late easing to 5.8% at the end of June from 6.5% at the end of 2025, then getting worse again in July. Another reading puts July at a record 6.13% for those weak-credit bundles, against 0.49% for bundles of loans to solid borrowers. A Federal Reserve survey has 34% of adults earning under $25,000 failing to pay all their bills last month, against 7% of those earning $100,000 or more. Borrowers with good credit are fine. The bottom of the income ladder is at levels not seen since the last recession. Now look at what investors think. Credit Acceptance, which lends to the weakest car buyers, closed Friday 0.4% off its high and up 4.3% on the month. Ally is 2.8% off its high, Capital One 3%, Synchrony 1.3%. Only Affirm, the buy-now-pay-later lender, has sagged, 7.9% off its high after falling 6.9% in a week. Both pictures can be right. A lender's shares are worth the gap between what it charges and what it loses. If it has raised the rate on new loans and bundled the old ones together and sold slices of the income to investors, the losses land on the buyers of those slices, not on the lender. That is the same migration we track everywhere else: the missed payments show up in a trust, and the stock does not flinch. What we cannot see from here is how much of the lending from 2022 and 2023 is still on the lenders' own books. If the answer is 'a lot', the market is wrong. None of this moves the odds today. The figures are two months old and this is not new stress; the tinder was already in the pile and already in our number. It is a picture worth redrawing while the lenders' shares sit at their highs, because the one thing that would change the picture, the lo

Fed, Treasury and policy3 Sephigh

Who pays for the power lines to the data center?

Households have been quietly paying, through their electricity bills, for the power lines that serve AI data centers. States are starting to make the developers pay instead, and the biggest tenant is fighting to keep the old arrangement.

Microsoft has asked Virginia's Supreme Court to strike down rules that make data-center developers pay upfront for the transmission lines built to serve them, the FT reports. It filed the appeal quietly, months after signing a pledge organized by President Trump to shield electricity customers from exactly those costs. The filing does not say which parts of the rules it objects to. This is a fight over who pays for the grid, and until this year the answer was households. Dominion Energy, the utility for Virginia's data-center belt, asked regulators this year to recover more than $1.5bn it had spent on transmission lines. Historically that money went into customers' bills and was paid back over decades. In July the State Corporation Commission, which regulates Virginia's utilities, ordered Dominion to bill data centers directly and upfront instead. Governor Abigail Spanberger's office had urged the commission to protect customers from 'severe stranded-asset risks' and from utility 'capital biases that prioritise shareholder profits over Virginia's ratepayers'. In other words, the state was worried that households could be left paying for wires to data centers that never got built or lost their tenant. Under the old arrangement the AI build-out had a hidden subsidy. The big cloud companies ordered the power, the utility borrowed to build the wires, and households repaid the debt through their bills. If the data center is canceled or its tenant walks away, the wires are stranded and the households still pay. Under the new arrangement the developer pays first, which raises the cost of every project and cuts its return. Microsoft, which canceled a 244-acre site in Caledonia, Wisconsin, after local opposition in October, evidently prefers the old one. This matters for how fast the bubble grows. Data Centre Watch counts more than $130bn of US projects blocked or delayed in the first quarter of 2026. Household electricity prices are up 15% since the president took office, according to the EIA, the government's energy statistics agency, two months before midterm elections being fought over the cost of living. The limit on AI spending may turn out to be neither chips nor money but permission, and whether the tenants will pay for their own power lines once someone else stops. On its own this moves our number very little, and we are including it for what it says about direction. A subsidy being withdrawn makes the marginal project less worth building, which slows how

The dollar, gold and reserve status7 Sepmedium

Tokyo's reserves fell $88bn. We can guess what it sold.

The biggest seller of US government bonds in the August rout was a friendly central bank doing something it had no choice about, and that is precisely the kind of selling no speech from the Federal Reserve can talk down.

Japan's Finance Ministry said on Monday that its holdings of foreign securities fell by $87.8bn in August, the largest monthly drop it has ever recorded. It had already admitted spending ¥15.4tn ($98.6bn) to prop up the yen in the month to August 26, also a record. The two figures are close enough that analysts quoted by Bloomberg conclude Tokyo paid for the rescue by selling US government bonds, most likely ones due to be repaid soon. A ministry official would say only that intervention was "a factor" behind the fall. The mechanics are not complicated. To buy yen you have to sell dollars, and Japan keeps most of its dollars in US government bonds. So the harder Tokyo defends its currency, the more of Washington's debt lands on the market, and this happened in the same weeks that the rate the US government pays to borrow for ten years (the 10-year) reached 4.82%, its highest since 2023, and Treasury Secretary Bessent doubled the department's purchases of its own long-dated bonds to at least $4bn per operation from September 9. None of this was hostile. The United States joined the yen operation on July 31, the first time the two countries acted together on the currency since 1998. It is still one of the largest holders of US government bonds selling into a falling market. What makes it worth watching is that it feeds itself. High American interest rates weaken the yen; Japan sells US bonds to buy yen; the supply of US bonds goes up. Sales of bonds due soon do less damage to long-term borrowing costs than the headline suggests, and we do not know what mix Japan sold. But Japan's own government now pays more than 3% to borrow for ten years, the first time since 1996, which gives Japanese money one less reason to stay in dollars at all. Put that beside the rest of the ledger. Norway's sovereign fund has proposed cutting roughly $80bn from the $215bn of US government bonds it holds, though not before 2027. Chinese commercial banks are buying, paying 3–4% on the dollar deposits that fund the purchases. The Treasury Department starts buying back its own bonds today. The sellers are governments. The buyers are either borrowing to do it or are the issuer itself. That is why this sits in fragility rather than ignition: nothing caught fire, but the natural buyers are leaving.

Bond market dysfunction7 Sepmedium

Two debt offices, one problem: too many long bonds nobody wants

When both Washington and London are pulling their longest-dated bonds out of the market by hand, those bonds have lost their natural buyer, and the government is trading the risk that prices fall for the risk that it cannot borrow again when the debt comes due.

The Treasury Department, which borrows the money the US government spends, begins the active phase of buying back its own debt today, September 7, with a weekly limit of $14.5bn. From September 9 through November 4 the operations aimed at the longest bonds, the 10-to-20-year and 20-to-30-year ones, double from $2bn to at least $4bn each. Bessent calls this support for orderly trading, not a sign of strain in the public finances. A government buying back its own bonds is not odd in itself. Doubling the size mid-program is a decision, not a routine. Across the Atlantic the same problem is being attacked from two directions at once. On September 17 the Bank of England, Britain's central bank, sets the pace at which it will actively sell the government bonds it holds over the coming year. Only about £30bn of those holdings come due in that period, so keeping its balance sheet, the list of what it owns and owes, shrinking at the same pace means selling bonds into a market where Britain paid 5.89% to borrow for thirty years last week (the 30-year gilt), the most since 1998. Then on September 24 the Debt Management Office, which does the British government's borrowing, tests a "switch auction", a device last used at the turn of the century that pulls bonds of one maturity out of the market and hands back bonds of another. The office insists no bonds will actually change hands this quarter. FT Alphaville reads the test as a signal that it is getting ready to reshape what it owes, and points out who the largest unhappy holder of long British government bonds is: the Bank of England itself. What a buyback or a switch actually does is worth holding onto. The state takes debt that does not come due for twenty or thirty years out of private hands and sells shorter debt to pay for it. Investors are relieved of the risk that those bond prices keep falling. The government takes on a different one: more debt coming due, more often, to be borrowed again at whatever the short-term rate happens to be then. That works right up until the short-term rate is where the trouble is. In the US the short end is where the Federal Reserve, America's central bank, may raise rates next week, and where, according to Bloomberg, Japan was selling. For the odds of a crash this is pile, not spark. Nothing is on fire; two debt offices are showing us they know the buyers have thinned out. One caveat: the US buyback detail comes from a single market note, and we have not seen the Treasury Departm

What would change our mind

The specific, observable things that would move the number - in either direction.

Today's 11am buyback announcement: a size of $8bn or more that pulls the thirty-year borrowing rate below 5.1% and keeps it there over the following week would lower ignition; $4bn or less followed by a sell-off in long-dated bonds would raise it.

Would move the number

The Fed's September 16 decision: a rate rise over the president's public objection, or no change with a split vote, and how the thirty-year rate and the dollar react in the 48 hours after.

Would move the number

Cogent's refinancing either going through at 11% with the pari plus structure or being withdrawn, and whether the extra interest shaky companies pay compared with the government climbs above 3.2 percentage points as other 2027 debts come up for refinancing.

Would move the number

The yen continuing to climb 1% a day with the VIX above 20 and the shares of the big banks that lend to hedge funds (Goldman Sachs, Morgan Stanley) falling: the signature of hedge funds being forced to close out bets funded with cheap yen, rather than of a policy success.

Would move the number