Two days of Treasury buybacks. Yields up both days. Gold down.

Archived reading, published Thu, 10 Sep 2026 13:41:13 UTC (3 hours ago). This is not the current state of the meter.

See the live reading →

CRASH-O-METER

Cracking

how close are we

+1 since the last reading

Fragility91

how much tinder is stacked up — moves slowly

Ignition43

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 25 sources and rewrites this page.

Why it moved: Up one to 67. Ignition goes from 40 to 43. The Treasury Department's doubled buying of government IOUs that do not come due for twenty or thirty years is on its second day, and the rates it was meant to hold down rose again: the ten-year is at 4.92% and the thirty-year at 5.35%, up roughly 0.08 and 0.06 percentage points from Tuesday's close and the highest in the eight readings on our trail. The VIX, the market's gauge of how much turbulence traders expect over the next month, is 17.7, up 23% on the week; under 20 is calm, and it passed 80 in March 2020. The S&P 500 is 2.6% below its high. Gold is down 2.2% in five days while the dollar rose, so nobody is hiding in bonds or in bullion, and the policy's first public test is going the wrong way. What has not confirmed is the debt of companies too shaky to be considered safe: the extra interest they pay compared with the government is still Monday's reading of 2.67 percentage points. That is why three points and not more. Fragility holds at 91. A tenth of Blackstone's BCRED investors asked for cash against a 5% cap and 16% of Cliffwater's did the same, but those are exit limits we already carry, working as designed. SK Hynix's borrowing costs closing in on Amazon's, and the Bank of Korea's warning about hedge funds betting with borrowed money, are better measurement of AI debt and hedge-fund bets already in the number. Nothing was sold under duress.

Reporting from 7 Jul to 10 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.

Bond market dysfunction10 Sepmedium

The Treasury is buying $4bn of its own bonds a session. It is not enough.

A program announced to hold down the government's long-term borrowing costs has now run for two days. Those costs rose on both, which tells you the government is not the one setting the price.

The US government's borrowing costs rose for a second straight day this morning and now sit above every level in our last eight readings (the ten-year rate at 4.92%, the thirty-year at 5.35%). This is the second day of a Treasury program announced specifically to stop that from happening. A government stepping into its own bond market to push down what it pays, and losing on both days it tried, is not routine, and it tells you who actually sets the price. The program itself: on August 19 the Treasury Department, which borrows the money the government spends, said it would double its buybacks of bonds coming due in ten to thirty years, from $2bn to at least $4bn per operation, running from September 9 to November 4. One tally puts September's total at about $38bn (sweep). The Treasury buys old long-term bonds from the dealers who hold them and pays for them by selling new IOUs that come due in weeks. The market ends up holding less of the debt whose price falls badly when rates rise, and more of the kind that barely notices. Bessent, who runs the Treasury, has called it a 'Treasury Twist'. The twenty- and thirty-year rates fell sharply when it was announced in August. They have risen since the buying actually began. Why might $4bn a session fail to move a $28tn market? Because the people who set the price at the margin are not the Treasury. Hedge funds are financing about $3tn of government bond holdings with cash borrowed overnight against bonds put up as security, and roughly $830bn of that is one trade: borrowing heavily to profit from a small gap between two nearly identical prices (Reuters). Japan sold down $88bn of foreign securities in August to pay for a record defense of the yen. One ally's monthly selling was twice the Treasury's entire September buyback. And Waller, one of the Federal Reserve's governors, said on September 3 there is 'no more premium for safe, liquid U.S. government debt'. In other words, investors no longer pay extra to hold it because it is no longer scarce, so the interest rate that neither speeds nor slows the economy is higher than it was (sweep). Two details are worth watching. The five-year rate rose more than the ten-year today, even though the buybacks target the longest bonds, which suggests the market is repricing borrowing at every length rather than reacting to how much long debt is on offer. And gold fell 2.2% over five days while the dollar rose. That is not money fleeing America. It is money demanding more interes

The AI capex bubble10 Sephigh

Amazon now borrows at Korean chipmaker rates

When the world's most creditworthy companies borrow at the same rates as companies in developing countries, because of how much they plan to borrow, the part of the AI building spree paid for with debt gets more expensive just as it gets bigger.

SK Hynix, the Korean memory-chip maker, now borrows at almost the same rate as Amazon: its bonds due in 2031 pay just 0.09 percentage points more than Amazon's, a third of the gap a year ago. It is now cheaper to buy insurance against SK Hynix failing to pay than against several US tech giants. And for the first time on record, the extra interest that companies in developing countries pay compared with the safest US companies has closed entirely. That gap has historically run 0.3 to 0.5 percentage points. It is now a few hundredths (Bloomberg). This is not because Korea got safer. It is because US tech companies have borrowed more than $600bn around the world to pay for building AI data centers, with new borrowing expected to run into the trillions by 2030, and bond investors charge for supply. The interest rate one of the giant cloud companies pays increasingly reflects how much the buyer expects it to borrow next year, not the chance it fails to pay. Puneet Sharma of Zurich Insurance puts it plainly: developing-country debt is a 'passive beneficiary'. JPMorgan expects the gap to close fully by year-end. The same force is visible in Switzerland, where AI-related names are 26.4% of all corporate borrowing in Swiss francs this year, and Swiss finance chiefs time their own bond sales to avoid the weeks when an American giant is in the market (FT). AI debt is about a fifth of all borrowing by well-rated US companies this year, against 1% in 2024, and Goldman estimates borrowing will pay for a third of AI building spending in 2026 (Reuters). Read it alongside this morning's government bond move. Waller of the Federal Reserve says investors no longer pay extra to hold US government debt because it is no longer scarce; the corporate market is saying there is no longer a premium for America's largest companies over a Korean memory maker. Both are supply stories. Both mean the cost of paying for the buildout rises the more of it there is, which is the opposite of what made the first $300bn cheap. For the odds of a crash, this is the slow kind, and it is measurement rather than growth. Nothing here is a failure to pay. It is a price rising until enough investors are willing to hold the debt, and that price is now being set by whoever buys the fifth trillion, not the first. The pile of AI debt was already in our number. Today we learned what the market charges to hold it.

Hidden leverage and shadow banking10 Sephigh

A central bank names the hedge fund that moved its market

Bets on the AI boom are being placed with borrowed money, through contracts traded in markets whose regulators can see the price swings but not the borrowing behind them.

South Korea's central bank told its parliament on Thursday that contracts traded abroad and tied to Samsung Electronics and SK Hynix need closer watching, because their growth 'risks amplifying domestic market swings'. It named one firm: Situational Awareness, a US hedge fund focused on AI, reported to have bet with borrowed money of up to four times its own stake while building and then dumping positions in memory-chip stocks worldwide during July's selloff (Bloomberg). Central banks warn about foreign speculators all the time. Writing a single fund's name into a report to lawmakers is not something they do. The plumbing is what matters. Korea's main stock index is dominated by two chipmakers that supply the memory for AI servers, so a bet on the AI buildout can be placed as a bet on Korea, with borrowed money. Foreigners do this through instruments Korean regulators do not supervise. BlackRock's US-listed Korea fund, a quarter of which is SK Hynix, took in a record $2.8bn in a single week in July. Hong Kong lists funds that magnify the daily moves of each company. Hedge funds add borrowing from their banks on top. When the trade reverses, the selling lands in Seoul from balance sheets, the lists of what a firm owns and owes, that sit in Connecticut and Hong Kong, financed by banks whose exposure Korea cannot see. This is the migration we keep describing, in a new place. The risk is not on a Korean bank's books. It is in a hedge fund's account with the American bank that lends it money and holds its positions, in the machinery by which an exchange-traded fund creates and cancels its shares, and in structured products sold to ordinary savers in Hong Kong. A central bank has now put a fund's name into its monetary policy report because it had no other tool. What the report does not tell us is how large the remaining positions are, or who lent the money for the four-times bet. Bloomberg reported in July that Citadel bought AI stocks from Situational Awareness during the rout, which means the position moved rather than vanished. Goldman and Morgan Stanley, the two lenders to hedge funds we track, are within 1% to 2.5% of their share-price highs. The market's view is that this is Korea's problem. The Bank of Korea's view is that it is somebody else's borrowing. This does not raise the odds of a crash, and it does not lower them. Nothing was added and nothing came unwound. What changed is that a regulator said out loud a thing we already carry: an AI bet made w

Crypto and TradFi contagion10 Sephigh

Two kinds of 'tokenized stock'. Only one votes.

Digital tokens that stand for shares are starting to be pledged as security for loans in crypto markets, and whether the token is the share itself or a promise from the broker that holds the share is the whole question when everyone runs for the exit at once.

Nasdaq is putting $100mn into Payward, the parent of the crypto exchange Kraken, at a valuation of $21bn, and Kraken will offer digital tokens that stand for Nasdaq-listed shares to its customers. The tokens, Bloomberg reports, carry 'the same voting rights as regular equities' (Bloomberg). That last detail is the reason this matters, because a rival product from Robinhood does not. The same day, Robinhood's Vlad Tenev was on CNBC defending his company's tokens tied to AMC shares against AMC's chief executive, who objects to them. Tenev's tokens are built differently. They are debt, issued by a separate Robinhood company and backed by shares that company holds. Holders get no vote. Asked how Robinhood will vote the shares it holds on their behalf, Tenev said the company 'hasn't really announced plans for the voting aspect' (CNBC). The word is the same. The instruments are not. A Nasdaq token, on the description available, is a share in a new wrapper: the company knows who holds it and the holder has an owner's rights. A Robinhood token is a promise from Robinhood, with nothing pledged behind it beyond shares that Robinhood controls, votes (or does not), and could in principle lend on to somebody else. The buyer owns Robinhood's promise to pay, not a piece of AMC. Why this belongs on a page about how trouble spreads: the point of putting stocks on a blockchain is to let people pledge them to a lender, trade them around the clock, and borrow against them on crypto venues. Something pledged to a lender behaves very differently in a panic when it is a promise from a middleman than when it is the asset itself. The crypto lenders that failed in 2022 failed on exactly this distinction. Tenev's argument that companies 'don't control other companies issuing their own securities that reference those shares' is correct in law. It is also a description of how a market in imitation shares grows without the real company knowing. We do not know how Nasdaq's design carries voting rights onto a blockchain, where the underlying shares sit, or who holds them in safekeeping. Those are the questions that decide whether this is a cheaper way to keep a shareholder register or a new layer of firms who have to be good for their promises, wearing a familiar ticker. For the odds of a crash, this changes very little today, and we are including it because the design choice will matter later. It is slow tinder at most: a new kind of pledge that has never been tested in a run. Nothing h

The dollar, gold and reserve status10 Sepmedium

The story that the world is dumping the dollar is smaller than it sounds

If the selloff in US government bonds is about lenders demanding more interest for tying money up, and about heavily borrowed holders, rather than about the dollar losing its place, then both the fix and the way it fails look different from what the dollar-decline story implies.

A study by two economists at the New York Federal Reserve, Linda Goldberg and Sneha Parthasarathy, finds the dollar's share of the reserves that central banks hold fell to 56% last year from 64% a decade earlier, and that almost all of the decline is the work of a handful of countries, with China and Russia doing most of the active shifting between 2015 and 2019. In every period they looked at, roughly as many countries raised their dollar holdings as cut them. Their conclusion: 'little evidence of widespread official de-dollarization' (sweep citing Bloomberg). A fall of eight points in a decade sounds like a trend. Two countries doing most of it is a decision, not a trend. This cuts against a story we have carried, so it is worth saying plainly. This morning's markets agree with the Fed's researchers. US government bonds sold off hard, and the dollar rose 0.4% while gold fell. That is the pattern of a market demanding more interest, not a market abandoning the currency. A 'sell America' day would look like the reverse. What the study does not cover is the thing actually moving government bonds. Japan's holdings of foreign securities fell $87.8bn in August, the size of its record defense of the yen; the biggest official seller of US debt right now is an ally propping up its own currency, in coordination with Washington, not a rival leaving the system. China's central bank bought 19.9 tonnes of gold in July, its 21st straight month, but at a third the pace of its previous buying spree (Caixin). The Dutch central bank moved 86 tonnes of gold from New York to London, so it could get at it in a crisis rather than to get out of dollars. So the picture is this: a few determined diversifiers, a large ally selling for reasons that have nothing to do with the dollar's standing, and a base of private holders, hedge funds financing their bonds with cash borrowed overnight against bonds put up as security, and now foreign private investors more than foreign governments, that borrows more and flinches faster than the central banks it replaced. That is a weakness in the market for US government debt, not in the dollar's place in the world. On the odds of a crash, this one lowers them on this front, and we should say so. The risk we were carrying under 'the world dumps the dollar' is smaller than we thought, because the evidence for it is thinner. The risk in the government bond market itself is not smaller, and the bonds beat carries it. We are keeping this beat cooler

Private credit and BDCs9 Sepmedium

A tenth of Blackstone's biggest lending fund asked for cash. Half got it.

These funds cap how much money can leave each quarter, which turns what would be a forced sale into a waiting list. That is why the strain in lending by investment funds shows up in the managers' share prices rather than in the interest rates their borrowers pay.

Blackstone's Private Credit Fund, a $77bn pool that lends to companies in place of banks, told its investors it would honor withdrawals equal to 5% of shares this quarter. Requests came in at about double that (roughly 10%, or about $4.3bn). The Cliffwater Corporate Lending Fund, at $31bn, got requests for about 16% of its shares and will also pay out 5% (sweep). KKR's Income Trust I, which hit its 5% cap in an earlier quarter, saw requests fall back to 2.5% this time. One fund hitting its exit cap is not unheard of. Two of the largest hitting it in the same quarter, one of them asked for three times what it will pay, is a queue forming. The cap is working as designed, and that is the point. These funds are sold to wealthy individuals as 'semi-liquid': you can get out, but only a slice per quarter, and only if not too many others want the same slice. When requests exceed the cap, nobody is forced to sell a loan at whatever price it fetches, so no valuation is tested and no loss is made real. The strain shows up instead as a line of people waiting. The line is information. The price the Blackstone fund says its own holdings are worth fell about 4% in six months (from $24.68 a share at the end of January to $23.64 at the end of July). One in ten holders would rather have cash today than that number in a year. Reuters looked at 44 listed funds that borrow money, lend it to mid-sized private companies, and pass the interest to shareholders. Their loans are carried at 97.57% of what was lent, 3.4% have stopped paying interest at all, and 81% of their software loans have been written down this year (Reuters). The listed funds trade every day, so the market prices them every day. The unlisted ones publish a number and let the exit line do the talking. What the line does to the machine is the part that matters. A fund honoring 5% a quarter has to hold or raise that cash, so it lends less and renews fewer loans as they come due. A borrower who counted on the same lender to lend again in 2027 now faces a lender managing its own outflow. Blue Owl's OBDC cutting the value of its Loparex loan from 63 cents on the dollar to 5 cents in a single quarter is what it looks like when that new loan does not arrive. What we cannot see is how many holders who were turned away left their request in the queue rather than withdrawing it. If next quarter's requests exceed 10%, the cap has stopped being a feature and become the story. Does this move the odds of a crash? Not today. We

Signs of the times

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

Amazon borrows at Korean chipmaker rates

SK Hynix's bonds due in 2031 pay only 0.09 percentage points more than Amazon's, a third of the gap a year ago, and it is now cheaper to buy insurance against the Korean company failing to pay than against several US tech giants. Bond investors are pricing how much Amazon plans to borrow, not whether it can pay.

ai_debtspreadshierarchy

10 SepBloomberg

One state audit knocked a third off national power growth

The EIA, the government's energy statistics agency, cut its forecast for growth in US electricity demand in 2027 from 2.9% to 1.8% because Texas paused new grid connections for data centers while it runs an audit. Texas has requests for 474 gigawatts waiting, more than five times the state's record peak demand and roughly 474 nuclear reactors' worth.

powerdatacentrestexas

9 SepBloomberg

A central bank names the hedge fund

The Bank of Korea's report to parliament names a single US hedge fund, Situational Awareness, as a driver of Korea's market swings, betting with borrowed money of up to four times its own stake in memory-chip stocks, and notes that BlackRock's Korea fund took a record $2.8bn in one July week. The AI trade is now big enough to appear, by name, in another country's central bank report.

leveragekoreahedge_funds

10 SepBloomberg

Swiss finance chiefs now dodge the weeks when American tech giants borrow

AI-related companies account for 26.4% of corporate bond sales in Swiss francs this year; Amazon's SFr2.82bn deal in May was ten times the size of a typical Swiss issue and close to 2% of the entire corporate market. 'Swiss treasurers now know to avoid weeks when a hyperscaler is working the order book,' says Oddo BHF's Arthur Jurus. In other words, the locals wait until the giant has finished borrowing.

ai_debtswitzerlandcrowding_out

9 SepFinancial Times

The rumour mill

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

Confirmed

“The US government intervened in the bond market by increasing Treasury buybacks of long-term debt from $2 billion to $4 billion.”

The Treasury Department announced the doubling on August 19, taking effect September 9 through November 4, for bonds due in ten to thirty years. The buying began this week, and the government's ten- and thirty-year borrowing costs have risen on both days since.

Claimed by ITM Trading (Daniela Cambone)

Partly true

“Bessent is buying back and retiring long-dated Treasuries early, funded by borrowing at the short end, and the recent week-long decline in 20- and 30-year yields is a direct result.”

The bigger buyback, paid for by selling short-term IOUs, is real. The claim that it made long-term rates fall is not supported: those rates dropped when the plan was announced in August and have risen since the buying actually started, with the thirty-year at 5.35% this morning.

Claimed by Heresy Financial

Partly true

“China is buying more gold than it officially reports.”

Analysts at Goldman and BMO estimate that gold flows linked to China through London's private dealer market ran to roughly 88 tonnes in May and June, against 25 tonnes officially reported, and the Chinese central bank's 21-month buying streak is on record. The gap is an estimate, not a disclosure. Beijing has confirmed nothing.

Claimed by ITM Trading (Daniela Cambone)

Partly true

“The Bank of Japan announced an $87 billion yen intervention across July to September, largely financed by selling US Treasuries.”

Japan's Ministry of Finance confirmed a record ¥15.4tn (about $98bn) spent defending the yen from July 30 to August 26, and its holdings of foreign securities fell $87.8bn in August. The $87bn figure is the fall in reserves, not the size of the intervention. That US government bonds specifically were sold is the market's guess; the data do not break holdings out by type.

Claimed by Meet Kevin

Earlier developments

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

Fed, Treasury and policy9 Sepmedium

Two tests: one already in the prices, one next week

The buyback and the campaign to lift the yen are being run as two separate policies, but both draw on the same $1.1tn pile of US government bonds that Japan owns, and next week's Bank of Japan meeting decides which way that pile flows.

The Treasury Department, which borrows the money the government spends, was due at 11am Wednesday to say how many of its own bonds it would buy back on Thursday, the first time it has put a size on the enlarged program it promised on August 19. The interest rate the government pays to borrow for thirty years closed higher on the day regardless (5.29%, with the ten-year at 4.84%, both above where they stood when the promise was made). The Treasury already had a schedule of small buybacks penciled in; doubling them in a hurry, two weeks after publishing that schedule, is an attempt to hold its borrowing costs down. On the first day the tool was used, it did not hold them. The August promise was to 'at least double' the $2bn operations penciled in two weeks earlier, and this week's repurchase covers bonds that come due in ten to twenty years. Bloomberg's preview set the terms for what would count as enough. Dealers expected more than $4bn, Lou Crandall of the research firm Wrightson thought $5bn to $6bn plausible, Morgan Stanley put the practical ceiling at $10bn, and anything at $4bn 'could prove a disappointment'. The announcement itself is not in the reporting in front of us, so we cannot say what the number was. What we have is the close. The ten-year finished up 0.03 percentage points on the day. Whatever the size, the people who buy government IOUs that do not come due for twenty or thirty years did not read it as a cure. The second test is Japan's, and it is next week. Bessent, who runs the Treasury, has spent a fortnight telling Tokyo to 'do the right thing' on interest rates and claiming 'asymmetric information' on the Bank of Japan. In other words, he says he knows something about the bank's intentions that the rest of the market does not. Markets now price a quarter-point rate rise from Japan's central bank. Some investors, Bloomberg reports, have moved to expecting half a point, or two rises in a row. Seisaku Kameda, a former chief economist at the bank, calls those expectations 'excessive' and says they will probably need to be scaled back. The two tests are one test because of where Japan's money comes from. The yen is the best-performing of the ten major currencies this month, up 4%. Japan's holdings of foreign securities fell $87.8bn in August to pay for a record ¥15.4tn of buying its own currency, and Japanese economists think the most plausible way it raised that cash was by selling US government bonds. If the bank's governor, Ueda, delivers

Private credit and BDCs9 Sepmedium

Blackstone down 12% in a month. Its loans, 0.1%.

Investors have decided the trouble in lending by funds rather than banks is a problem for the firms that manage the loans, not yet for the loans themselves. Which of those two gets tested first decides how orderly the exit is.

Blackstone's shares fell 3.7% on Tuesday, are down 5.6% on the week and 12.3% over twenty trading days, and sit 13.6% below their high. The manager behind Blue Owl fell 3.2% and is 14% off its high; Ares lost 2.7% and KKR 1.6%. The funds those firms run barely moved: Blackstone Secured Lending fell 0.3%, Ares Capital 0.2%, Blue Owl Capital Corp 0.4%, and the fund that tracks the debt of companies too shaky to be considered safe fell 0.2%. That is the odd part. The firms that manage the loans are being sold as if something is wrong, and the loans themselves are priced as if nothing is. The extra interest those shaky companies pay compared with the government was 2.67 percentage points on Monday's reading, about where it has sat for a month. So the two prices are two different statements about the same loans, and both hold together. A manager earns a fee on the money it looks after. The engine of the last five years has been funds that never close and do not trade on an exchange, like Blackstone's $77bn BCRED, which grow by wealthy individuals continuously subscribing new money. That engine depends on new money arriving faster than old money leaves. Last week BCRED received requests to withdraw roughly 10% of its shares, about $4.3bn, and honored 5%, which is its cap. Cliffwater's $31bn fund received requests for 16% and paid out 5%. Every financial adviser in the country now knows the exit door is not guaranteed to open. New subscriptions do not stop, but they slow, and a fee stream that was compounding stops compounding. That is what the manager's share price is pricing. The fund itself barely moves because its value is the price the manager itself says its holdings are worth. BCRED's stated value per share went from $24.68 in January to $23.64 in July, a 4% decline. And the loans, for the most part, are still paying. Lenders see interest arriving. Shareholders see a growth story ending. For now this changes the odds of a crash very little, and we should say that plainly. Nothing was sold in a hurry, and a cap on withdrawals is doing what it was designed to do, which is to stop a run before it starts. The standoff resolves when withdrawal queues build for several quarters in a row and a manager must choose between selling loans at whatever price other investors will pay, which tests whether its stated values are real, or telling investors to keep waiting, which tests whether they keep trusting it. The first choice would move the fragility half of our meter

The AI capex bubble9 Sepmedium

474 gigawatts waiting to plug in. Texas pressed pause.

Getting connected to the grid, not raising money, is now what limits the build-out of AI data centers, and the debt already borrowed against Texas sites was priced as if it would not be.

The Energy Information Administration, the federal agency that forecasts American energy use, cut its forecast for how much electricity demand will grow in 2027 to 1.8% from the 2.9% it projected in August. The reason is a single state decision. Last month Texas ordered an audit of every proposed data center seeking to connect to its grid, a requirement Bloomberg says has 'effectively halted development'. The state's grid operator is sitting on about 474 gigawatts of connection requests, more than five times the most power Texas has ever drawn at once. One gigawatt is roughly one nuclear reactor. That number is not a forecast of what will get built. It is a queue, and queues in this business are padded, because developers file for several sites to secure one. But the region that includes Texas was expected to account for nearly 40% of the nation's growth in electricity sales in 2027, and that share was the premise behind a good deal of borrowing. The borrowing is where it bites. The site-by-site structures we have been tracking work like this: one of the giant cloud companies holds about 20% of a separate company set up to hold the data center off its own books, and lending funds and bond investors put up the other 80%. The debt is repaid out of lease payments, and the lease payments begin when the facility is switched on. Nscale is reported to have closed up to $1.85bn of loans, ranked first for repayment and backed by the property, for a campus in Ward County, Texas, last week. We do not know whether that project is caught by the audit, and we do not know who pays the interest while a switch-on date slips; that depends on lease terms nobody has published. But a grid pause moves the day the revenue starts without moving the day the interest is due, and somebody is on the wrong side of that gap. The stock market has decided it is nobody. Vistra and Talen, two power producers, are both up about 9% on the week. The smaller companies that rent out computing power for AI rallied 16% to 23% over five days before giving back 3% to 5% on Tuesday. This does not add to the pile so much as show us the pile more clearly, so it leaves fragility where it was. In one respect it cuts against our argument: the constraint here is physical, visible and sitting on a public regulator's desk, which is the opposite of a hidden risk. In another it does not. The debt the pause delays is not visible at all.

Crypto and TradFi contagion9 Sephigh

Tether, which issues digital dollars, will lend them to small businesses

It is our whole argument in miniature, lending drifting to the least-measured corner it can find, with the twist that the lender's motive is demand for its coin rather than getting paid back.

Tether, which issues USDT, the $183bn digital token meant to be worth exactly one dollar, launched a $400mn lending fund on Wednesday with London-based Fasanara Capital and is aiming to raise up to $3bn more from outside investors. The fund, called StableFund, will lend to small and medium-sized businesses 'that conventional funding channels have historically underserved', with the financing 'USDT-linked'. Fasanara, which manages $6bn, will find the borrowers and decide who is creditworthy. In June it began lending against Ferraris. A firm that started lending against sports cars in June is deciding in September which small businesses get a crypto company's money. The logic from Tether's side is simple. Its business is the gap between what it pays and what it earns on its reserves: issue a dollar token, hold the dollar in US government bonds, keep the interest. That only grows if the number of tokens in circulation grows, and USDT's total value has been flat all year. A lending fund denominated in USDT creates demand for the coin at the moment somebody borrows, and gives people chasing a return a reason to hold it. The FT frames the fund explicitly as an effort 'to boost the use of its dollar-linked cryptocurrency', which is an unusually honest description of what a credit fund is for. The timing is the interesting part. Traditional lending funds are, in the FT's words, writing down portfolios, facing defaults and gating redemptions. In other words, they are admitting loans are worth less than the books said, watching borrowers fail to pay, and refusing to let investors take all their money out. Into that market steps a lender whose borrowers are, by definition, the ones banks turned down, in an instrument whose reserves were audited in full for the first time last month. The FT does not say where the initial $400mn comes from, Tether's own profits or something else, and does not say whether reserve assets are involved. The outside money, if it arrives, sits in a fund, not behind the promise that each token is worth a dollar. So on what we know this is not a risk to USDT's backing, and it does not move our number. It is something smaller and more characteristic of the period: another pool of lending outside any regulator's line of sight, to borrowers nobody else would take, valued by the people who made the loans. That is one more twig on the pile, not a spark, and a small one.

Hidden leverage and shadow banking9 Sephigh

A country borrowed €1bn without calling it borrowing. Nobody can say what it is.

It is the first time this cycle that borrowing dressed up as a derivative has reached the point where the borrower cannot pay, and how it is sorted out will tell banks what happens to the security they hold when the same thing happens somewhere larger.

Senegal, whose debt crisis began when billions of dollars of loans turned out to have been left out of the official figures, has borrowed more than €1bn in a way nobody can now classify. International banks lent it hard currency and took Senegal's own bonds, issued in its local currency, as something valuable pledged to the lender, which they keep if the country cannot pay. The contracts are called total return swaps. On paper they count as a derivative, a side bet on a bond, rather than as a loan, which is how they stayed out of the reported debt. Now the country is heading for default, and nobody can say whether the banks are in line with the foreign bondholders or protected alongside the local ones. Last week Senegal announced a 'treatment plan' for the debt it owes abroad, alongside a draft $2.2bn rescue loan from the IMF, which lends to countries in trouble, and the price of its foreign bonds plunged. It then said it would keep paying that foreign debt for now, including an interest payment due September 13, and that debt issued at home in the CFA franc, which is pegged to the euro, is excluded from any restructuring. S&P Global, one of the three firms that grade how likely a borrower is to pay you back, cut Senegal's grade to CC on Friday and called default 'a virtual certainty'. The swaps sit across that line. What was pledged is domestic debt; the cash came from foreign banks. If they count as foreign, the banks share the losses with bondholders. If they count as domestic, the banks are protected, and the foreign creditors, as S&P warns, may refuse to be the only ones taking a loss. Senegal has not said. The FT flags one more mechanism: the swaps can trigger a demand from the lender for more cash, right now, if the interest rate on Senegal's local bonds rises and the pledged bonds lose value. So the borrowing that helped Senegal keep going may also have limited how far it could let its own bond market fall. In dollars this is small, and it does not move our number. It matters to us as a specimen. A total return swap lets a lender hold a risk that appears nowhere as a loan, and lets a borrower carry debt that appears nowhere as debt, right up until the moment somebody has to decide who gets paid first. That is the same property that runs through the structures we spend most of our time on. Here it has arrived at the decision point, in public, with a payment date attached.

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

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

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.

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

What would change our mind

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

The extra interest paid by companies too shaky to be considered safe rising above 3 percentage points, or by companies considered safe above 1 percentage point, when Tuesday's and Wednesday's data arrive. Company debt confirming what government bonds and the lending-fund managers are already saying would take ignition well past 50.

Would move the number

The thirty-year rate closing back below 5.20%, where it stood after the August announcement, through the September 24 operation aimed at the longest bonds. That would mean the Treasury's tool works, and ignition should come down.

Would move the number

Blackstone's or Cliffwater's fund reporting next-quarter withdrawal requests above 15%, or any large unlisted fund suspending withdrawals outright rather than paying everyone a slice, which would turn a queue into a forced sale that tests what the loans are really worth.

Would move the number

A hedge-fund lender or the Bank of Korea disclosing how large the remaining borrowed positions in Korean chip contracts are; if July's selling was the bulk of it, that risk is smaller than we are carrying.

Would move the number