The gap between what Italy and Germany pay to borrow for two years almost doubled on Thursday to 0.55 percentage points, the biggest one-day jump at the close since 2020. France's equivalent gap widened by as much as 0.22 percentage points, the most since 2012. German bonds rallied hard. In an ordinary selloff every country's bonds fall together. The core being bought while everyone else sells is what the euro area called fragmentation in 2011, and Barclays' Rohan Khanna said as much: the price action is "reminiscent of periods when bond market fragmentation was a major concern, such as during the European sovereign debt crisis" (Bloomberg).

The trigger was a French budget that tries to narrow the deficit by holding down spending and raising taxes, and may topple the prime minister in the attempt. The odd part is what happened to expectations for interest rates at the same time. Traders stopped fully betting on three more quarter-point rate rises from the European Central Bank; on Tuesday they had bet on at least four. Lower rate expectations normally lift bonds everywhere. On Thursday they lifted German and Dutch bonds only. The market is treating higher borrowing costs as a growth problem for the weaker countries rather than a policy choice.

Across the Channel the interest rate Britain pays to borrow for thirty years hit its highest since 1998 (6.03%), and British bank shares fell 4.1% on the FTSE 350 banks index, with NatWest down 5.2%, after Sky News reported that bank chiefs had been summoned by the chancellor ahead of the October 28 budget (Reuters).

The relief this bought the United States is real but borrowed. The Journal describes money flowing into US government bonds as a place to hide while hedge funds, the investment firms that manage money for the wealthy and can bet in ways ordinary funds cannot, closed out crowded European bets (WSJ). The ten-year rate, which touched 5.34% during the day, is 5.19% this morning; the S&P 500 is 0.8% off its high and the VIX is 15.6, which is calm. "Markets have a faint smell of a crisis in the making," said Columbia Threadneedle's Ed Al-Hussainy. Fidelity's Mike Riddell: "France has been slowly but steadily breaking. But today feels like the first day that broader financial markets have noticed."

One session cannot tell a reshuffle from a new regime. The 2012 episode ended when the European Central Bank promised to stand behind the weaker countries. Nothing in Thursday's reporting suggests it has yet been asked. Our read is that this is an ignition story, not a fragility one, and it moved sideways: the spark got closer in Europe by exactly as much as it moved away from the United States.

The interest rate to borrow the current ten-year US government note, by lending cash overnight against it, traded as low as 2.70% on Thursday and closed at 3.75%, according to Curvature Securities. The rate for borrowing any old Treasury traded at 3.86% and closed at 3.92% (Bloomberg). Normally the two sit close together. That gap is the sound of a crowd all wanting the same bond.

To bet that a bond will fall you have to sell one you do not own, which means borrowing it first. You borrow it by lending cash overnight to whoever owns it, with the bond put up as security. When many traders want to borrow one particular note, its owners can demand a cheaper cash loan in return: the rate falls, and the note is said to trade "special". A rate 1.2 percentage points below the general one means the people betting against the ten-year were paying, in effect, more than a percentage point a year for the privilege.

What makes this odd is that the note is not scarce. Roughly $92bn of it exists, the Federal Reserve, America's central bank, holds about $10.6bn of it, and the Treasury Department will sell $39bn more of it at an auction on October 7. "There's a deep short base in the market and no surprise there are a lot of shorts in the 10-year," said Curvature's Scott Skyrm, who expects the note to stay volatile for two weeks. In other words, a great many traders have bet against this bond, and he is not surprised.

Yesterday we described the forced selling of government bonds by holders of mortgage bonds as a loop that runs one way. A deep crowd of bets against the bond is the same machine wired in reverse. Thursday's rush into safety took the ten-year rate from 5.34% back to 5.19% this morning, exactly the move that hurts a crowded bet against it and forces those traders to buy, which is one reason a US bond rally built on French distress can run further than its cause deserves. A note trading special is not a cash crisis; the general rate is still in line with SOFR, the benchmark overnight rate, at 3.87%. But a market that is simultaneously being sold by forced hedgers and squeezed by forced buyers is not one where interest rates are telling you much about the economy. For the odds of a crash this is tinder rather than a spark: nothing has broken, but the world's benchmark rate is being set by who is trapped where.

The extra interest companies worldwide pay to borrow compared with governments widened by about 0.05 percentage points this week, the biggest move since March, leaving it at its widest in six months on a Bloomberg index. The equivalent gap for Asia's safest corporate borrowers opened Friday another 0.02 to 0.04 of a percentage point wider (Bloomberg). A twentieth of a point is not a rout. The reason it matters is the trade it undoes.

For the past year the fashionable view was that a company like Microsoft was a safer borrower than most governments, and company borrowing costs stayed close to government ones even as government bonds slumped. That worked while the supply of new company debt was manageable. It is not now: Paramount Skydance's $41bn of bonds sold to pay for a takeover fell as soon as they began trading, SoftBank ran a huge sale of debt too shaky to be considered safe to fund its AI push, and record new borrowing is colliding with investors who have started worrying about what companies owe. "It's that choppiness that keeps people away from the market," said T. Rowe Price's Sheldon Chan. Last week the Bank of England put AI-related borrowing at $450bn for the year to September; that supply has to be absorbed by someone, at some price.

The same selloff reached the market that is supposed to be asleep. American state and city bonds, measured by the Bloomberg Municipal Bond Index, lost about 4.4% in September, their worst month since Lehman collapsed in September 2008, with the interest rates on them at their highest since at least 2011 (Bloomberg). Thursday brought a second day of relief, the ten-year rate on those bonds falling 0.07 percentage points to about 4% and the thirty-year to 5.2%, as cash came back into the funds that hold them.

Our own numbers say the same thing more slowly: the extra interest paid by the shakiest companies stood at 3.12 percentage points on Wednesday against 2.7 in mid-September, wider every day this week. That is still less than half the 8 percentage points a crash trigger needs, and the move is orderly. But the one corner of the market that had been pricing no stress at all has begun to price some. That nudges the odds up on the fragility side: the buyers the AI borrowers need have just become choosier.

Broadcom's Wall Street bankers are about to invite lenders into a $42bn senior slice, the part that gets paid first and is backed by security, and Blackstone is leading an $18bn junior slice, the part that takes losses first, putting in $9bn from its own funds and selling the rest to others. The $60bn package, which has been taking shape for weeks, exists to let Anthropic and other companies buy Broadcom chips and data-center equipment (Bloomberg).

This is vendor financing: the seller's bankers arrange the debt that pays the seller. It is an old trick, telecoms equipment makers did it in 1999, and the question is always the same. Who eats the loss if the equipment is worth less than the loan? The reporting does not say who the borrower is, whether an Anthropic company or a vehicle that owns the chips, what the senior slice is secured on, or whether Broadcom stands behind any of it. Those three facts decide whether this is a loan to an AI lab, a loan against silicon, or a debt Broadcom may one day have to pay itself.

The structure does tell you one thing. The senior piece goes through banks to be sold on; the junior piece, the part that absorbs the first losses, sits with Blackstone's funds and whoever they sell it to. That is the migration this site exists to track: the riskiest slice of a $60bn AI loan lands outside the banking system, in vehicles whose investors will not see a price put on it until something forces one.

The same day, Bloomberg reported Anthropic aims to start marketing its stock-market listing the week of November 9 and trade before Thanksgiving, with prospective investors valuing it at $1.8tn to $2tn (Bloomberg). The prospectus that leaked earlier this week showed a $42bn net loss against $518bn of commitments to buy computing power. The lenders behind the $60bn are, in effect, betting on the listing: the cushion of shareholders' money beneath their chips is a valuation that does not yet exist. This adds to the pile. It is fragility, not a spark, and it is exactly the kind we count.

Amazon has been sounding out investors on moving about $8bn of Nvidia Grace Blackwell chips into a separate company set up to hold them off Amazon's books, which would borrow from outside investors and lease the chips back to Amazon. The chips are already running in more than a dozen US data centers across five states. Amazon would offer up to 10% of the vehicle's shares to others and keep none itself (FT).

This is selling something and renting it back, and the something loses value every year. The FT is explicit about the purpose: a balance sheet, the list of what Amazon owns and owes, that looks lighter and so protects its credit grade, alongside the promises about leftover value that tech companies have been giving lenders instead of borrowing directly. Investors expect the vehicle to be graded safe enough for the most cautious buyers, borrowing off Amazon's AA grade, which opens it to insurers and pension funds.

That grade is the tell. A box full of chips Amazon does not own cannot borrow cheaply on its own merits. It can only do so if Amazon's lease payments are, in substance, an Amazon debt. So the risk has not left Amazon; it has been renamed. The piece does not say how the lease is accounted for, how long it runs, or who bears the leftover value when Grace Blackwell is superseded by Nvidia's Vera Rubin, which the FT notes is imminent. If the vehicle's investors bear it, they are taking the risk that the technology goes out of date in exchange for the low interest a safe borrower pays. If Amazon bears it through the lease, the balance sheet is not lighter at all.

Eight billion dollars is small for a company that has committed up to $83bn to Anthropic alone. What makes it notable is who is doing it. The small AI cloud companies borrow against chips because they must. Amazon borrows against chips because it would rather not show the borrowing, and the people being invited to hold the debt are the insurers and pension funds this site has watched absorb loans from investment funds, insurer loans and data-center debt all year. The shares were down 0.4% on Wednesday; the market has not yet decided this is a statement about anything. For the odds, this is risk moving rather than risk growing, so it does not raise the number. It does make the next measurement harder.

The average thirty-year fixed mortgage rate reached 7.28% on October 1, up a quarter of a point in a week, the largest weekly rise since October 2022 and the highest level since late 2023, Freddie Mac said (FT). In late February it had dipped below 6% for the first time since 2022, days before the strikes on Iran. The round trip is the government bond selloff arriving at the kitchen table: US government bonds had their worst month in four years in September, and mortgage rates are set off them. KB Home, the builder, reported third-quarter revenue down 20% and homes delivered down 19%, and said conditions had weakened since June.

The same afternoon the housing regulator moved in the opposite direction. The FHFA, which oversees the two government-backed mortgage giants, plans to tell Fannie Mae and Freddie Mac to accept credit reports from two bureaus rather than the current three, an announcement that could come as soon as October 12 and take effect within one to three months, having already decided to treat VantageScore 4.0 as equivalent to Classic FICO, the standard credit score (Bloomberg). TransUnion fell 4.7% and Equifax 5.9% after hours. The stated aim is lower closing costs; the Biden administration shelved the same idea over concerns about judging risk accurately, and eight Senate Banking Republicans have asked Pulte, the regulator's director, to abandon it.

The mechanism is simple. Fannie and Freddie guarantee the loans and sell the bonds; the credit report is how they decide what to charge. Less data at the moment a loan is approved does not change how many borrowers fail to pay, only how well the guarantor sees them coming. Doing that at 7.28%, with Credit Acceptance, a lender to the riskiest car buyers, down 3.7% on Wednesday and 16% in a month, is a bet that the affordability problem is a fee problem. On the evidence of the market, it is a rates problem. This is fragility, slowly added: more loans made with less known about them, and the taxpayer standing behind the lot.

Some of Australia's largest retirement savings funds (superannuation funds) have been unable to take their money out of Partners Group's $12bn Global Value Fund for more than four months, the Australian Financial Review reported overnight. The Swiss manager, which runs more than US$185bn, paid out 51% of the withdrawals requested in May and none at all in June, July or August, according to investors who spoke to the paper. The reason given is worry about how the fund's unlisted holdings are performing and what they are worth. Funds like this promise investors their money back on a schedule. Four months of silence is not a delay, it is a door being held shut, and the industry word for it is a gate.

On Wednesday Metrics Credit Partners, one of Australia's biggest firms lending fund money to companies rather than bank money, froze withdrawals from two more funds and paused new lending for 90 days, a day after admitting three of its funds listed on the Australian stock exchange were worth a combined A$160m less than it had said. Metrics had already shut its $6bn fund for professional investors on Monday, after its auditor refused to sign off on the prices it put on its loans. The lending pause is the new fact, and it is the one that travels. A lender that stops making loans for a quarter has pulled away from borrowers who were counting on it to let them borrow again when their current debt comes due.

The mechanism in both cases is the same. A fund with no end date promises to let investors out at set intervals, against holdings that cannot be sold quickly. That works for as long as nobody doubts the price the fund says its holdings are worth. Once they do, everybody with a right to leave wants out at today's stated value before it is cut, and the manager's only way to avoid proving the price wrong by selling into a falling market is to stop paying out. The retirement funds, meanwhile, carry the holding in their members' balances at whatever price Partners Group gives them.

This is our running argument about risk migrating to places nobody checks, played out in one afternoon: two sets of untested prices, one at a firm that buys companies with borrowed money and one at a fund that lends to them, being tested by the people trying to leave. It does not add to the pile. It tells us how much was already there, which is a fragility story and not an ignition one. The publicly traded firms in the same business had already made up their minds. Apollo, KKR, Ares and Blue Owl fell a further 2% or so on Tuesday; Blue Owl is 26% off its high.

The interest rate the US government pays to borrow for ten years ended September at 5.3%, up more than half a percentage point in the month, the worst month for these bonds in four years, and touched 5.33% on Thursday morning, a level last seen in 2002. The rate for thirty years is 5.67%. US government bonds are a $32tn market and the reference point for the price of almost everything else, and it has, in the words of JPMorgan Asset Management's Priya Misra, had a 'vicious loop' take hold: 'No one wants to step in front of the freight train.' Rates move most months on some piece of news. What is happening here does not need news.

What makes this week different from last is that the selling now has a named mechanical engine. AllianceBernstein's head of trading, Matthew Scott, told the FT that the big forced sellers are hedge funds and listed firms that borrow to own mortgage bonds (mortgage REITs). As mortgage rates rise, homeowners stop refinancing, so the bonds backed by those mortgages pay off more slowly and in effect become longer-term debt. A holder who had protected himself against a seven-year bond now owns something closer to a ten-year one, and he fixes that by selling government IOUs that do not come due for twenty or thirty years. That pushes rates up, which stretches the mortgages further, which forces more selling. The trade has a name, convexity hedging, and it does not read inflation data: inflation, measured by PCE, came in at 3.4% on Wednesday against a 3.7% forecast, the kind of surprise that would ordinarily send bond prices up, and it did nothing.

The things that were supposed to lean against this are not. The Treasury Department, which borrows the money the government spends, raised the cap on its buyback of its own bonds in August to $6bn from $2bn, but has been taking around half of what sellers offer and falling short of the cap in each recent operation, apparently because it thinks the prices asked are too high. ING's Padhraic Garvey says it is 'perfectly entitled' to decline; Jefferies' Thomas Simons says anyone who really needed to raise cash 'will give an aggressive offer in'. Both are true, and both mean the backstop is a buyer at a price of its own choosing, not a buyer.

What it costs to be a forced seller in a market like this is visible in the bonds sold by US states and cities. Nuveen put its entire $190m holding of bankrupt Brightline's senior bonds up for an all-or-nothing sale on Monday, days after the company filed for Chapter 11, and sold at 45 cents on the dollar. Diameter, Redwood and FourSixThree bought it. Two days later the bonds traded at 55. That gap is the price of having to sell today, and it is why this is an ignition story: a selloff that forces people to sell stops only when rates stop rising, and rising rates are what it produces.

The interest rate Britain pays to borrow for thirty years went through 6% on Thursday for the first time since 1998, and the FTSE 100 fell 2%, its worst day since May. Markets now expect the Bank of England to raise rates in November and to have Bank Rate at 4.75% by next July, on the view that a prolonged energy shock from Iran will push winter bills and inflation higher. US-Iran talks have stalled; Marco Rubio ordered Iranian negotiators out of the US overnight. A government paying 6% for thirty years is not a crisis in itself. Britain has not done it for 28 years, and the line was crossed on a day with no British news to explain it.

Two things make Britain's long-term debt a particular pressure point rather than one more line on a global chart. The first is that the Bank of England's own Financial Policy Committee, the body that watches for threats to the system as a whole, said on Wednesday that heavy borrowing by hedge funds betting on British government bonds (gilts) was making a sharp correction more likely. The last time the thirty-year rate moved like this, in 2022, the forced sellers were the pension system's funds built to match their future payouts; this time the committee is pointing at the hedge funds that took their place. It also said overall vulnerabilities had grown in recent months, naming energy prices, inflation, government debt and rising rates, and noted that the AI industry's borrowing worldwide this year is set to exceed what the British government itself sells in bonds.

The second is the calendar. A Budget is weeks away and the Institute of Directors says 65% of bosses are pessimistic about the economy and more than 70% plan to freeze or cut investment, largely for fear of tax rises. A government that must raise taxes to satisfy a bond market that is at the same time demanding 6% for thirty years has very little room in either direction.

What would calm it is obvious and not within Westminster's gift: a settlement in the Gulf, or a US government bond market that stops selling itself. Neither was on offer on Thursday morning. For the odds of a crash this is the ignition side. The borrowed money the regulator has just pointed at was already in the system; the market most likely to force it out is now moving fast.

France presents its 2027 budget on Thursday paying 4.8% to borrow for ten years, the most since 2008, and 1.22 percentage points more than Germany, the widest gap in 14 years after it grew by 0.44 percentage points in September alone. France now pays more to borrow than Greece or Italy, the two countries that used to be the euro zone's problem. That is the significance in one line: Greece and Italy are the countries the euro zone spent years worrying about, and being charged more than them is the fact to hold onto.

The Economist's point is that this is not a problem of finding buyers. France 'has no difficulty issuing sovereign debt'. What the market is pricing is politics. Prime minister Sébastien Lecornu has promised €54bn of savings, from freezes on higher-rate pensions and some civil-service pay, a third year of a supposedly one-off surcharge on large companies, and squeezes on almost every department except defense. The finance minister calls that 'ambitious'. Yet even if every measure passes, the deficit stays at 5% of national output, the fifth year running at or above that level and well over the EU's 3% ceiling; this year's figure has already slipped to 5.4%. And it may not pass: the government has no majority, the lower house is fragmented, and Marine Le Pen, favorite for next spring's presidential election, says her party will vote against any measure touching pensions.

So the premium is a bet that getting the budget under control grows more remote with every parliamentary round and every election. That is a different kind of stress from the forced selling in US bonds or the energy fear in Britain: slower, more durable, and harder to reverse with one good piece of economic data.

Who holds the debt matters. A government that pays more than Italy is a weaker floor under everything in Paris whose price is set off it, including the balance sheets, the lists of what they own and owe, of the French banks and insurers that, as everywhere, hold their own government's bonds. This adds a little to fragility and nothing to ignition. The budget's reception in parliament over the coming months, not the headline savings figure, is what we would watch.

The Bank of England's Financial Policy Committee put a number on the thing we have been circling for a month: borrowing to build AI reached $450bn in the year to September, more than double all of last year, on Morgan Stanley's estimate, and the AI industry's bond sales worldwide this year are expected to exceed what the British government sells in gilts. For scale, that is a single industry out-borrowing one of the world's largest governments. The committee's conclusion was that this 'broadens the exposure of capital markets to developments in AI' and that the 'risk of a sharper correction persists', notably if expectations for profits take a knock. If the productivity gains do not arrive, it added, the damage would reach government bond markets, not just the share prices of AI companies.

The same day, Bloomberg reported what the people who actually sign data-center leases are doing about it. Brokers at CBRE, the largest commercial property services firm, are telling neoclouds, the firms that rent out access to AI chips, that they will not enter serious talks for sites unless a partner with strong credit stands behind the lease. With empty space in the main markets at a record low 1.4%, landlords can be choosy, and they do not want tenants graded as risky borrowers who owe large sums on their chips, many of them former bitcoin miners. If a neocloud folds, the guarantor takes over the lease or pays the fee to end it.

Put the two together and the $450bn gets a shape. A growing share of it can be borrowed only because a handful of companies whose balance sheets, the lists of what they own and owe, are among the strongest anywhere, Big Tech and Nvidia, have agreed to stand behind tenants the landlord would otherwise turn away. The risk is not spread across the system; it is concentrated on the guarantors and on what the chips are worth if a tenant fails, a resale price nobody has had to test. That is risk moving to where nobody puts a daily price on it, which is our running argument, and it belongs on the fragility side of the meter: nothing here is close to catching fire, but the pile is bigger than it was and we now know roughly how big.

The market has already applied CBRE's filter. IREN is down 13% in five sessions, Core Scientific 12% and Applied Digital 10%, while Nvidia and Microsoft sit within 1% of their highs. The guarantors are being priced as safe. The guaranteed are not.

On the day Partners Group's four-month freeze came to light and Metrics shut two more funds, the SEC, the agency that polices American markets, voted to make it easier for private funds to raise money from ordinary Americans. Three proposals, approved unanimously by an all-Republican commission on Wednesday, would let people into funds that buy companies with borrowed money, funds that lend to companies, and venture funds if they pass a test of financial sophistication or hold a professional license; would widen the set of funds allowed to charge fees based on performance; and would give managers of interval funds more leeway over when they buy back investors' shares. 'At its core, this is a question of freedom and fairness,' chairman Paul Atkins said. The rules can take effect after a 60-day comment period, and follow the Labor Department's March rule admitting private equity to 401(k)s. Regulators loosen rules on ordinary days. Doing it on the day two funds lock their investors in is a coincidence worth noticing.

Read the third measure slowly. Interval funds are the packaging that sells fund lending to ordinary savers: they promise to buy back a slice of shares at set intervals, which is what makes them acceptable to people who expect to be able to get their money back. Giving managers more say over the timing of those buybacks is a legal way of saying that locking investors in will be easier. Institutions in Australia are learning this week what that discretion feels like from the inside.

The buyout industry needs the money. The WSJ describes a sector struggling to attract new clients and to sell what it owns at a profit, and the listed managers tell the same story: Blackstone is 22% off its high, Blue Owl 26%, KKR 17%. Money from ordinary savers that cannot easily leave is the most patient money there is.

The Federal Reserve, America's central bank, added its own loosening on the same day, finalizing an overhaul of its stress tests of the banks that will average two years of results and, by its own estimate, cut the year-to-year swings in how much capital banks must hold by about half. None of this breaks anything today, and it does not move the odds this week. What it decides is who holds the next round of untested valuations, and the answer is: more people, with fewer exits. That is fragility being stacked for later.

The 30-year Treasury yield went through 5.59% on Tuesday, a level last seen in 2002, on its sixth straight daily rise. The 10-year traded at 5.24%, its highest since 2007. The two-year, at about 4.9%, is the last major maturity below 5%. Treasuries have lost 2.6% this year after a 6.3% gain last year.

What is new is not the level but the cause. On Tuesday afternoon John Williams, vice-chair of the Fed's rate-setting committee, said that after September's hike "there is no need for urgency" and pointed to late this year rather than October for the next move. Odds of an October rise fell from around 70% to near 50%. Crude eased on the day. The two stories we have been telling about this selloff — oil feeding inflation, the Fed chasing it — both softened, and the long end sold off anyway. Citi calls it a "light buyer's strike". Yardeni Research blames an unwind of the yen carry trade. Heavy corporate supply, led by Paramount's bond sale, weighed as well.

A selloff that continues when its stated drivers reverse is a selloff about term premium: the extra yield buyers want for holding long duration at all, whatever the Fed does next. Since the Fed hiked on 16 September the 10-year has gone from 5% to 5.25%. The long end has been tightening faster than the central bank.

The one thing pulling the other way is positioning. Open interest in 10-year futures has risen in 13 of the last 14 sessions and in the five-year in 11 of 12; asset managers added more than 100,000 10-year shorts in the week to 22 September, one of the largest weekly additions since 2023, and Bank of America says trend-followers are "firmly short". A crowd that size is fuel for a violent reversal on a soft PCE print on Wednesday or a weak jobs number on Friday. Jim Bianco has turned bullish on Treasuries for the first time since 2020. Falling yields would ease the pressure on everything downstream of them; the speed of a squeeze would hurt the people currently making money.

Some 117,419 municipal bond trades were reported on Monday, the most in a single day since at least 1995, according to data from the Municipal Securities Rulemaking Board compiled by Bloomberg. The par value traded, $21.8bn, was above the one-year average of about $14.6bn but not a record. The muni market has lost 4.7% this month, its worst month since 1987, and benchmark yields are at their highest since at least 2011. Supply this month already tops $51bn. Yields rose another 3bp on Tuesday.

The gap between a record number of trades and an unremarkable par value is the telling part. Munis are held disproportionately by households and their advisers, in small lots. A day with a record count of trades but ordinary dollar volume is a day of many small sellers, not a few dealers moving blocks.

Andrew Clinton of Clinton Investment Management gives three reasons: higher absolute yields drawing new buyers, sharp rate moves creating tax-loss harvesting opportunities — selling a bond at a loss to offset gains elsewhere — and some investors being spooked. He adds a fourth: a heavy new-issue calendar meant holders were "making room for the new issues that they purchased by selling other bonds that they own".

None of this is a default story. What it shows is transmission. A term premium built in the Treasury market is arriving in the borrowing costs of states, cities and school districts through the one bond market ordinary savers hold directly, and they are turning the portfolio over at a rate not seen in three decades to cope with it.

Clearlake Capital has raised $1bn through a collateralised fund obligation, or CFO, after reshuffling the collateral following discussions with investors, Bloomberg reports. Goldman Sachs was sole structuring and placement agent.

A CFO is a securitisation of stakes in private funds. Here a new Clearlake vehicle holds $600m of the firm's private-credit exposure, $200m of a private-equity secondaries fund and a further $200m private-equity fund stake. Against that $1bn it sold $775m of Class A, B and C bonds and $225m of equity. The Class A tranche priced at 285bp over SOFR, 15bp tighter than first marketed; with SOFR at 3.88% that is a coupon of about 6.7% for the safest slice.

The bonds are serviced by distributions from the underlying funds. The reason the deal exists is that those distributions have slowed: a prolonged M&A drought has made it hard for managers to return capital to their backers, so they are borrowing against the stakes instead. That is the circularity. The collateral is valued at marks set by the manager whose funds they are; the debt is sized at 77.5% of those marks; and the cash to pay it depends on exits that are not happening. Last week Metrics in Australia showed what happens when an auditor tests marks of this kind.

The rejig is the interesting detail. Bloomberg says the collateral mix was changed to improve pricing and cut the debt-to-equity ratio, and that a high-profile $3bn CFO was shelved last week when buyers spurned old private-equity stakes. Investors, in other words, are still buying this paper but are now dictating what goes into it. The market is open. It is no longer indiscriminate.

At $1bn this is small. As a template — leverage layered on private-fund marks to manufacture liquidity that exits used to provide — it is exactly the migration Crash Lab tracks.

Paramount Skydance is finally selling the debt for its $110bn purchase of Warner Bros. Discovery, and every month of delay has cost it. Bloomberg reports that borrowing this week rather than three months ago will add somewhere between $250m and more than $500m a year in interest, because global inflation fears have lifted rates in the meantime.

The package is about $44bn of bonds and $7.5bn of loans. After the deal closes the combined company will carry more than $87bn of investment-grade and high-yield debt, according to CreditSights, ranking it among the largest borrowers in Bloomberg's junk-bond universe. David Ellison's plan to service it rests on $6bn a year of cost cuts, to be reached within three years, at a business heavily exposed to old-fashioned television networks where revenue is hard to grow. Some investors passed on the deal citing execution risk. "If their overall debt expense is higher, that could lower their cash flow," Stephen Flynn of Bloomberg Intelligence said, which is the whole problem in one sentence.

There is a feedback loop worth naming. The $44bn bond sale was cited on Tuesday as part of the corporate supply weighing on Treasuries and pushing the 30-year to its 2002 high. A deal this size moves the risk-free rate it prices off. The delay raised yields; the sale itself raises them further; and the company pays both.

Who ends up holding this matters. The investment-grade tranche goes to insurers and bond funds; the junk and loans go to CLOs and high-yield funds already digesting a 302bp spread. A $6bn savings promise in a shrinking industry, sitting under $87bn of debt priced at two-decade-high rates, is the kind of leverage that looks fine until the first missed target.

Two $30bn AI financings surfaced on Tuesday, and neither touched a public market.

OpenAI is seeking at least $30bn in a new round at a valuation of about $1.4tn before the new money, Bloomberg reports, after pushing back its IPO. Sam Altman has said the company will not list this year, calling it an "ill-advised moment". The round is explicitly a bridge in place of an IPO; OpenAI last raised $122bn in March at $852bn including the money. Talks are early and led by investors. The day after Anthropic's prospectus laid its losses and $518bn of compute commitments in front of everyone, its rival has chosen not to.

Tesla, meanwhile, filed three credit agreements totalling $30bn: a $20bn delayed-draw term loan with Citigroup as agent, an $8bn five-year revolver and a $2bn one-year line with Wells Fargo. They replace a $5bn facility and are undrawn. The company plans to spend more than $25bn this year on factories, robotaxis and AI. "We should be spending on capex as fast as we can," Elon Musk said in July. A programme that ran on cash now has $30bn of bank commitments behind it.

On the same day a buyer pushed back. PGIM anchored a $500m CLO issued by Allstate and arranged by BNP Paribas with a first-of-its-kind cap: no more than 15% of the loans may be AI-related. AI paper is only 2–3% of the $1.4tn CLO market today, but Edwin Wilches of PGIM says the concern is that "AI financing has become ubiquitous across credit and equity markets" and losses could arrive in a downturn even where there is no explicit data-centre exposure. PGIM is working on similar deals.

Put together: the biggest AI balance sheets are raising from private investors and banks, where there is no daily price and no prospectus, while the first large institutional lender has started writing a hard limit on how much of it it will hold. The money is still flowing. The plumbing is starting to grow valves.

The Bank of Canada is now "regularly" enlarging its two-week repo operations to hold down CORRA, the Canadian overnight repo rate, deputy governor Toni Gravelle told a Bloomberg conference in New York on Tuesday. The bank "can tolerate only so much upward pressure", he said, and is supplying extra term cash ahead of quarter-end reporting dates for Canadian banks and large government bond maturities. From the first quarter of 2027 it will route repo through a new collateral management service to cut inefficiencies.

CORRA measures what banks pay to borrow overnight against government bonds and bills. It is supposed to sit on the policy rate. When it drifts above, collateral is plentiful and cash is scarce: there are more bonds looking for overnight funding than there is money willing to fund them at the target. Bloomberg notes the drift follows multi-year strains as the bank wound down quantitative tightening.

This is a footnote, and the honest reading is that the central bank is managing it competently. It is in this edition because the mechanism is the general one. When a central bank shrinks its balance sheet while the government keeps issuing, the private sector must warehouse the bonds, and it warehouses them on repo. Every basis point CORRA sits above target is a small tax on holding government debt, and a small reason to hold less of it. The same arithmetic is at work in the Treasury market, at far greater scale, where a crowded short and a buyer's strike are the visible symptoms.

What to watch is whether "regular" becomes "standing", and whether the same quarter-end pressure shows up in SOFR this week; it closed Monday at 3.88%, unremarkable.