Nvidia's guarantee covers the land and the power, not the chips

Archived reading, published Sun, 30 Aug 2026 02:16:58 UTC (22 days 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

Fragility90

how much tinder is stacked up — moves slowly

Ignition34

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 13 sources and rewrites this page.

Why it moved: Down one to 62, on ignition: this is the fifth consecutive reading on the same August 28 tape (the VIX, the market's gauge of how much turbulence traders expect over the next month, at 14.43; the extra interest risky borrowers pay over the government at 2.63 percentage points; the same gap for safer corporate borrowers at 0.79 percentage points) and no new funding channel closed in this window. The Australian fund gates are now being re-reported by the regulator rather than extended. Fragility holds at 90 because the window's structural items (the shape of Nvidia's $105bn Ohio guarantee, Google's $44bn rent guarantee for Anthropic, the Volta and Nebius debt pipeline) are sharper measurements of vendor financing and AI borrowed money we already count, not new risk stacked on this morning.

Reporting from 28 Aug to 29 Aug

Crash points

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

What’s moving the needle

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

The AI capex bubble29 Augmedium

What the $105bn guarantee actually covers

The chip vendor is now carrying the property risk of its customer's landlord. That is exactly the kind of exposure that sits outside a bank and only gets a price when someone demands payment on it.

The structure behind Nvidia's OpenAI backstop has been described in enough detail to see the plumbing. The site is PORTS-Pike in Ohio. SB Energy builds it, owns it and operates it. OpenAI takes the tenancy, eight IT-gigawatts of it. Nvidia gets exclusive rights to supply computing power at the campus and, in exchange, promises that the initial 4.25 IT-gigawatts of infrastructure will hold a certain value, with its total obligation capped at $105bn. The promise covers land, power and buildings. It does not cover the chips (TrendForce, via 新頭殼). That carve-out is the interesting part. A residual value guarantee is a promise that an asset will be worth a specific amount at a stated future date, and whoever writes it absorbs the gap if it is not. Nvidia has written one on concrete, transformers and dirt, the parts of a data center with a long life and, in principle, a buyer other than OpenAI. It has not written one on the depreciating silicon it sells, which is the part of the stack whose second-hand value nobody can currently observe. Google has done a version of the same thing for Anthropic, guaranteeing data center rent to a total of $44bn and taking roughly 20% equity in the associated power and building projects. Broadcom, which co-designs Google's custom AI chips, has separately pledged to buy back chips that do not sell. Our fact-check of the circulating $125bn figure found the number is $105bn: talks reportedly began nearer $250bn and were revised down to under $120bn before the announced deal. What must stay true for the structure to work is that the Ohio campus has a tenant at the end of the lease. Not necessarily this one.

Hidden leverage and shadow banking29 Augmedium

Worth $2.4bn, borrowing $5bn

Debt is being sized against customer contracts and depreciating chips rather than against the borrower's own cushion of equity. That is how a downturn in one tenant's demand becomes a credit event at five separate borrowers.

Volta Infra Holdings raised $300m of venture money in August at a $2.4bn valuation. On August 27, JPMorgan began early outreach to lenders on a $5bn debt package for its AI data center buildout (AI Business Weekly, citing Bloomberg). The debt being arranged is roughly twice what the equity market says the whole company is worth. That is not automatically absurd. Project debt is sized against contracted cash flows and the assets it funds, not against the sponsor's equity value. But it tells you what the underwriting now rests on. The same week: Nebius upsized a planned raise from $4.5bn to $5bn against a spending target of as much as $25bn this year, having already borrowed $775m in July secured on existing GPUs and the cash flows from a contract with a well-rated customer. Applied Digital is sounding out about $3.5bn of debt from companies too shaky to be considered safe-grade borrowers for its Delta Forge campus in Louisiana. Goldman is marketing $800m for two Novva sites in Nevada and is expected to lead a $1.085bn secured note for a CoreWeave-leased campus, issued through a financing vehicle affiliated with PowerHouse and Chirisa. Every one of those deals converts a promise from a well-rated tenant into borrowing capacity for a company that has none of its own. The lender's protection is the contract and the equipment. The contract is only as good as the tenant's willingness to keep paying for computing power it may not need. The equipment is GPUs, whose resale market has never been tested by a large forced seller. Companies have borrowed more than $410bn for data centers and AI this year.

Bond market dysfunction29 Augmedium

Three former Treasury officials cannot explain the buybacks

Official support for government bonds that do not come due for twenty or thirty years flows through the books of fifty hedge funds. That is a fragility you cannot see in the interest rate.

Nathan Sheets, Citigroup's chief economist and a former Treasury under secretary, on the decision to more than double buybacks of long-dated government bonds: "I do not think there were any meaningful signs of illiquidity in the market, and it is not clear what bar was met that triggered the expansion of the buybacks at this time. It was an unusual action. I did not find it shocking, but I found it puzzling" (Nikkei). Treasury buybacks are not unusual in themselves. Doing one at double the normal size, into a market that the Federal Reserve (America's central bank, which sets interest rates) and former Treasury officials both describe as working, is. On August 19 the Treasury Department said buybacks of government bonds maturing in ten to thirty years would rise from as much as $2bn per operation to at least $4bn, running September 9 to November 4. The interest rate the government pays to borrow for thirty years had just touched 5.31%, the highest in nineteen years. Treasury Secretary Bessent said on CNBC that "part of it is signaling" and that "we believe that the yields don't reflect the underlying fundamentals." Interest rates round-tripped within a day. On August 23 the Fed's Kashkari said the Treasury market was functioning as it should. So an intervention was launched into a market that officials on both sides describe as working. That leaves the question of whom the support actually reaches. The buyer holding most of the long-dated government bonds is not a pension fund. Federal Reserve research puts large hedge funds' total Treasury exposure at about $4tn as of September 2025, financed by roughly $3tn of cash borrowed overnight against bonds put up as security, with about $830bn of that in a trade that profits from a small gap between two nearly identical prices and 90% of the total concentrated in fifty firms. We cannot show that buybacks were designed to support that position. We can say that improving the price of older government bonds improves the economics of holding them with borrowed money, and that a policy which makes a crowded, heavily borrowed trade more profitable tends to make it larger.

Household credit29 Aughigh

Brazil's borrowers spend 26.6% of income on debt

Consumer credit stress is real but concentrated at the bottom of the score distribution, which is a social fact before it is a financial one. The market is right to price it that way for now.

Brazil's ratio of household debt payments to income, excluding mortgages, hit a record 26.6% in June, up from 25.7% at the end of last year. More than half of all credit card users (52.8 million people) are carrying revolving balances or installment loans. The central bank is now preparing measures, and the sources briefing Reuters say it will start by restricting lenders rather than adopting the IMF's suggested cap on individual debt-service ratios (Reuters). Why this belongs on a page about the US bubble: it is the same shape. Policymakers name the problem as "the widespread availability of costly credit products, weak transparency requirements and insufficient financial education." That is, the supply side finding borrowers who are already stretched, at rates that only work if nothing goes wrong. The American version is quieter and more selective. Fitch has late payments on the riskiest car loans at 6.13% in July, up from 5.80% at midyear and 5.60% a year earlier. New York Fed data show 5.5% of auto debt more than ninety days past due, the second-highest reading in more than twenty years. But FICO finds ninety-plus-day late payments on mortgages and car loans rising only in the lowest score bands and flat everywhere else, and bank card thirty-day late payments at about 2.85%, down year on year. So the US household picture in aggregate is fine and the bottom tenth is not. That is a distributional problem, not yet a systemic one. The bundles of loans that hold this paper are small relative to bank books. It becomes a market problem when the extra interest investors demand for holding those bundles moves, and it has not.

The AI capex bubble28 Aughigh

The lenders arrived. The shareholders left.

When equity investors and lenders disagree this sharply about the same borrower, the resolution is usually a refinancing that one of them does not survive.

IREN announced $2.4bn of debt from Blue Owl and Pimco at 9%, coming due two and a half years after funding, to buy Nvidia Blackwell Ultra GPUs for its Mackenzie campus in British Columbia. On the same day it forecast as much as $30bn of capital spending in the 2027 financial year. The shares fell almost 13% (Bloomberg). That combination (credit shows up, equity runs) is now the defining pattern at the heavily borrowed end of AI. Over five sessions IREN is down 15.4% and sits 21% below its recent high. Core Scientific is down 20% over twenty days and 31% off its peak. Applied Digital fell 7.6% in a day, Nebius 4.2%, CoreWeave 4.1% on the week, Talen Energy 11.4% over twenty days. Now look at what did not move. The extra interest that risky borrowers pay over the government is 2.63 percentage points and has tightened 7.4% over twenty days. HYG, the largest fund tracking that debt, is a quarter of a percent from its high. Microsoft closed at its period high, Amazon rose 4% in a session, Alphabet 1.8%. The stock market is repricing the companies that borrow to buy chips while leaving untouched both the companies that buy chips out of their own cash flow and the credit market that funds the borrowers. One of those two markets is wrong. The shareholders are junior and are being asked to fund spending programs many times their own market value. The lenders are senior, secured on the equipment, and paid 9% for a loan shorter than the useful life of what they are lending against. Our reading is that both can be right for about two more years, and only one after that.

The dollar, gold and reserve status28 Aughigh

Treasury bought yen and won't say how much

The United States is now a discretionary participant in currency and bond markets it also regulates and issues into, with no published position and no appropriation behind it.

In July the US Treasury Department sold euros held in the Exchange Stabilization Fund and bought yen, after the Japanese currency hit a 40-year low. It was the first joint intervention with Japan in fifteen years. In a one-page reply to Senator Elizabeth Warren, Treasury Secretary Bessent confirmed the mechanics and rejected the suggestion that Japan owes Treasury anything: "Treasury exchanged existing Exchange Stabilization Fund foreign-currency assets for yen. No new congressional appropriation was involved." He did not disclose how much yen was bought, at what rate, or what the position is now worth (CNBC). A Reuters photograph of his notepad on July 31 read: "Buy Japanese Yen (JPY) $5-10 bil." Strip the politics out and this is a government running an undisclosed currency position, funded by selling one reserve asset to buy another, and marking it to market. The Federal Reserve separately widened a facility so the Bank of Japan could pledge its US government bonds for cash rather than sell them. That was a quiet acknowledgment that the alternative was Japanese official selling into the long end at the same time the thirty-year rate was making nineteen-year highs. The yen has not held the gains: FXY, the yen-tracking fund, is down 0.78% over five days and 0.69% over twenty. Sheets, who supported the intervention, was clear about the limit: for a sustained stronger yen "we will need to see some adjustment, particularly by the Bank of Japan in its monetary policy." Meanwhile the slow story continues. The IMF's reserve-composition data puts the dollar at 57.13% of allocated reserves in the first quarter of 2026, against 71% in 2000. The Reserve Bank of Australia cut its dollar share from 55% to 45% in 2024/25 and lifted euros from 20% to 30%. Nobody is fleeing. Everybody is trimming.

Signs of the times

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

Worth $2.4bn, borrowing $5bn

Volta Infra Holdings raised $300m of venture funding in August at a $2.4bn valuation. Ten days later JPMorgan started calling lenders about a $5bn debt package for its data center buildout, roughly twice the whole company's equity value.

ai capexleverage

28 AugAI Business Weekly (citing Bloomberg)

Bitcoin miner plans $30bn of spending

IREN, a Sydney-listed bitcoin miner turned AI landlord whose logo appears on Golden State Warriors jerseys, forecast as much as $30bn of capital spending for the 2027 financial year on the same day it borrowed $2.4bn at 9%. Its shares fell 13% in response.

neocloudscapex

28 AugBloomberg

An AI found the counterfeiting bug

In May, a researcher working with Shielded Labs used an auditing tool built around Anthropic's Claude Opus 4.8 to find, in roughly six hours, a bug in Zcash's Orchard shielded pool that had been live since 2022 and missed by multiple human audits. In a test environment it could mint unlimited counterfeit ZEC. Our fact-check rates this confirmed.

cryptoai

29 AugCrash Lab fact-check (claim 1078)

A billion dollars a day

Nvidia reported $96.2bn of revenue for the quarter ended July 26 (about $1.07bn per day) and guided to roughly 70% revenue growth in the next fiscal year, describing even that as supply-constrained. It is also now writing residual value guarantees on its customers' buildings.

scalevendor financing

29 AugCrash Lab fact-check (claim 1070)

Record week for the scarcity trade

Gold and bitcoin ETFs together drew about $7bn over five trading days, which Bloomberg's Eric Balchunas called by far a record for a five-day period. Gold is up 12.3% over twenty days and down 3.2% over the last five.

goldflows

29 AugCrash Lab fact-check (claim 1010)

The rumour mill

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

Partly true

“Nvidia wants to give OpenAI a $125bn guarantee.”

The announced residual value guarantee is $105bn, per Reuters. The Wall Street Journal reported the two sides discussed as much as $250bn before revising Nvidia's exposure to under $120bn. The direction of the rumor is right; the number is not.

Claimed by Wealthion

Partly true

“The Chicago PMI fell from 58.3 to 47.1.”

The August reading is indeed 47.1, a 10.5-point collapse, but from 57.6 in July. The 58.3 figure was the consensus expectation, not last month's print. A ten-point one-month drop in a regional activity survey is worth noticing while the new Fed chair says conditions are not restrictive.

Claimed by Thoughtful Money (Adam Taggart)

Partly true

“The annual payroll benchmark revision came in at minus 79,000, against Goldman expecting an upward revision of 50,000 to 450,000.”

The BLS preliminary benchmark revision for March 2026 is indeed −79,000, confirmed by Reuters and Bloomberg. The Goldman range is sourced only to ZeroHedge quoting the bank, so treat the comparison as directionally right and the counterfactual as unverified.

Claimed by Thoughtful Money (Adam Taggart)

Partly true

“Treasury will use the roughly $1tn general account to fund buybacks of long-term Treasuries.”

The buyback doubling to at least $4bn per operation is confirmed and official. The Treasury General Account funding it is described by officials as something they could do, with no figure and no announced decision. Do not treat a stated possibility as a plan.

Claimed by Peter Schiff

Earlier developments

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

Private credit and BDCs29 Augmedium

The first gates go up, eleven thousand miles away

Freezing withdrawals is how a fund that holds hard-to-sell assets discovers that its promise of easy access was a marketing document. Australia just ran the experiment first.

On August 27, MA Financial told investors it would limit withdrawals to 1% of the money it manages per month across its affected private credit products. On August 28, CVS Lane went further and "determined to temporarily suspend the processing of application and redemption requests across the funds." No money in, no money out. Australia's market regulator, ASIC, described this as the "first significant cracks" in Australian private credit, tied to the collapse of several large borrowers and a growing number of non-bank lenders running into cash constraints (JD Supra summary). The mechanism is the oldest one in finance and it has nothing to do with whether the loans are good. These are funds that let investors pull money out monthly, but the loans they hold take three to five years to pay back. When more investors want their money back than the fund has cash coming in from borrowers, the manager has two choices: sell loans (which establishes a price and forces every remaining loan to be valued against it) or freeze withdrawals (which protects the valuations). Everyone freezes. Freezing is not evidence that the loans are bad. It is evidence that the promise of easy access was never backed by enough cash. The scale here is small: Australian retail funds against a roughly $2.1 trillion global private lending market. What makes it worth your attention is that the same pressure is measurable in the US, where withdrawal requests at funds that do not trade on an exchange hit a record 12.4% of the funds' stated value in the second quarter, up from 10.4% in the first, and where loans that have stopped paying interest at the largest listed lending funds have risen to between 2.4% and 3.6% of their books. No US fund of this kind has frozen withdrawals. That is the line to watch, and it has not been crossed.

Fed, Treasury and policy28 Aughigh

The examiners are told to look at less

Our running thesis is that risk migrates to wherever nobody is measuring it. This makes the measured zone smaller by rule change rather than by migration.

The OCC, the regulator that supervises national banks, and the FDIC announced a new supervisory framework on Thursday, built around "refocusing supervision on what matters most — material financial risk." It creates a new definition of "unsafe or unsound practice", previously open to interpretation, and sets a higher bar for issuing formal notices that tell banks to fix deficiencies. Reputational risk was already removed from the framework earlier (FT). The case for it is not stupid, and the regulators make it themselves. Silicon Valley Bank received several notices about governance, controls and anti-money-laundering, and none that fixed the interest-rate and cash-flow risk that killed it. "The notices to Silicon Valley should have been about the actual risks to the bank rather than a random assortment," one regulator involved told the FT. Fewer, better-aimed notices is a coherent theory of supervision. The problem is one of timing and perimeter. Bank examination is the one place in the American financial system where somebody with legal power to demand answers looks at a loan book and asks what it is really worth. The volume of lending inside that perimeter has been shrinking for a decade; the growth is in direct lending funds, insurance company holdings and separate companies set up to hold data-center debt off the parent's books, where the examiner does not go at all. Narrowing what gets flagged inside the perimeter, at the moment the risk outside it is being tested for the first time, means the measured share of the system gets smaller in both directions at once. The filings do not say how many formal notices the new bar would have eliminated. That number, when it appears, is the one worth having.

The AI capex bubble29 Augmedium

Meta's next data center may never touch its books

If the biggest building boom in a generation is financed through separate companies that keep the debt off the builder's books, the borrowed money in the system cannot be read from the borrowers' accounts.

Meta is reported to be preparing a $12 billion financing backed by BlackRock, run through a separate company set up for the purpose, for a nearly 1-gigawatt Texas data center. The structure is larger than its record 2025 deal (Stocktwits/Perplexity sweep). It sits alongside Meta's Hyperion joint venture with Blue Owl, which issued $27 billion of debt against Meta's long-term lease commitment, and alongside a separate $6.6 billion borrowing package for OpenAI's chip purchases, where the chips themselves serve as the security and Nvidia provides a backstop. Here is the plumbing. The separate company owns the building and the debt. Meta signs a lease. Bondholders are repaid out of the lease payments, so their real exposure is Meta's promise to pay rent, but the debt belongs to the separate company, and the lease appears in Meta's disclosures as an obligation rather than as borrowing. Everything works so long as the tenant keeps paying and the building keeps being worth something at the end. If the tenant stops, the bondholder owns a very large, very specialized building in Texas. The total is what makes this a system question rather than a corporate-finance one. Companies have borrowed over $400 billion globally for AI investment in 2026 alone, per Bloomberg-compiled data, and Nvidia has partnered with Wall Street firms on financing platforms intended to mobilize $500 billion of outside capital. The Bank for International Settlements, the central bankers' own bank, calls the leases-and-separate-companies portion "shadow borrowing." Two cautions. The $12 billion figure comes from a single collection of press reports, not a filing. And the accounting treatment (whether and how the separate company gets folded into Meta's books) is not something we have seen documented. Both matter, and neither is public yet.

Bond market dysfunction28 Aughigh

Japan's interest bill just rose 27% in one budget

The largest foreign buyer of US government bonds that do not come due for decades now has a rising interest bill at home and a domestic bond market that finally pays a meaningful return.

Japan's Ministry of Finance published its fiscal year 2027 budget request on Friday. Debt service comes to ¥36.64 trillion, up ¥5.36 trillion on the current year's initial budget. The interesting line is inside it: interest and discount charges of ¥16.59 trillion, up ¥3.55 trillion, roughly 27% in a single budget year (MOF). That is what returning to normal interest rates costs when you have the largest government debt stock in the developed world. Japan's ten-year interest rate has gone from 0.4% to near 3%, a thirty-year high, and each year a slice of the outstanding debt rolls over at the new rate. The bill compounds from here whether or not rates rise further. The part that reaches across the Pacific: Japan has spent ¥27.13 trillion (about $170 billion) supporting the yen this year, the largest annual currency intervention on record, including ¥15.39 trillion between July 30 and August 26. To buy yen you need dollars, and Japan raises dollars in part by running down its holdings of US government bonds, now about $1 trillion as of June, down 16% from a $1.25 trillion peak in 2021 (Nikkei). So the biggest foreign government buyer of long-dated US bonds is now a seller, for domestic currency reasons, while its own domestic bonds pay 3% for the first time in a generation. Japanese investors turned net sellers of ¥2.8 trillion of foreign bonds and stocks in a single week after the intervention, though August as a whole was still net buying. "It's too early to say the trend is shifting," NLI Research's Shingo Ide told Nikkei. Correct, but the direction of the incentive is no longer ambiguous.

Crypto and TradFi contagion28 Augmedium

The stocks fell 7%. The coin did not.

A company whose entire business model is holding bitcoin and issuing shares at a premium to the value of those holdings has to find another way to service its debts once that premium disappears.

On Thursday's close, Strategy fell 7.35%, BitMine 7.18%, SharpLink 7.44%, Galaxy Digital 8.22% and Coinbase 6.43%. Bitcoin is at $78,202, up 1.0% on the day and down 0.97% over five sessions. Over twenty days the coin is up 20.6%, and Strategy is up 36.5%, BitMine 37.7%, SharpLink 33.9%. That gap is the entire business model of a company that exists to hold digital coins, expressed as a price. These companies are listed pots of coins plus a premium to the value of those coins. The premium is not decoration; it is the engine. Issue shares above the value of the underlying holdings, use the proceeds to buy more coins, and the coins per share goes up without anyone taking a risk. It works in exactly one direction. When the premium shrinks, the stock falls without the coin moving at all, and the moment the shares trade at or below the value of the holdings, the share-issuance channel closes, because issuing then destroys coins per share rather than creating it. What is left is debt. The FT reports $72 billion of bonds that pay no interest and convert into shares, issued globally this year, a market Michael Saylor's Strategy did much to popularize, sold to buyers who want volatility rather than income (FT). We do not have current premium-to-value figures for any of these companies, and one session is not a trend (over five days the stocks still beat the coin). But the shape of the risk is clear enough: the stock is a leveraged bet whose leverage is highest precisely when the coin stops going up.

Household credit28 Aughigh

Three million people, thirty-eight points

Lending to borrowers with the weakest credit is the one part of the machine already in visible deterioration, and the companies that do the lending are priced as if the problem is over.

Fitch, one of the three firms that grade how likely a borrower is to pay you back, puts US subprime auto loan late payments at 6.13% in July, up from 5.80% at midyear and 5.60% a year earlier. The improvement from tax-refund season has fully faded (AutoRemarketing). New York Fed data has 5.5% of all US auto debt more than ninety days late, the second-highest reading in more than twenty years. Separately, FICO reports about 3.2 million consumers with a recent student-loan late payment, carrying an average 38-point year-on-year credit score decline, while consistent payers gained 6 points and those who were never late gained 16 (Equifax/Moody's). The 38 points are the mechanism, not the headline. A late payment is a loss to one lender once. A 38-point score drop reprices that borrower's next car loan, next credit card limit and next mortgage application for years. It converts a single missed payment into a permanently higher cost of borrowing for three million households, which raises the odds of the next missed payment. FICO's own breakdown shows late payments rising past ninety days only in the lowest score bands and flat everywhere above. The lenders' shares are not remotely bothered. Capital One is up 3.5% over twenty days and 4.8% off its high, Credit Acceptance up 4.4%, Synchrony up 3.5%, Affirm 1% off its high, Oportun up 33%. Only Ally is soft, down 2.6% and 6.7% off its high. The market's reading is that this is a contained, low-score, already-accounted-for problem. That has been the correct call for two years. It stops being correct when the deterioration moves up a score band, and nothing in this week's data says it has.

The AI capex bubble29 Augmedium

The rating grades the lease. The buyer prices the shed.

If investment-grade data-centre paper needs 7%+ to clear, the credit index is no longer measuring where the AI buildout's borrowing costs actually are.

QTS Realty Trust sold $3.9bn of bonds to fund a Georgia data centre tied to Microsoft. The bonds carried high-grade ratings. They cleared at a yield of about 7.23%. Hold those two facts together. The investment-grade index option-adjusted spread was 79 basis points on 27 August, close to the tights of this cycle, and the ten-year Treasury was 4.67%. A generic investment-grade bond therefore yields somewhere in the fives. This one paid seven and a quarter. BlackRock's July financing for a Texas data-centre project, also high grade, went at about 7.53%. The gap is the part worth understanding. In these structures the rating largely reflects the lease: a long contract from a creditworthy tenant, secured on a specific building. The agency is grading a covenant. The buyer is pricing something else — what the asset is worth if the tenant does not renew. A data hall is a shed with a substation, a cooling plant and a single-purpose fit-out built for one customer's rack density. The residual question is not whether Microsoft pays. It is who leases it afterwards. The pipeline behind it keeps loading. Applied Digital is sounding out roughly $3.5bn of high yield for its Delta Forge campus in Louisiana; Goldman is marketing $800m for two Novva sites in Nevada; $1.085bn of secured notes for a CoreWeave-leased campus are expected to be issued through a financing vehicle affiliated with PowerHouse Data Centers and Chirisa Technology Parks. That is this site's thesis in one line. Risk is leaving the place where it is aggregated and priced in public, for deals rated one lease at a time. The index says 79 basis points. The deals say seven per cent. Both numbers are real, and only one of them shows up on a spread chart.

Bond market dysfunction29 Augmedium

A former Treasury official calls the buybacks puzzling

An official bid supporting the long end matters most to the levered funds holding it, which is a fragility you cannot see in the yield itself.

Nathan Sheets, now Citi's global chief economist and previously Treasury under secretary for international affairs, told Nikkei that he saw no meaningful signs of illiquidity in the Treasury market and that it was "not clear what bar was met" to trigger the expanded buybacks. "I did not find it shocking, but I found it puzzling." He added that the biggest driver of higher long rates is fiscal policy and the volume of issuance. The action he is describing: on 19 August the Treasury said it would at least double liquidity-support buybacks of 10-to-20 and 20-to-30 year nominal coupons, from as much as $2bn to at least $4bn per operation, running from early September to 4 November. It followed the 30-year touching about 5.30–5.34%, a level last seen in 2007. Yields fell on the announcement and round-tripped within a day. What makes the intervention interesting is who is on the other side of the long end. Federal Reserve research covering September 2025 puts large hedge funds' gross Treasury exposure at about $4tn — $2.4tn long, $1.6tn short — equal to roughly 8.5% of privately held Treasuries, up from about 4.5% in early 2023. Cash-futures basis trades were around $830bn, financed by roughly $3tn of repo borrowing, with about 90% of it concentrated in the fifty largest funds. So the marginal holder of long-dated Treasuries is a levered relative-value book funded overnight. A buyback puts an official bid under off-the-run bonds — the leg those books are long. Whether that is the intent, the filings do not say. But it is the effect, and it means the Treasury's liquidity programme and the basis trade's survival are now pointed the same way.

The dollar, gold and reserve status28 Aughigh

Japan spent $96.5bn in a month. It has to sell Treasuries to do it.

The largest foreign holder of Treasuries is now a periodic seller of them to defend its own currency, at exactly the moment the US long end is struggling to clear.

Japan's Ministry of Finance disclosed that it spent ¥15.39tn ($96.5bn) buying yen between 30 July and 26 August — the largest single intervention round on record. For 2026 the total is now ¥27.13tn ($170bn), the biggest annual figure ever, eclipsing the whole of 2024's ¥15.3tn. The mechanism is circular and worth stating plainly. To buy yen you must sell dollars, and Japan's dollars are largely held as US Treasuries. Japan's holdings stood at about $1tn in June, down 16% from a $1.25tn peak in 2021. Selling Treasuries pushes US yields up. Higher US yields widen the rate gap against Japan. A wider gap pushes the yen down again — which is what the intervention was for. Nikkei notes the yen has fallen nearly 30% against the dollar in five years, and that the 10-year JGB yield has climbed to a 30-year high near 3% without stopping it. Japanese private money is only partly cooperating. Ministry data showed investors net-selling ¥2.8tn of foreign bonds and equities in one week after two weeks of net buying above ¥2.5tn each. But the first three weeks of August were still net purchases of ¥2.2tn, on track for the largest monthly buying in fourteen months. "It's too early to say the trend is shifting," one strategist told Nikkei. One further detail, from a single secondary sweep and so held loosely: the yen-support operation was reportedly funded with euros rather than dollars, and the Fed widened a facility so the Bank of Japan could pledge its Treasuries for cash rather than sell them. If accurate, that is the tell — the plumbing was rearranged specifically so Japan would not have to sell into a weak long end.

Hidden leverage and shadow banking28 Aughigh

Borrow at zero, provided your shares are unstable enough

Zero-coupon converts move the downside protection out of the contract and into the assumption that the share price keeps moving.

Companies have issued $72bn of zero-coupon convertible bonds so far this year, according to Dealogic figures reported by the FT, against a full-year record of $73bn set in 2025. Total convertible issuance is $147bn, more than 50% above the same point last year. Bloomberg reports that a measure of how equity-like the convertible market has become is at its highest since 2021 — the year Peloton and Beyond Meat sold billions of 0% notes. A convertible pays you twice: a coupon, and an embedded call option on the shares. Take the coupon to zero and the whole return is the option. What disappears is the bond floor — the stream of interest payments that used to cushion the holder when the equity fell. What is left is an unsecured, non-interest-bearing claim on a company that may be burning cash, plus a call struck above today's price. The pricing logic is what makes it dangerous. Options are worth more when the underlying is volatile, so the more erratic the issuer's shares, the less it needs to pay in cash. "The companies coming to this market now are far more volatile than the ones they replaced," Mirabaud's Nicolas Crémieux told the FT. "If you're selling something more valuable, you can offer less for it." The market is therefore cheapest for exactly the issuers most likely to gap. And because so much of the issuance is AI-linked — the FT names ON Semiconductor's $1.3bn and Ciena's $2.9bn, both at zero — buyers end up with many separate positions expressing one theme. Bloomberg quotes the warning directly: a portfolio can be heavily exposed to the same trade without that being obvious from any single line item.

Fed, Treasury and policy28 Aughigh

Hiking bias meets a 47.1 print

September is now a live meeting, and the rate is the single variable that reprices private credit, AI project debt and the long end simultaneously.

Fed Chairman Kevin Warsh told Jackson Hole that this summer's PCE and CPI readings "do not tell me that underlying trends have meaningfully improved" and that financial conditions are not currently restrictive. Two-year yields rose as much as nine basis points to 4.32% and September hike odds moved above 50% from around 36%. The FOMC meets 15–16 September; August CPI lands on the 11th. Set against that, two data points that came through our fact-checking desk this window. The Chicago PMI printed 47.1 in August, down from 57.6 in July — a 10.5-point fall against expectations near 58, and the claim checks out with the correct prior month. And the preliminary annual benchmark revision to March 2026 payrolls came in at –79,000, where Goldman had looked for an upward revision of between 50,000 and 450,000. The revision itself is small in absolute terms; the direction of the surprise is not. One regional survey is one survey, and consumer credit data is mixed rather than collapsing. But the shape matters for everything else on this site. The policy rate is the input to floating-rate private credit coupons, to the 9% GPU loans neoclouds are signing, to subprime auto affordability, and to the term premium at the long end that Treasury is buying back bonds to contain. A central bank tightening into decelerating activity data is the standard geometry of a policy error, and it is one of the few sparks that can be lit on a specific date. A related correction: one channel told viewers markets still price about 40% for September. They did before the speech. They do not now.

Household credit28 Aughigh

Brazil reaches for the lever the Fed doesn't use

When a central bank starts rationing credit by rule rather than by price, it is telling you the rate path is not going to rescue the marginal borrower.

Brazil's ratio of household debt service to income, excluding mortgages, hit a record 26.6% in June, up from 25.7% at the end of last year. More than half of the country's credit card users — 52.8 million people — carry revolving balances or instalment loans. Reuters reports that the central bank is preparing measures aimed at high-cost household debt, and that it plans to start by putting limits on lenders rather than adopting the IMF's recommended cap on individual borrowers' debt-service ratios. This is macroprudential policy substituting for monetary policy: instead of changing the price of credit for everyone, the authorities restrict who is allowed to be sold what. It is the tool you reach for when the policy rate is already doing other work, and it is worth watching because the same logic will present itself elsewhere. The US comparison is instructive because the stress there is sorted by borrower rather than by product. Fitch has subprime auto delinquencies back up to 6.13% in July after falling to 5.80% at midyear, and New York Fed data has 5.5% of auto debt 90-plus days past due, the second-highest reading in more than twenty years. Meanwhile commercial-bank credit card 30-day delinquencies fell to 2.85% from 3.04% a year earlier, and 60-plus-day delinquencies across all cards fell to 2.69% from 2.87%. FICO finds 90-plus-day delinquency rising only in the lowest score bands and flat everywhere else. There is no aggregate household credit crisis in either country. There is a large, clearly delineated group of borrowers for whom the current level of rates is not survivable, and one central bank has now decided that is a regulatory problem rather than a rates problem.

The AI capex bubble28 Aughigh

Microsoft rents the chips, Lambda owes the money

The strongest balance sheets in the AI buildout are consuming the compute while the weakest are carrying the debt against it.

Lambda has raised about $1bn of private, short-dated debt, arranged by JPMorgan and placed with private-placement investors. The proceeds buy Nvidia GPUs. The GPUs will be leased by Microsoft. It is the company's second such deal in a month, following a $926m loan that closed in early August (Bloomberg). Look at where each piece of risk ends up. Microsoft gets the compute without the capex line, without the debt, and without owning an asset whose useful life is the single most contested number in this cycle. Lambda — an Nvidia-backed company currently in talks to raise up to $3bn before a possible IPO — owns the hardware and owes the money. Nvidia sells the chips and holds equity in the buyer. What has to stay true for this to work: Microsoft keeps paying, and at maturity the GPUs are either still under contract or still worth something. The debt is short-dated; the assets depreciate on a schedule nobody agrees on. If the lease term is shorter than the debt, Lambda is refinancing into whatever the GPU resale market looks like then. The reporting does not say how long Microsoft is committed for, or whether the lease is cancellable. That is the number that matters and it is not public. This is the migration thesis in one transaction. The compute sits with the most creditworthy balance sheet in the world. The financing sits in a private placement, on a private company, at a price no index publishes. Bloomberg-compiled data puts global AI-related debt issuance above $400bn in 2026 alone.

Private credit and BDCs29 Augmedium

Marked at 63 cents, still counted as performing

A loan marked at 63 that still counts as performing is the exact place where private credit's losses are allowed to sit unrecognised.

Two 2021-vintage software loans — to Cornerstone OnDemand and to Symplr — are carried by their lenders in the 63–65 range and are not on non-accrual (JD Supra). Both things can be true at once. The mark says the lender thinks it gets back roughly two-thirds. The accrual status says the borrower is still making its interest payment, which in this vintage often means paying in kind — adding to the principal rather than sending cash. That gap is why the headline non-accrual series moves so slowly. Morningstar DBRS puts average BDC non-accruals at 3.4% of portfolios in Q2 2026, up from 3.1% in Q4 2025: a gentle line. Jefferies' platform numbers are steeper — Blackstone Secured Lending at 3.6% against 0.3% in Q1 2025, Golub at 2.9%, Blue Owl at 2.8%, Ares at 2.4%. The marks are the faster instrument. One fund actually tested one. BlackRock TCP Capital sold $523m of loans, cut leverage from 1.38x to 0.4x, and absorbed a 10.4% hit to net asset value — $0.68 against a $6.58 June NAV. That is a real price, discovered by an actual sale, and the fund is now far less levered than it was. On our reading, that is the rare item that reduces fragility rather than adding to it. The share prices disagree with all of it. Over twenty sessions Blue Owl's manager is up 19.1%, FS KKR 15.9%, Ares Capital 6.3%, Blackstone Secured Lending 7.1%. Rising non-accruals, rising unit prices. One of those is wrong.

Fed, Treasury and policy28 Aughigh

Regulators narrow what counts as a risk

The measurement apparatus for bank risk is being re-scoped at the moment most of the risk sits one counterparty away from the bank.

The OCC and the FDIC announced new supervisory rules on Thursday built around a single phrase: "refocusing supervision on what matters most — material financial risk." The package issues a new definition of "unsafe or unsound practice," sets a higher bar for Matters Requiring Attention — the formal notices that tell a bank to fix something — and follows the earlier removal of "reputational risk" from the framework (Financial Times). The stated rationale is Silicon Valley Bank. Its supervisors issued notices on governance, controls and anti-money-laundering while its duration and liquidity risk went unaddressed. A regulator involved told the FT the notices "should have been about the actual risks to the bank rather than a random assortment." On the merits, that is a fair reading of what went wrong in 2023. The question is what counts as material now. The interesting bank risk in this cycle is not on the bank's own books. It is the warehouse line to the private credit fund, the subscription facility to the buyout vehicle, the repo book financing hedge funds' $3trn of Treasury borrowing, JPMorgan arranging GPU paper it then places privately. Those exposures are financial and they are material — a narrowed lens could sharpen the focus on exactly them, or it could shrink the field of view to what sits on the balance sheet. Nothing about this changes any bank's leverage today, so it does not move our fragility number. It changes who is looking, and at what. We will know which reading was right only after something breaks.

Bond market dysfunction28 Aughigh

Japan's interest bill jumps 27% in one budget

The 30-year Treasury's problem is not a single bad auction; it is that its largest structural buyer now has a 3% alternative at home and a rising interest bill.

Japan's Ministry of Finance published its FY2027 budget request on Thursday. Debt service comes to ¥36.64tn, up ¥5.36tn on the current year. Inside that, the interest-and-discount line goes from ¥13.04tn to ¥16.59tn — a rise of ¥3.55tn, or about 27%, in a single budget (MOF). That is what normalisation costs. Japan's ten-year yield has gone from 0.4% to near 3%, a thirty-year high, on a debt stock built when money was free. The bill arrives with a lag and then arrives all at once. The part that matters outside Japan is the funding loop. Tokyo has spent ¥27.13tn — about $170bn — buying yen this year, the largest annual intervention on record, including ¥15.39tn between 30 July and 26 August (Nikkei). Buying yen requires selling dollars, and Japan's Treasury holdings stood at roughly $1trn in June, down 16% from the $1.25trn peak in 2021. Meanwhile Japanese private investors turned net sellers of ¥2.8tn of foreign bonds and equities last week, after two consecutive weeks of buying above ¥2.5tn each — though August is still on track for the largest net buying in fourteen months. So: a domestic bond that now pays 3%, a budget that must fund it, an official sector selling Treasuries to defend the currency, and a US Treasury doubling long-end buybacks to at least $4bn per operation from 9 September. The largest foreign owner of American duration has reasons to own less of it. None of this is a crisis. It is a slow change in who the marginal buyer is.

Crypto and TradFi contagion28 Aughigh

The coin didn't move. Its holders fell seven percent.

Digital-asset treasury companies are a leveraged bet on monetary policy expectations, not on the coins they hold, and their equity prices say so before their balance sheets do.

On Friday's close, Strategy fell 7.35%, BitMine 7.18%, SharpLink 7.44%, Galaxy Digital 8.22% and Coinbase 6.43%. Bitcoin over the comparable stretch did essentially nothing: down 1.6% over five sessions, up 0.35% on the latest print, and still up 19.8% over twenty. That gap is the whole business model of a digital-asset treasury company, visible in one session. These firms are not a wrapper on the coin. They are a levered claim on the coin plus an option to issue equity above the value of what they hold and buy more. When sentiment is good the premium expands faster than the asset — Strategy is up 36.5% over twenty sessions against bitcoin's 19.8%, roughly twice the move. When it contracts, it contracts the same way, and the underlying does not have to fall at all. What compressed it was not crypto news. Gold fell 1.73% on the day and 3.22% over five; silver fell 4.37%. Kevin Warsh told Jackson Hole that inflation is not meaningfully slowing and that financial conditions are not restrictive, and September hike odds moved above 50% (Bloomberg). Everything that trades as a bet against the dollar's purchasing power repriced together. The treasury companies just have more leverage on that bet than the metal does. No funding stress is visible. The channel these firms use — zero-coupon convertibles, a market Michael Saylor did much to popularise — is running at $72bn of issuance this year, near its full-year record. The premium is what is trading. It is also what has to hold for the next issue to work.

Household credit28 Augmedium

The average borrower is fine. 3.2 million are not.

Consumer credit aggregates are averages over a bifurcated population, and the lending indices show underwriting still loosening into the weaker half.

FICO reports that about 3.2 million consumers with a student loan payment due recorded a recent delinquency, and that those borrowers saw an average 38-point drop in their credit score year-on-year. Borrowers who paid consistently gained 6 points; those with no recent delinquency gained 16. FICO also finds that 90-plus-day delinquency on mortgages and auto loans rose only in the lowest score bands and stayed flat everywhere else. Read those two findings together and the aggregate series stop being confusing. Commercial-bank credit card 30-day delinquency was 2.85% in Q2, down from 3.04% a year earlier — improving. Fitch puts subprime auto delinquency at 6.13% in July, against 5.80% at midyear and 5.60% a year before; prime auto is 0.49%. The New York Fed has 5.5% of all auto debt 90-plus days late, the second-highest reading in over twenty years. The average is fine because the average is taken over a population that has split into two. The mechanism worth watching is the score itself. A 38-point drop does not cause a default; it causes repricing. Those 3.2 million borrowers now qualify for less credit, at worse terms, on their next car or card. That shows up as weaker demand and tighter supply, several quarters out, not as a credit event now. Supply has not tightened yet. Cox Automotive's Dealertrack credit availability index was 105 in July, up 6.9% year-on-year and near its loosest in eleven years. Lenders are still leaning in. The lenders' own shares are unbothered: Capital One down 1.05% over five sessions, Synchrony down 1.76%, Ally down 1.76%, Credit Acceptance down 1.0%.

What would change our mind

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

The extra interest risky borrowers pay over the government moving through 3.50 percentage points, or HYG breaking more than 2% below its recent high. The credit market has not confirmed a single thing the stock market has said about the heavily borrowed AI companies.

Would move the number

A US fund that lends to mid-sized private companies, or a fund that limits redemptions to set windows, blocking withdrawals the way CVS Lane and MA Financial have in Australia. That would turn a foreign funding story into a domestic one.

Would move the number

A residual value guarantee or a GPU-secured loan being tested: a heavily borrowed AI company missing a payment, or a lender admitting its chip collateral is worth less than the books said. That would give us the first observable price on equipment nobody has had to value.

Would move the number

A weak government bond auction where dealers get stuck with an unusual share, or a visible widening in the gap between swap rates and Treasury rates. That would mean the big leveraged Treasury trade is unwinding rather than being subsidized.

Would move the number