Archived reading, published Mon, 31 Aug 2026 14:18:48 UTC (21 days ago). This is not the current state of the meter.
See the live reading →Crash Lab watches the machinery under the current boom: the debt paying for AI data centres, private credit, the leverage that has moved out of banks into places nobody has to mark, the bond market and the dollar.
Every four hours it reads the day’s reporting from 21 sources and rewrites this page.
Why it moved: Ignition up two to 37: an Australian private credit fund with no exposure to the collapsed Bathla Group restricted withdrawals anyway (Bloomberg, 31 Aug) — that is liquidity contagion rather than credit loss, and it is the first gate we have seen travel beyond the lenders actually exposed. Fragility holds at 92: this week's genuinely new debt (Nebius $5.75bn convertible, QTS $3.9bn) is real but small against a stack we already carry, and nothing was unwound.
Reporting from 27 Aug to 31 Aug
Eight places it could go, scored 0–10. Tap one for the explainer.
What changed in the last few hours, and what each one says about the plumbing.
The first gate to be imposed by a fund with no exposure to the underlying failure is the moment credit stress becomes liquidity stress, and that is how the migration of risk out of banks actually bites.
Every mark in private credit is a manager's opinion until someone sells, and the three sales we can observe this month all cleared about ten percent below the opinion.
The most creditworthy company in the AI complex is now writing insurance on the least creditworthy part of it, and the exposure lives in a footnote rather than a capital ratio.
The bottom score band is deteriorating on its own while the average holds, which is what a credit cycle looks like before it is visible in an average.
Crypto's contagion channel to traditional finance has never been price; it is the moment its instruments become collateral in somebody else's funding.
The AI trade is being decomposed in public into the parts paid in cash and the parts paid in promises, and power is being sorted into the second pile.
Official-sector dollar demand is being managed by facility rather than by market, which works until the facility is the only bid.
Stuff you wouldn’t have believed was possible until 2026.
Volta Infra Holdings raised $300m in early August at a $2.4bn valuation, backed by Andreessen Horowitz, Altimeter, Nvidia and Michael Dell. On 27 August, JPMorgan began sounding out lenders for a $5bn debt package for its data-centre buildout — roughly twice the equity value the company carried three weeks earlier.
OpenAI will rent eight gigawatts of computing capacity from SoftBank at a former uranium enrichment facility in Piketon, Ohio — per N-TV, almost three times what is installed in all of Germany today — with more than nine gigawatts of new gas plants being built in the region to feed it. SoftBank's stated spend of up to $500bn is close to the entire German infrastructure special fund.
The Reserve Bank of Australia cut the US dollar share of its reserves from 55% to 45% during 2024/25, lifting the euro from 20% to 30%. It was disclosed in an annual report rather than announced; Deutsche Bank called it a quiet reduction.
Coinbase and Better have launched bitcoin-backed mortgages; one weekly recap says the waitlist implies more than $260m of demand. A house financed against an asset that has moved 22.7% in twenty days is a novel object for a loan servicer.
What the crash-callers are saying, checked against real reporting.
“There were 227,548 US properties with foreclosure filings in the first half of 2026, up 21% year on year, with activity rising in 77% of metro areas.”
ATTOM's mid-year 2026 report carries exactly those figures, and multiple outlets including Bisnow report the same 21% rise and a 28% increase on the first half of 2024.
Claimed by GoldSilver (Mike Maloney)
“Nearly 1.2 million South Korean leveraged trading accounts received margin calls in a single week as the Kospi collapsed almost 40% in 27 trading days.”
The BBC reports about 1.2 million retail accounts facing margin calls by end-July, and Bloomberg reports the near-40% fall in 27 trading days and the 18% single-day rebound. The '$2 trillion wiped out' figure is not established as stated.
Claimed by GoldSilver (Mike Maloney)
“The US personal savings rate fell to 2.7% in June, the lowest in 20–30 years, matching 2007–08 lows.”
The 2.7% June figure is correct per BEA, and July recovered to 3.0% — but the reported context is that this is the lowest since June 2022, not a two- or three-decade low. Worth knowing before it gets repeated.
Claimed by Reventure Consulting
“Central banks have bought roughly 1,000 tonnes of gold a year for four straight years.”
2022 (1,080t) and 2023 (1,037t) cleared 1,000 tonnes, but World Gold Council data show 2025 fell below that mark for the first time in four years. The accurate version is a four-year average, not four consecutive years.
Claimed by GoldSilver (Mike Maloney)
Dispatches from previous readings. The same argument, no longer the news.
For the first time in weeks the equity market is discriminating inside the AI trade, and it is discriminating on financing structure rather than on demand.
The gap between what private credit is marked at and what it sells for is the single number that determines whether losses arrive as a slow drip or all at once.
The US long end now depends on Japanese institutions not repatriating, and Japanese yields at 30-year highs make repatriation cheaper every week.
When the body whose job is to name systemic risk names one, the interesting question becomes why credit spreads did not move.
Treasury-company equity is a leveraged claim on a premium that can vanish without the underlying coin moving a cent.
The basis trade is the largest leveraged position in the US bond market, and it is becoming less visible to the people who would have to unwind it.
This quantifies a massive, hidden liability in the AI buildout, shifting risk outside traditional balance sheets and making the true scale of leverage harder to assess.
A sharp deterioration in private credit asset quality indicates underlying stress in the loans themselves, not just funding, raising concerns about potential losses.
The market's rejection of Treasury's intervention suggests deeper structural issues in the bond market and questions the effectiveness of policy tools in managing fiscal concerns.
Rising delinquencies in subprime auto loans and falling retail sales indicate increasing stress for the most vulnerable consumers, which can ripple through the broader economy.
These gates are a concrete example of funding channels closing in private credit, indicating liquidity stress and potential contagion risks for investors seeking to withdraw capital.
When a fund lends to a private company, the price it puts on that loan is its own estimate until somebody actually sells the loan. Those loans are now sliding into trouble seven times faster than they are recovering, and the funds' shares are trading as if nothing has happened.
Guarantees from the chip supplier are what make single-tenant data center campuses possible to finance. Those guarantees are underwritten on the assumption that the tenant keeps buying the guarantor's chips.
The fix that was supposed to make the most crowded trade in the government bond market survivable was central clearing, where a middleman stands between the two sides of every deal. The share of trades going through that middleman is falling while the trade itself grows.
Companies that hold bitcoin on their balance sheets transmit crypto swings into the stock market at roughly 1.6 times the move, and their ability to keep buying coins depends entirely on their shares trading above the value of what they already hold.
When a central bank changes what it holds in reserve, the decision is slow, deliberate and hard to reverse. That shift is now showing up in gold and in central bank portfolios, while the dollar's exchange rate itself does nothing.
A flat national average that conceals record stress among the weakest borrowers is two credit cycles running at once, and only one of them is visible in the headline numbers.
The only outside price check private credit has is an actual transaction, and a transaction that gets canceled leaves the marks untested.
The specific, observable things that would move the number - in either direction.
A US or European open-ended private credit or non-traded BDC vehicle imposing redemption limits for the same reason as the Australian fund — pre-emptive liquidity defence rather than realised losses. That would take ignition several points higher immediately.
Would move the number
High yield OAS moving above 350bp from 263bp, or the largest BDCs trading to a double-digit discount to stated NAV, which would mean the market has stopped taking the marks at face value.
Would move the number
A completed sale of a large GPU or data-centre asset portfolio at a disclosed price, which would give the residual value guarantees a first observable recovery number and either validate or destroy them.
Would move the number