Three price prints landed last week and every one of them behaved. Consumer prices came in exactly where the street had them, on all four lines. Producer prices came in softer than expected. Jobless claims ticked up and stayed historically low.

Then on Friday the consumer buckled. Retail sales fell six tenths of a percent against a consensus looking for a small gain, and consumer sentiment gave back its entire summer recovery in a single month.

In between, on Wednesday afternoon, the Treasury reported that the federal government ran a 432 billion dollar deficit. Not for the quarter. For the month. That is the largest July on record and the largest month of any kind since March 2021.

The ten-year Treasury yield fell on the news.

That combination is the week. The number everyone says will eventually break the bond market got printed, and the bond market rallied. The number that did move things arrived two days later with no headline attached to it.

There is also an asterisk on that deficit that almost nobody put in print, and it matters.

CONTINUE READING for the asterisk on the record deficit, why deficits do not set long-term interest rates and what does, the part of the PPI report that was not soft at all, the strongest evidence against the bearish case stated in full, and where gold, the miners and bitcoin sit against their own trend lines. Join us beyond the paywall, if you have not done so already.

Read more

On Monday, August 10, NVIDIA announced that it had signed on with Apollo, BlackRock, Blackstone, Brookfield, Goldman Sachs and KKR to build financing platforms aimed at mobilizing more than 500 billion dollars of third-party capital for AI infrastructure.

Half a trillion dollars. Morgan Stanley’s forecast for all global AI-related debt issuance in 2026 is about 570 billion. The release never says over how many years the 500 billion is meant to be raised. Spread over one year it is the entire AI debt market. Spread over five it is ordinary. Nobody has said which.

The coverage treated it as a done deal. It is not a done deal. The final line of the release, sitting directly above the media contacts, says the partnerships remain subject to execution of the final agreements.

That is not a technicality. That is the whole thing. Let’s follow the money.

CONTINUE READING for what was actually signed on Monday, why the 25 percent backstop everyone is quoting is not what it sounds like, which company actually ends up borrowing this money, and the one number that would tell you the structure is breaking. Always get the full picture through the lens of George Gammon for just $0.54/day with a premium subscription to RebelCapitalist.com.

Read more

On July 29 the Federal Reserve held rates at 3.50 to 3.75 percent. It was not a quiet hold. Three voters dissented, and all three wanted a hike.

Nine days later, payrolls printed negative 23,000.

So Wednesday at 8:30 in the morning is the scorecard. If July inflation runs hot, the three dissenters had the better read, the hold starts to look like a mistake, and a September hike goes from live to likely. If it runs cool, the committee threaded it.

Here is the part nobody on television will mention. There is a market where people are already betting real money on Wednesday’s number, and it is not pricing what the commentary implies.

CONTINUE READING for what that betting market is actually pricing, the numbers in Friday’s jobs report the headline buried, why the credit market is dead calm at the index level and screaming on a handful of specific borrowers, and the call for Wednesday morning. Join us beyond the paywall, if you have not done so already.

Read more

The S&P 500 closed Friday at 7,757.64, a record. It got there on the back of a jobs report that showed the American economy shed 23,000 payroll jobs in July.

Read that again, because the sequence matters. The economy lost jobs. The index printed an all-time high. Best week since April.

There is a version of this story where those two facts are in tension, and every headline you saw over the weekend told it that way. Bad news is good news, rate cuts are coming, buy. That version has one problem, which is that the futures market is not pricing a cut at all. It is pricing something close to even odds on a hike.

So what exactly did the market like about a negative payroll print?

Subscribe now

Monday and Tuesday: the numbers that got no coverage

The week did not start on Friday. It started on Monday, with two prints that nobody put on a chyron.

Construction spending for June came in at negative 0.1 percent against a median forecast of positive 0.3. Factory orders for June, out Tuesday, printed negative 0.3 percent against a forecast of positive 0.3. Job openings landed at 7.4 million, exactly in line, down from 7.5 million the month before.

There is a ritual that follows misses like these. Within about four hours, someone explains that construction spending was soft because of a specific line item, or weather, or a base effect, and the number gets filed under “noise.” Fine. But if the explanation is that obvious after the fact, why was it not in the forecast before the fact? A median forecast is not a guess pulled out of the air. It is the collective output of economists at Dow Jones Newswires and the Wall Street Journal who are paid to know about the line item, the weather, and the base effect. When the entire profession misses by 0.4 percentage points on a series it models every month, the interesting question is not what the excuse is. It is why the model needed one.

To be fair, and this matters, the same two days carried real strength. ISM manufacturing printed 55.6 against a 54.0 forecast, up from 53.3. That is not a soft number. That is expansion accelerating. Anyone building a bear case has to carry that on the books, not leave it off.

Wednesday: 44,000

Then ADP.

Private payrolls grew by 44,000 in July. The forecast was 75,000. The prior month was 95,000. So the number came in 41 percent below consensus and less than half the previous print, and the previous print was already the weakest in six months.

ISM services, same day, missed at 54.1 against 54.5. Still expansionary. Still a miss.

Thursday: the one number that argues the other way

Here is where the honest version of this gets uncomfortable, and it is worth sitting in the discomfort rather than skipping past it.

Initial jobless claims for the week ending August 1 came in at 199,000, better than the 204,000 forecast. The four-week moving average sits at 198,750. Two weeks earlier, for the week ending July 18, the advance figure printed 187,000, which is the lowest weekly reading since September 1969. Americans were still watching Apollo footage the last time this few people filed for unemployment insurance.

Continuing claims are not screaming either. 1,801,000 for the week ending July 25, against 1,963,000 a year ago. That is lower, not higher. Anyone telling you continuing claims are at a multi-year high is not reading the series.

Challenger’s July layoff count backs it up: 33,429 announced job cuts, the lowest monthly total in two years, down 27 percent from June and 46 percent from a year ago.

And Q2 productivity came in at 1.4 percent against a 0.6 percent forecast. Productivity growth is genuinely good news. It is the only free lunch in macro.

So the bear case has to answer this: if the labor market is deteriorating, why is nobody getting fired?

The answer is that initial claims measure exactly one thing. They measure the rate at which employed people are pushed out the door. They say nothing whatsoever about whether there is a door open on the other side. Firing is at a 57-year low. Hiring is what stopped. Those are two different faucets, and only one of them shows up in the weekly claims number.

That is also why claims are still the right timing tool despite all of that. In every genuine contraction on record, claims eventually spike. They have to, because at some point the frozen hiring turns into active firing. Which means the useful discipline is not to wait for claims to confirm the story, but to know in advance what confirmation looks like. Not 220,000. Not 250,000. Something north of 300,000 and heading toward 350,000 on the four-week average. Until then, claims are a fact that argues against the bear case, and pretending otherwise is how you end up with a thesis that cannot be falsified.

Friday: minus 23,000

The July employment report landed at negative 23,000 against a consensus of positive 83,000. A 106,000 swing.

Then the revisions, which are where the actual story lives. May was cut from 129,000 down to 63,000. June was cut from 57,000 down to 20,000. Combined, the prior two months lost 103,000 jobs that had already been reported, celebrated, and priced.

Average hourly earnings rose two cents to $37.62, up 3.2 percent year over year against a 3.5 percent forecast and a 3.4 percent prior. Wage growth decelerating.

And the unemployment rate went down, to 4.1 percent from 4.2.

The mechanics nobody walks through

The falling unemployment rate is doing enormous narrative work right now, and it deserves about ten seconds of scrutiny before it gets to keep doing it.

Most people carry an intuition that the unemployment rate is some kind of derivative of how many people are working. It is not. It is a ratio, and the denominator moves.

Look at what actually happened inside the household survey in the same release. Household employment fell by 87,000. The employment-to-population ratio fell from 59.0 percent to 58.9. Labor force participation fell from 61.5 percent to 61.4 percent. U-6, the broad underemployment measure, did not improve at all. It sat at 7.9 percent, unchanged.

So fewer people were working, a smaller share of the adult population held a job, and the headline rate improved. It improved because people left the labor force faster than they lost jobs. That is the entire mechanism.

Think about a restaurant that starts bragging its wait time has dropped to zero. The wait time really did drop to zero. It dropped because the neighborhood moved away. Nothing about the kitchen got better.

Now, the objection at this point writes itself, because it is the one that gets deployed every single month: the labor force is shrinking because of immigration policy, so of course payrolls are soft, and it does not mean anything about the economy.

Take that seriously, because the underlying data is real. Census reported net international migration falling from 2.7 million in the year to July 2024 to 1.3 million in the year to July 2025, a 54 percent collapse. Foreign-born workers are 19.1 percent of the civilian labor force. Shrink the inflow and you mechanically shrink job creation. That part is correct.

Here is where it stops working.

Run it to the extreme, which is how you test whether an argument has structure or just direction. Assume a labor force of 300 million and assume 250 million of them leave the country tomorrow, along with the 250 million jobs they held. What happens to the unemployment rate? Close to nothing. The numerator and the denominator fall together. What happens to the economy? It ceases to exist. Those 250 million people were not just supplying labor. They were paying rent, buying groceries, servicing car loans, and filling restaurant seats. One person’s spending is another person’s income, and that arithmetic does not care about anyone’s visa status.

So the migration argument does not rescue the data. It explains why the unemployment rate is not moving while the economy weakens. Those are opposite conclusions. When it is 250 million people, everyone agrees it would be catastrophic. When it is a few hundred thousand, it becomes a rounding error. The number changed. The logic did not.

There is a second problem with leaning on it, which is durational. This administration has been in office for roughly a year and a half. At some point, everyone who is going home has gone home, and the flow effect washes out of the year-over-year comparisons. You cannot run the same explanation indefinitely against a series that keeps deteriorating. Eventually a negative payroll print is just a negative payroll print.

One correction worth making, because precision matters more than momentum: this is a single negative month, not a streak. June was revised down hard, to positive 20,000, but it was still positive. The story is not consecutive contraction. The story is a 103,000 two-month revision and a headline that broke below zero for the first time in this cycle.

The other side of the ledger, stated properly

Now the part that a bearish desk is obligated to put in front of you, because leaving it out would make everything above worthless.

The Atlanta Fed’s GDPNow estimate for Q3 2026, as of August 6, sits at 5.83 percent annualized. Not 1.8. Not 0.8. Nearly six percent.

Q2 real GDP came in at 1.5 percent on the advance estimate. ISM manufacturing is accelerating. Claims are at a 57-year low. Productivity beat by more than double. Layoffs are the lowest in two years.

That is not a cherry-picked list. That is a genuinely strong economy on most of the instruments people actually watch.

So how do both readings coexist? The same way a car’s speedometer can read 90 while the fuel gauge reads empty. Neither instrument is broken. They are answering different questions. One tells you how fast you are moving right now. The other tells you something about the next 200 miles. GDPNow is a nowcast built largely from data that has already printed, and early in a quarter it is famously jumpy. The labor market is the transmission mechanism into the quarter after next.

And this is the part that decides whether any of it matters: the NBER does not date a contraction off a nowcast. It dates one substantially off the labor market. You can print negative GDP and get no recession call if employment holds. You will not get a labor market that breaks without negative GDP following it. Which is why the payroll series carries more weight than its headline volatility deserves, and why 5.83 percent does not settle the argument.

The index does not care, and it never did

Look at that chart without a narrative attached and it is a good chart. It broke out. It is above the 50-day and miles above the 200-day. The oldest rule in trend following, popularized by Martin Zweig in Winning on Wall Street in 1986, is that the trend is your friend and you do not fight the tape. That rule has made vastly more money than any valuation model.

Is it expensive? Absurdly. Forward twelve-month P/E is 20.0 against a ten-year average of 19.0. The cyclically adjusted P/E sits at 42.39, against a long-run mean of 17.40 and an all-time high of 44.19 set in December 1999. We are inside a percent and a half of the dot-com peak on that measure.

Both things are true at once, and the failure mode of the bearish crowd is refusing to hold them simultaneously. Overvalued is not a timing signal. It has never been a timing signal.

What is worth flagging is the quality of the reasoning being used to justify each new leg. The argument moves. Six months ago the market rallied because the Fed was going to ease. Now the Fed is on hold, the futures market has flipped to pricing a possible hike, and the market is rallying anyway, because the economy is resilient. Those are contradictory premises producing an identical conclusion. When the conclusion is fixed and the reasoning rotates to fit it, that is not analysis. That is a search for permission.

None of which tells you when it ends. Nobody in this business is good enough to pick the top, and the ones who claim they are have a track record you can look up.

But there is a mechanism worth watching, and almost nobody is charting it.

Supply

Forget sentiment. Forget valuation. Think about shares as units of something, the way you would think about currency units.

If the number of dollars in the system goes up faster than the goods those dollars chase, each dollar buys less. The same arithmetic applies to equities. If the number of shares outstanding grows faster than the earnings behind them, each share is a claim on less. Price is set at the margin by supply against demand, and for fifteen years the supply side of that equation has been running in reverse.

Since the financial crisis, the single most reliable bid under this market has been corporate America shredding its own paper. Buybacks are still running near 1.2 trillion dollars annualized. Every dollar of that is share count going down.

Now look at what just happened on the other side.

US IPO proceeds hit 251 billion dollars in the first half of 2026, a first-half record that beat 2021, with eleven deals above a billion. Total US equity new issuance for the half came to 307.7 billion. UBS’s full-year estimate runs 200 to 350 billion in IPOs plus roughly 400 billion in secondaries, so somewhere between 600 and 750 billion dollars of new paper in a single year.

The individual deals are the part that should stop you. Alphabet raised roughly 85 billion dollars in equity in June, upsized from 80, with 10 billion of it taken down by Berkshire. That is the largest equity offering in history, and the stated purpose was AI infrastructure and compute. SpaceX priced its IPO at 135 dollars a share on June 11, raising 75 billion, roughly 86 billion with the greenshoe, at about a 1.75 trillion dollar valuation. That is the largest IPO ever completed, by a factor of nearly three over Saudi Aramco. OpenAI filed confidentially in June targeting something close to a trillion-dollar valuation and a 60 billion dollar raise. Anthropic filed confidentially on June 1.

And the reason all of it is happening at once is not a coincidence of the calendar. The four largest hyperscalers are guiding to more than 700 billion dollars of capital expenditure in 2026. Alphabet alone guides 180 to 190 billion. Meta guides 125 to 145. That spend has to be funded, and increasingly it is being funded with stock.

Paul Tudor Jones laid out the mechanism publicly this spring on Invest Like the Best, and it is worth stating plainly rather than quoting at length: the supply of stock rises while buybacks fall, because the same capital expenditure commitments that force the issuance also consume the cash that used to fund the repurchases. He described the US market as over-equitized at 252 percent of GDP, and drew the parallel to 2000, where the damage came not on the IPO date but later, when lockups expired and the restricted stock hit the tape.

Here is the honest state of play, and it cuts against the dramatic version of this story: net supply has not flipped yet. At 1.2 trillion in buybacks against 600 to 750 billion in issuance, corporate America is still retiring more stock than it creates. UBS expects that to hold through year end.

So this is not a signal that has fired. It is a signal to watch, and the specific thing to watch is the ratio, not either leg on its own. The moment buyback authorizations start getting trimmed to fund data centers while issuance keeps setting records, the machine that has absorbed every dip since 2009 starts running backward. That is a mechanical change in market structure, not a mood.

Michael Burry, who deregistered Scion in November 2025 and now publishes only through his own newsletter, wrote on August 4 that a “1987-type fall” is possible while conceding that new highs will pull new money in. One caveat on him that the headlines never carry: he is not short-only. His own recent post titles include “4 Shorts, 3 Longs,” and he covered his Palantir and Oracle shorts in the same stretch. The book is net bearish. It is nowhere near as apocalyptic as the coverage implies.

Where the money might already be lying in the corner

Gold spot closed the week around 4,324. Look at the shape rather than the level. Roughly 5,500 in February, a long grind down to just under 4,000 by midsummer, and now a sequence of higher lows and higher highs off that base. It is still below its 200-day at 4,483, but it has taken back the 50-day.

Silver looks similar in shape and worse in structure. Around 63, just above a 50-day at 62.30, and a long way below a 200-day at 70.41. It bounced off the mid-50s. It has not repaired anything.

The miners are the interesting one. GDX closed Friday at 89.89, up 7.11 percent on the day. Peak near 115 in March, a brutal slide to roughly 70 by July, and now a move that has carried it back above its 200-day at 87.37, with the 50-day at 78.44 turning up underneath it.

That is the whole argument for preferring the miners here, and it is not a story about gold bugs or debasement or any of the usual furniture. It is that of gold, silver, and the miners, exactly one has reclaimed its 200-day. The metal is the underlying. The miners are the leveraged expression, and they are the ones printing the better structure. If the macro read is a weakening economy and eventual pressure on real rates, the leveraged expression is where the convexity lives.

Bitcoin is bouncing too, at roughly 64,900, off a June low near 57,000. But it is still under a 200-day at 70,015 that has been sloping down for the better part of a year, and it is a long way from the 115,000 area it traded last autumn. Same bounce, much worse chart.

Given a choice among the four on chart structure alone, the miners are the only one where the trend has actually turned rather than merely paused. None of this is investment advice, and the discipline that matters more than the pick is not buying into a downtrend just because the macro thesis feels right. Falling knives do not care how correct you are.

The most hated bond on earth

The 30-year Treasury yield closed Friday at 5.20 percent. Trough near 4.55 last October, a steady climb all year, a 50-day at 5.04 and a 200-day at 4.88, both rising.

There is not a more universally despised asset in the world right now. The consensus says the long end goes to six, then eight, then fifteen, and the reasoning is always some combination of deficits, issuance, and de-dollarization.

The contrarian case is not that the consensus is stupid. It is that the consensus is already positioned. When everyone who was going to sell has sold, the marginal seller stops existing, and the asset stops going down on bad news. That is the setup that makes a hated long interesting.

But the chart says wait. Higher highs and higher lows in yield means the trend is still up, which means the price trend is still down, and buying it here is fighting the tape for the sake of being early. The condition to watch is straightforward: yields need to start printing lower highs and lower lows, and then break back below the 5.10 area toward five. That is where the macro view and the tape would finally agree. Until they do, the view is a view, not a position.

The week ahead

Wednesday, August 12: CPI. Headline is forecast at 0.1 percent for the month and 3.4 percent year over year, down from 3.5. Core is forecast at 0.3 percent monthly and 2.5 percent year over year, down from 2.6.

Look at the spread between those two, because it is the single most under-discussed number in this market. Headline at 3.4 and core at 2.5 means the inflation problem is almost entirely energy. And energy has already broken. Brent touched 126 dollars intraday on April 30, a four-year high, on Strait of Hormuz blockade fears, and settled at 114.01. As of Friday, Brent was near 85 and WTI near 78. That is a third off the high. The Middle East risk premium is real and it can return without warning, but the thing that has been keeping the inflation narrative alive has been retracing hard for three months while wage growth decelerates to 3.2 percent. That combination is disinflationary, not inflationary.

Thursday, August 13: initial claims are forecast at 203,000, and PPI at 0.2 percent. Claims is the number that matters. Not because 203,000 tells you anything, but because the four-week average is the tripwire, and it is nowhere near tripped.

Friday, August 14: retail sales forecast at 0.1 percent, and preliminary consumer sentiment at 54.5. Sentiment has been sitting near record lows for months while the index makes record highs. That gap is the entire story of this economy in one comparison.

Then September 16, the FOMC. The Fed has been on hold at 3.50 to 3.75 percent for five straight meetings under Kevin Warsh. Futures are roughly 55 to 60 percent on another hold and 40 to 45 percent on a quarter-point hike, with those hike odds having dropped sharply after Friday’s payroll print. No cut is priced at all.

Sit with that. Payrolls went negative, wage growth is decelerating, oil is a third off its high, and the market’s base case is still that the next move is up. If that is right, the equity market is trading at a 42 CAPE into a tightening cycle. If it is wrong, something in the labor data is about to become impossible to explain away.

What would prove this wrong

The bear case here rests on one load-bearing claim: that hiring has frozen while firing has not yet started, and that the second half always follows the first. If that is right, the four-week claims average moves through 250,000 and toward 300,000 in the coming months, the household survey keeps bleeding, and the participation rate keeps falling for reasons that have nothing to do with anyone’s border policy.

If it is wrong, it will look like this. Claims stay pinned near 200,000 through the autumn. Payrolls revert positive in August and September without a downward revision. Participation stabilizes. GDPNow is vindicated and Q3 prints somewhere north of four percent. In that world, July was a statistical air pocket in a genuinely strong expansion, and the correct move was to own the index and ignore the noise.

Both of those are live. Nobody gets to know yet. There are no certainties here, only probabilities, and the honest version of the probability is that the labor data has been deteriorating for three consecutive months while every explanation offered for it has a shorter shelf life than the last one.

Watch the claims average. Watch the buyback authorizations against the issuance calendar. Everything else this week was commentary.

Facts > Narrative. Always.

Stand up for freedom, liberty, and free market capitalism.

Monday’s tape said nothing happened.

The Dow closed up 262 points, about half a percent, at 52,210. The S&P 500 finished up a point and change, call it flat. The Nasdaq slipped a fifth of a percent. Brent settled 6.3% lower at $85.87 after the United States and Iran paused their strikes, with West Texas Intermediate down harder still, more than 8% at its worst. Glance at the closes, go back to your day, and you would conclude the market chewed on the weekend’s news and decided it was fine.

The weekend’s news was that Nvidia is in talks to guarantee roughly $250 billion of financing so that OpenAI can lease a data center campus in southern Ohio.

One floor away, on a desk most equity investors never bother to look at, the reaction was not fine at all. Five-year credit default swaps on Nvidia’s debt, which are nothing more exotic than insurance contracts against Nvidia failing to pay, posted their biggest one-day jump on record, widening about 14 basis points to roughly 82 at the intraday peak. That contract only began actively trading last November. It has never moved like that before.

Two markets looked at the same headline on the same day. One shrugged. One repriced the risk of the most profitable company on earth. Somebody is reading this wrong, and the tell for which one is sitting in a public filing that takes about ninety seconds to find.

A note for free readers. This one is on the house, in full, no gate. The paid edition is where the weekly rates and positioning work lives, plus the levels being watched into each print. If this is the kind of read you want two or three times a week, consider joining us on the paid tier.

Subscribe now

What Nvidia’s $250 Billion OpenAI Backstop Actually Does

Start with the strongest version of the bull case, because it is genuinely strong and it deserves a fair hearing before anyone takes a swing at it.

Vendor financing is not a scam. It is one of the oldest moves in industrial capitalism. When you make a capital good, and your customer needs a decade of financing to buy it, and the banks do not know your customer as well as you do, you put your balance sheet behind the deal. Every railcar builder, every aircraft manufacturer, every equipment maker in history has done some version of it.

And Nvidia has the balance sheet. Revenue of $81.6 billion in a single quarter. Net income of $58.3 billion in that same quarter. Cash, marketable debt, and marketable equity securities north of $80 billion. The demand for compute is real, it is unmet, and the buyers are lining up.

Look at the structure and it gets more reasonable, not less. Nvidia is not writing a check. It is guaranteeing a lease. The guarantee is contingent, it only bites if the tenant defaults, and it shrinks over time as the tenant makes its payments. SB Energy, the SoftBank power subsidiary developing the 10 gigawatt site on a decommissioned uranium enrichment plant about fifty miles south of Columbus, gets to borrow against Nvidia’s credit instead of its tenant’s. OpenAI has no investment grade rating, so without a backer, the debt either does not get raised or gets raised at a price that kills the project. Japan is putting $33 billion into the gas generation. The Commerce Department controls the power allocation. This is not two guys in a garage.

So the steelman goes: strongest balance sheet in the world underwrites a lease for a customer it knows better than any lender could, unlocks a project the country’s own energy department is sponsoring, and takes warrants for the trouble.

Fine. Now walk the mechanics.

The Number Buried in Nvidia’s Own 10-Q

Nvidia already does this. It has a guarantee book, it discloses it, and the disclosure is one paragraph long.

Go to the quarterly report filed with the SEC for the three months ended April 26, 2026, and find the facility lease guarantee note. Maximum gross exposure across every such agreement Nvidia has signed: $3.5 billion. Terms running five to seven years. Partners have parked $712 million in escrow to cut the exposure further. The company classifies them as credit derivatives and tells shareholders the guarantees were “not material.”

$3.5 billion.

The reported Ohio backstop is $250 billion. That is roughly seventy-one times everything Nvidia has guaranteed to date, combined. It is more than Nvidia’s entire revenue for fiscal 2026. It is several times its cash and marketable securities. And it sits alongside a separate conversation, reported at the same time, about financing up to $350 billion of chip purchases.

Is a jump in scale automatically a problem? No. Companies grow into bigger commitments all the time. But scaling a business line and scaling a contingent liability by seventy-one times in one negotiation are different animals, and the second one does not show up on the balance sheet until the day it does.

Where Nvidia’s Profit Came From Last Quarter

Here is the part nobody put in a headline.

In that same filing, Nvidia reported income before tax of $69.9 billion. Operating income was $53.5 billion. The gap, $16.4 billion of it, came in under total other income. Almost all of that was gains on equity securities. Marks on Nvidia’s investment portfolio.

What is in the portfolio? Non marketable securities, the line holding Nvidia’s stakes in privately held companies plus its infrastructure fund positions, went from $22.3 billion in late January to $43.4 billion by late April. Nvidia bought $18.6 billion of them in three months. The year ago quarter, that line was $649 million. Investment commitments outstanding: $27 billion.

So roughly a quarter of the pretax income came from marking up positions in a private AI complex that Nvidia is simultaneously funding, supplying, and now proposing to cosign for. And on the revenue side, the same filing shows three direct customers accounting for 30%, 18%, and 16% of accounts receivable. Sixty four percent of what Nvidia is owed sits with three counterparties. One quarter earlier that number was 56%. It is getting more concentrated, not less.

Alphabet did a milder version of the same trick last week. The headline was an enormous earnings beat. Read the release and a $99.0 billion gain on equity securities added $77.1 billion to net income and $6.26 to diluted earnings per share. Underneath that, capital spending of $44.9 billion outran $39.1 billion of operating cash flow, free cash flow came in at negative $5.9 billion, and full year capex guidance went up to a range of $195 billion to $205 billion. The stock fell about seven percent, which tells you the buy side has started reading the cash flow statement before the press release.

Picture a kid who wants a car loan. His detailing business loses money every month, so the dealer says no. Then his father cosigns, and the dealer says yes, because the father has income. Nobody asks where the father’s income comes from. It comes from the kid’s detailing business, where he is the paid CEO. The cosigner and the borrower are the same credit wearing two coats. When the kid’s business fails, the father’s paycheck stops in the same instant the guarantee is called. The guarantee was never worth anything independent of the thing it was guaranteeing.

That is not a circle. A circle you can follow. Twist it into enough loops, add a sovereign, a trade deal, a power allocation and an escrow account, and you get something closer to a pretzel, which is exactly why it has held up this long. You have to trace it with your finger to see it comes back where it started.

The Bond Market Files Its Dissent

Credit desks do not get paid for upside. They get paid for being right about who pays and who does not, which makes them the least romantic readers in finance.

Nvidia’s swaps are not the only ones moving. S&P has cut Oracle to BBB minus, the bottom rung of investment grade, and its five-year protection is trading around 215 basis points. None of this is a solvency call on Nvidia. The company earned $58 billion in a quarter and has a fortress of liquidity. It is a call on how much contingent obligation the market wants to sit behind, and at what price.

The Bank for International Settlements got there first. In its Annual Economic Report published June 28, the institution that central banks answer to named AI financing as one of four global pressure points. The specific objection was not valuation. It was structure: chipmakers and cloud providers taking equity in AI labs that then commit to multiyear purchases from those same investors, on terms the report calls “poorly disclosed, with risks of the same asset being pledged multiple times.” The five largest hyperscalers are on pace to spend more than a trillion dollars across 2025 and 2026, already past what their earnings and free cash flow can cover.

Central banks are not usually early. When they are, it is worth a second look.

The second look arrived overnight. Seoul closed down 10.8%, its worst day since the opening of the Iran conflict in March. Samsung fell 13.4%, its worst session in nearly twenty years. SK Hynix dropped 14.7% at home, and its American shares closed below their IPO price for the first time since listing this month. Kioxia lost 18.3%. Those are the companies that actually make the memory that goes in the racks. They are not trading on a story about compute demand any more. They are trading on whether the financing behind the compute demand is real.

What To Watch Now

The honest framing is that none of this proves anything yet. A deal that is still in negotiation is not a deal. Nvidia has issued no statement, terms are unset, and a guarantee is not a loan. There are no certainties here, only probabilities.

But there are three things worth watching this week, and they are specific.

Meta and Amazon report in the next few days, and the number that matters is not earnings, it is capital spending guidance and how each one intends to fund it. If the pattern from last week holds, a beat plus a capex raise gets sold rather than bought. That would be the third straight confirmation that the market has stopped paying for spending and started charging for it.

Second, the swaps. One record day is noise. Nvidia protection sitting at 80 basis points or wider a week from now, with the equity flat, is the credit market maintaining a position rather than reacting to a headline.

Third, whether the guarantee actually gets signed, and at what number. A backstop that comes in at $250 billion tells you Nvidia believes the demand is durable enough to underwrite. A backstop that lands materially smaller, or gets restructured into something with a cap and a term, tells you the lawyers priced the tail risk and the lawyers won.

Here is the line in the sand. If Nvidia’s next quarterly filing shows the guarantee book stepping up from $3.5 billion into the tens or hundreds of billions while the customer concentration keeps climbing, then the supplier has become the credit for its own end market, and the equity should be valued like a lender rather than a chipmaker. If instead the guarantee stays contingent, capped, and small relative to the balance sheet, then Monday’s move was a headline reaction and the credit desks will unwind it quietly.

Either way, the filing is public and free. The market spent Monday trading a press report. Half the answer was already on the SEC’s website.

Facts > Narrative. Always.

Stand up for freedom, liberty, and free market capitalism.

Subscribe now

Refer a friend. Get paid in research.

If you’ve found the Rebel Capitalist News Desk useful, odds are you know a few people who would too…investors who are tired of consensus finance coverage and want a thesis they can actually build a position around.

Here’s how the referral program works:

Share your personal referral link (find it in your account settings under “Refer a friend”). When someone subscribes using your link, it counts. Hit the thresholds below and we add comp months to your subscription automatically…no action required on your end.

• 5 referrals → 1 free month added

• 10 referrals → 3 free months added

• 15 referrals → 6 free months added

There’s also a leaderboard on the site if you want to see where you stand.

No complicated signup. No affiliate dashboard to manage. Your link is already active…you just have to use it.

Refer a friend

Subscribe now

On Wednesday night Alphabet told the world its earnings per share had risen 294%.

By Thursday afternoon the stock had dropped about 7%, and the seven largest technology companies in America had shed roughly $797 billion of market value between them, the worst single session for that group since the tariff panic of April 2025.

So which is it? Best quarter in the company’s history, or the thing that took eight hundred billion dollars off the board in six hours?

Both. The reason sits in a small table near the back of the earnings release. Alphabet generated $39.1 billion in cash from operations last quarter and spent $44.9 billion on property and equipment. Subtract one from the other and free cash flow came to negative $5.855 billion, the first negative quarter in the twenty-two years Google has been a public company.

The objection writes itself, and it is fair. Negative free cash flow because your business is rotting is a catastrophe. Negative free cash flow because you are building the future is an investment. Those are genuinely different animals, and anyone who pretends otherwise is selling something.

Fine. Grant the point entirely. Then ask the question that decides it: at what point does the building stop being building and start being the cost of showing up?

CONTINUE READING for the number that turns a 294% blowout into a miss, how Alphabet actually paid for a quarter it could not fund out of pocket, the one Magnificent Seven name getting paid for refusing to play, what the last capital-spending mania did to shareholders, the honest math on the trade everyone is reaching for, and the week ahead. Join us beyond the paywall, if you haven’t done so already.

Read more

Inflation came in soft last week. Two prints, both below consensus, both cheered.

And roughly seven thousand miles away, 1.2 million leveraged retail accounts in South Korea were being liquidated by their brokers.

Nobody connected the two, because on the surface there is nothing to connect. One is a story about prices, the other about Samsung and SK Hynix. Different continents, different asset classes, different news cycles.

But run both stories down to the mechanism and they are the same story. Both are about what happens when the money doing the buying was borrowed.

So here is the question worth paying for: if borrowed money can do that to a market with a fraction of the leverage, what is the number in American accounts telling us right now?

Read more

Subscribe now

The most important number in markets right now is not a stock price. It is a single figure that just went negative for the first time in the history of the data.

For roughly two decades, the biggest software companies on earth were cash machines. Amazon, Alphabet, Meta, Microsoft, Oracle. They threw off free cash flow the way a garden hose throws off water, steadily, reliably, boringly. That era is over. Bank of America’s own research shows the twelve-month forward free cash flow of the hyperscalers peaked near $280 billion around 2024, then rolled over hard, and is now projected to go negative for the first time in the entire series going back to 2007.

Read that again. The most profitable business model of the last twenty years is about to stop generating cash.

Not because these companies stopped making money. They are minting it. The cash is being redirected, every dollar of it, into AI capital expenditure, and out the other side of that pipe it lands on the income statements of the semiconductor companies. The chip makers’ free cash flow is going vertical at the exact moment the hyperscalers’ is going through the floor. Same money. Different pocket. That is the whole story in one picture.

So here is the question nobody on financial television wants to sit with: if you are just moving the cash from one set of balance sheets to another, and the market cap of both sets goes up at the same time, what exactly is being created here? Or is it just being rearranged?

So here is the question nobody on financial television wants to sit with: if you are just moving the cash from one set of balance sheets to another, and the market cap of both sets goes up at the same time, what exactly is being created here? Or is it just being rearranged?

CONTINUE READING: below the paywall we walk the mechanism the mainstream is calling a healthy rotation and show why the numbers say otherwise, why demand and an economic model are two completely different things and how the twenty-dollar-bill trap applies directly to the AI build, what the 1840s railway mania and the dark fiber of the dot-com era tell you about who actually gets paid when a capex boom finally delivers, why Korea and the Kospi are the pressure gauge to watch, how the yen breaking 162 lights a second fuse under global risk, and the single line in the sand that tells you the moment the whole machine goes into reverse. Join us beyond the paywall, if you haven’t done so already.

Read more

Subscribe now

Here is a number that is supposed to be good news.

In May, Americans reduced what they owe. Total consumer credit outstanding actually shrank. Wall Street had penciled in an increase of about $17.5 billion, right in line with the $20.8 billion jump the month before. Instead the Federal Reserve’s G.19 report showed the whole thing went into reverse, the first monthly contraction since November 2024.

Turn on financial television and you already know the script. Households are deleveraging. Balance sheets are healing. The consumer is being responsible, paying down those high-rate cards, getting the house in order before the second half of the year. So a pullback in borrowing must mean strength, prudence, a soft landing you can bank on.

It is a tidy story. It is also almost certainly backwards.

Because there are two ways a country’s credit card balances go down. One happens when people are flush and retire some debt. The other happens when people are so far underwater they cannot borrow another dollar, and the balances that vanish are not paid off so much as charged off. Same direction on the chart. Opposite universe underneath.

So which is it? What does the rest of the data say once you stop reading the headline and start reading the fine print?

CONTINUE READING to see the delinquency chart that looks like 2008 all over again, the auto-loan number that just broke a record that stood for two decades, the reason your real cost of living is nowhere near the official inflation rate, and the exact three-signal dashboard that tells you when the last stage of this credit cycle finally arrives. Join us beyond the paywall for this weeks wrap-up and look ahead, if you haven’t done so already.

Read more

Subscribe now

Here is a number that should stop you cold. In June, existing home sales in the United States ran at a seasonally adjusted annual rate of 4.09 million. That is not a soft patch. That is the pace of home selling the country last saw digging out of the 2008 wreckage. Adjust it for a population that has grown by tens of millions of people since then, and on a per-household basis the American housing market is transacting at a rate below the depths of the Great Financial Crisis.

Now here is the number sitting right next to it. The median existing home price hit a brand new record of $440,600, the 36th month in a row of year-over-year gains.

Crisis-level volume. All-time-high price. Sitting in the same monthly report, describing the same market, in the same week.

So which one is lying?

Neither, actually. And that is the part almost nobody covering this beat is willing to walk you through, because it forces an uncomfortable conclusion about where prices go next. The health of a market is not measured by its last print. It is measured by whether anything is actually moving. And right now the patient has a pulse, the monitors are still beeping...but the thing is flat on its back, hooked to the machines, and the family is being asked to make some decisions.

If almost nobody is selling, and prices still cannot manage to rise, what does that quietly tell you about the buyer? And what happens to that record price the moment the supply side thaws even a little?

Keep reading and here is what is waiting for you below the line: why a “record price” in a market with no transactions is the weakest kind of record there is, the metro map that is already flashing red under the national average, the myth of the frozen homeowner who “will never, ever sell,” the willing seller who just walked away from an entire portfolio and told everyone why, and the homebuilder quietly setting the real clearing price while everyone stares at the headline. Join us beyond the paywall, if you haven’t done so already.

Read more

Subscribe now

The most expensive stock in the history of initial public offerings just spent its first month falling.

SpaceX went public on June 12 at $135 a share, ripped to almost $226 within four days, and has been bleeding lower ever since. It closed today down almost 7 percent, near $149, a hair above where it started life on the Nasdaq. The largest IPO ever printed, the one that briefly made Elon Musk the world’s first trillionaire, is now round-tripping back toward its debut price while the people who bought the top hold the bag.

That alone would be a story. A hot IPO cooling off is as old as markets. But the reason this one matters has almost nothing to do with rockets, and everything to do with a single line item buried in a regulatory filing days before the stock ever traded. Once you see how that line item works, you cannot unsee it. And it explains far more than one wobbly stock.

So let’s connect the dots.

The deal that showed up right on time

On June 5, a week before the offering, a filing disclosed that Google had agreed to pay SpaceX roughly $920 million a month for AI compute capacity housed in xAI's data centers. That is not a typo and it is not an annual figure. It is nearly a billion dollars, every month, running from October of this year through June 2029. Call it $11 billion a year, landing on the books of a company that was about to ask the public to value it at more than $1.75 trillion.

Now, the timing. Google, a company that runs one of the largest cloud businesses on earth and designs its own AI chips, agreed to rent roughly 110,000 Nvidia GPUs from a direct competitor in AI models. Days before that competitor’s IPO. At a price that landed as recurring revenue precisely when a recurring-revenue story was worth the most.

Ask yourself a simple question. If you were selling your business next Tuesday, and a friendly giant offered to sign a multi-year, eleven-figure contract this Friday, one that pumps the exact number your buyers are staring at, would you call that a coincidence? Or would you call it a favor with a very specific delivery date?

Here is the part that makes it more than a hunch. This was not Google's first move on SpaceX. Alphabet backed the company in 2015, sinking in around $900 million when the whole business was valued at roughly $12 billion. At the IPO, SpaceX was valued north of $1.75 trillion. So the same company writing the $920-million-a-month check is also sitting on a decade-old equity stake that just got marked to a valuation more than a hundred times higher. A commitment to pay eleven billion over several years is a rounding error against what that stake is now worth on paper. You do the math on which number Google cares about more.

And this was not even the first compute deal wheeled out ahead of the offering. A month earlier, Anthropic signed on to use SpaceX’s compute at its Colossus data center in Memphis. Two enormous customer commitments, both materializing in the weeks before the public got to buy. Surging demand, right on schedule.

So what does the average investor, staring at a $1.75 trillion price tag on a rocket company that loses money, actually need to understand before deciding whether the story holds together?

Subscribe now

Watch where the dollars actually go

Strip away the drama and look at the shape of it, because the shape is the whole thing.

Nvidia sells the chips. SpaceX and xAI buy the chips and stand up the data centers. Google rents the compute those chips provide and pays close to a billion a month for it. That payment lands as revenue, which helps justify buying more chips, which flows back to Nvidia, which has been busy investing in and backstopping the very companies doing the buying. Round and round it goes.

Nobody in this loop is technically lying. Every contract is real, every dollar is disclosed in a filing. That is what makes it so slippery. The genius of it is that a single pile of money can be counted as revenue at one company, as demand at another, and as a growth story at a third, all at the same time, as it moves around the ring. One dollar wearing four hats.

Give the bulls their strongest case here, because it is not nothing. Compute really is scarce. Even hyperscalers with unlimited balance sheets cannot build data centers fast enough to keep up with the permitting, the power hookups, and the construction timelines. Renting capacity that is already switched on is a legitimate way for Google to add compute faster than pouring its own concrete. Google itself framed the deal as bridge capacity for demand it says has outrun its own forecasts on its Gemini Enterprise product. Take that at face value. The AI buildout is real, the demand is real, and some of these deals are exactly what they look like.

But that is precisely where the sleight of hand lives. Because demand and profit are not the same word, and the whole edifice depends on you forgetting the difference.

Surging demand is not a profit

Imagine a business that sells twenty-dollar bills for ten bucks each. What would demand look like? It would go vertical. Your customers would form a line around the block. Your total addressable market would be roughly every human being with ten dollars. You would post the most breathtaking growth curve anyone had ever seen.

And you would go bankrupt.

That is the trap in every “surging demand” headline attached to these AI deals. Huge demand tells you the price is attractive. It tells you nothing about whether the seller makes money, and nothing about whether there is even a path to making money. When the demand exists only because someone is selling compute, or capital, or a story, for less than it truly costs, the demand curve is a distraction. The question that matters is downstream and quieter: does the cash actually come back?

Look at the fundamentals underneath the price and the trap comes into focus.

In its most recent reported quarter, SpaceX’s AI segment generated $818 million in revenue and posted a $2.5 billion operating loss. In the same period, company capital expenditures hit $10.1 billion, more than double a year earlier, with $7.7 billion of that poured straight into AI infrastructure. So the segment collecting the demand is losing money at a rate of billions a quarter, while the spending required to keep collecting that demand is running at ten billion a quarter and climbing.

Now set Google’s $920 million a month against that. For the arrangement to make sense as a business rather than a favor, Google needs to extract more value from that compute than it pays. If it turns out Google is paying a rate well above the going market price, and some of the people who read the filing closely think it is, then this is not a customer relationship. It is a subsidy dressed as a sale, timed to a stock offering.

And a subsidy has a shelf life. Read the fine print in the filing and the whole thing is built on quicksand. SpaceX disclosed that if it fails to deliver the committed GPUs by a September deadline, Google can terminate immediately. After this year, either side can walk with 90 days’ notice. So the eleven-billion-a-year headline that helped anchor a trillion-dollar valuation is a contract that can evaporate one quarter after the shares are distributed. Pump the number when it counts. Keep the option to cancel once the selling is done.

That is the mechanism. Not a conspiracy whispered in a back room, but an incentive structure sitting in plain sight in a public document, doing exactly what it was designed to do.

Subscribe now

The thing everybody knows and nobody can trade

Here is the uncomfortable part, and the reason this is so hard to act on.

You can see the whole scheme. You can lay out the circle, name the deals, quote the filing, point at the money-losing segment underneath the trillion-dollar tag. And it changes nothing about what the stock does tomorrow, because a game like this can run for years. There is an ocean of capital out there with an enormous appetite for risk, and as long as that appetite holds, the music keeps playing. Pointing at the emperor’s wardrobe does not put clothes on him, and it does not stop the parade.

So the temptation is to short it. Resist that temptation. Trying to stand in front of this thing is like stepping onto the tracks to stop a runaway freight train by holding up your hand. You are correct about where the train ends up. You will still get run over. The tape does not reward being early, it punishes it, and “early” and “wrong” cash out at the same window.

What actually breaks the spell is not an argument. It is arithmetic finally showing up in the cash flows. The day the market decides that the compute Google is renting cannot generate more than Google is paying, that the AI segment has no credible path from a $2.5 billion quarterly loss to daylight, that the demand was always just cheap money circling the drain, the story reprices in an afternoon. Nobody can tell you the date. Anyone who claims to is selling something.

But you can watch for the tells, and this month has produced a cluster of them worth respecting.

What to watch

Start with the stock itself. SpaceX going up and to the right, higher highs and higher lows, would have said the appetite for this game was still bottomless and the ruse had plenty of runway left. Instead it topped near $226 and rolled straight over, giving back nearly a third of its value in weeks. That is not proof of anything on its own. It is a chink in the armor.

Then look at who is quietly heading for the exits. Blackstone, one of the biggest names in data-center buildout, just sold its stake in three Northern Virginia hyperscale data centers and walked away from what was slated to be the largest data-center campus in the country. Meta, which raised its capex forecast to as much as $145 billion this year, is now reportedly building a business to rent out excess compute. Sit with that word. Excess. If demand is truly insatiable, how does anyone end up with too much? Either the demand is not as bottomless as advertised, or the buildout has run ahead of it. Maybe both.

And then there is the broad tape, which has been flashing a signal worth respecting. There is a market tell going back to 1971: when the Nasdaq underperforms the Dow by around 5 percent over a short stretch of trading sessions, the odds of sliding into a bear market run near two-thirds. That is exactly the pattern lately. The tech-heavy Nasdaq keeps taking the losses while the Dow, with far less exposure to the AI names, keeps holding up or hitting records, session after session, as investors rotate out of the most speculative corners. The same nerves tripped a circuit breaker in South Korea, where the Kospi fell hard enough to halt trading, chipmakers leading the way down. The riskiest assets are being sold first. That is usually how these things start.

Two-thirds is not certainty. There are no certainties here, only probabilities, and a two-thirds signal is wrong one time in three. So this is not a call to buy puts on the Nasdaq or to bet the tape is done. Taking a directional swing at a market this reflexive is a fast way to be right on paper and broke in practice.

The honest position is narrower and more useful than a prediction. The circle only spins as long as fresh capital keeps feeding it and everyone agrees not to ask where the cash actually ends up. The line in the sand is the moment that question gets asked in earnest, when one marquee compute deal gets cancelled instead of renewed, when one of these money-losing segments has to show real cash instead of booked demand. Watch the renewals, not the announcements. Watch whether “excess compute” spreads from Meta to the rest of them. Watch whether SpaceX keeps making lower highs. The rockets may well be the future. The financing structure strapped to them right now is a different machine entirely, and machines like this have always ended the same way. The only thing nobody knows is the hour.

Facts > Narrative. Always.

Stand up for freedom, liberty, and free market capitalism.

Subscribe now

Subscribe now

Two weeks ago the Atlanta Fed’s real-time growth tracker said the U.S. economy was cruising at better than four percent. As of July 1 it says 1.2 percent. Same economy. Same quarter. Nobody moved the goalposts on the field... they moved the entire stadium.

And here is the part that should make you sit up. That 1.2 percent reading came out the day before the June jobs report landed. So the single ugliest labor number in this whole cycle is not even baked into it yet.

The mainstream take on all of this is going to be soothing, and it is going to sound reasonable. Growth is “moderating.” The labor market is “normalizing.” The unemployment rate actually ticked down to 4.2 percent, so how bad can things really be? That is the story you will hear on the financial channels this week, delivered in the same calm voice they use for everything right up until the moment it isn’t calm anymore.

There is a different way to read the exact same data. It starts with one simple habit that separates people who understand the economy from people who just repeat the headline: read past the first line.

The Atlanta Fed nowcast just fell off a cliff

Start with what actually happened, because the shape of it matters.

The Atlanta Fed’s GDPNow model is not an opinion. It is not a pundit with a necktie guessing at the future. It is a running arithmetic estimate of where the current quarter’s GDP is tracking, rebuilt from scratch every time a new piece of hard data hits the tape. No feelings. No narrative. Just the numbers as they come in.

Here is the ride it took this quarter. The model opened Q2 at 3.7 percent back on April 30. It climbed to a peak of 4.3 percent in mid-May. The sun was out, the birds were singing, the economy was supposedly running on all eight cylinders. Then it started sliding. Down to 3.3. Down to 3.0. Down to 2.5 by June 25. And then, on July 1, straight down to 1.2 percent.

Think about that for a second. In roughly six weeks, the most data-driven real-time read of the U.S. economy went from sprinting to barely standing up. A drop of more than three full percentage points. That is not a soft landing gently kissing the runway. That is the wheels coming off while everyone in first class is still being told to enjoy the flight.

So what put the wheels in the ditch? Two things, and both of them are worth understanding, because “GDP fell” is the kind of sentence that lets your eyes glaze over. The mechanics are where the real story lives.

Now, before we go under the hood, ask yourself the obvious question. If a model with no agenda, running on nothing but incoming government data, just cut its own growth estimate by more than two-thirds in six weeks... and if it did that before seeing the worst jobs data of the cycle... what is that model going to say next?

CONTINUE READING: below the paywall we crack open the two mechanical drivers that gutted the nowcast, we go to page two of the jobs report where half a million people quietly vanished from the employed column in a single month, we walk through the participation-rate magic trick that let the unemployment rate fall while the labor market fell apart, and we lay out the one number that tells you whether this is a wobble or the start of something that does not stop at 1.2 percent. Join us beyond the paywall, if you haven’t done so already.

Read more

There is a number in last week’s jobs report that should have been the headline. It was not the 57,000. It was not the unemployment rate. It was sitting two tables down in the same government release, and it said that half a million fewer Americans were working in June than in May. Half a million, in one month. And the story you were handed was that unemployment went down.

So take the attention off the Middle East, where it has understandably been parked for three or four months, and put it back where it belongs. Back on the labor market. Because everything comes back to how many Americans are actually working.

Share

The June jobs report: a stuck-in-the-mud number, then something worse

Start with the warm-up act. On Wednesday, ADP reported private employers added 98,000 jobs in June, against a consensus of 110,000. Not a catastrophe, but in an economy of 340 million-plus people, it is the sound of an engine that will not quite turn over.

Then Thursday. The Bureau of Labor Statistics put out the June employment situation and the headline establishment number came in at 57,000 jobs against a Wall Street consensus of 115,000. Roughly half. And to make sure nobody mistook it for a blip, the BLS revised April down 31,000 and May down 43,000, a combined 74,000 jobs reported to exist a month ago that quietly did not.

Here is the part that reads like a magic trick. With a jobs number that soft, the unemployment rate did not hold at 4.3%. It fell to 4.2%.

How do you print a bad jobs number and get a better-looking rate in the same breath? The rate did not improve. The denominator shrank. Labor force participation dropped three tenths of a point to 61.5%, the lowest since March 2021. People did not find jobs. They stopped being counted. To be counted as unemployed you have to have actively looked for work in roughly the last four weeks, so millions can want a job and not find one, and as long as they have stopped hunting, the headline rate looks better, not worse. The number can improve for the worst possible reason.

The number nobody front-paged

Now flip to the household survey, the one the BLS actually uses to calculate that tidy 4.2%. From May to June, the number of employed people in it fell by 507,000, and the civilian labor force contracted by 720,000. Read those against the headline. The establishment survey says the economy added 57,000 jobs. The household survey says half a million people stopped working and nearly three-quarters of a million left the labor force. Same press release. One got wall-to-wall coverage. The other got silence.

...already well below expectations of 115,000. But a second, less-publicized measure showed 507,000 fewer people were actually employed that month. The unemployment rate fell to 4.2% not because more people found work, but because 720,000 people stopped looking for jobs entirely. When people stop looking, they are no longer counted as unemployed...making the job market look better than it really is.

And look at what led the June losses: leisure and hospitality, down 61,000. That sector has been one of the main engines of job growth for two years, and it just swung to an outright decline. Leisure and hospitality is discretionary spending, the thing a family cuts first when money gets tight. When hiring there goes negative while the savings rate is getting hollowed out, that is disposable income telling you something.

From above 4% to 1.2%, and it does not even include this yet

Here is where the labor data stops being a standalone story. The Atlanta Fed’s GDPNow tracker for the second quarter estimated growth above 4% in mid-May. Real GDP, annualized. Sunshine and birdsong. Then it walked off a ledge. By June 25 it was down to 2.5%. By July 1 it printed 1.2%.

And here is the detail that should make you sit up. That 1.2% is dated July 1. The jobs report landed July 2. The worst labor print in months is not even baked in yet, and the next update lands July 7. When a model has already fallen from above 4% to barely over 1% before the bad data, you cannot sweep the direction of travel under the rug.

Why does the labor market drag the whole thing down? Follow the money. Fewer people working means less income and softer aggregate demand, which the model reads directly. But the second channel may matter more. Every paycheck flowing into a 401(k) is an automatic, price-insensitive buy order for the S&P 500. It does not ask whether stocks are cheap. It just buys. That passive bid props up asset prices, and propped-up asset prices prop up the economy, largely through AI capital spending. Fewer workers means a thinner bid under the index and less appetite from the mega-spenders to keep the buildout going. Pull one paycheck out and you are pulling a thread that runs straight into the largest single source of growth in the economy.

The stool nobody wants to look under

That AI buildout is the leg of the stool worth examining, because the way it is funded has quietly changed. When this started, the hyperscalers paid for data centers out of positive free cash flow. Not anymore. In early June, Alphabet, one of the great cash cows of the modern era, priced an equity capital raise of $84.75 billion to fund AI infrastructure. Google. Selling stock. A company that had not meaningfully issued equity in two decades, going to market for the largest equity raise in corporate history because the cash it throws off is no longer enough to feed the machine, and that on top of tens of billions in fresh debt over the prior year. When the profitable players are selling equity and debt to keep capex running, you have left the “funded by free cash flow” phase for the “funded by capital markets” phase. And capital markets stay open right up until they slam shut.

Which brings us to last week’s other tell. Blackstone’s data-center arm, QTS, walked away from its portion of what would have been the largest data-center campus in the world, a 2,100-acre project in Virginia. Be fair to the facts: the proximate cause was legal. Zoning approvals were thrown out on a public-notice technicality, a development partner had already exited, and QTS was left holding the full utility-upgrade bill alone. But a private-equity giant deciding the crown-jewel fight is not worth it, right as the profitable tech names start selling stock to fund their own buildouts, is at least worth a raised eyebrow.

Put it all together and none of it, alone, is a crash. Together it has a name.

Below the paywall we lay out the seven textbook signatures of a late-stage credit cycle and check every one against the tape right now, we explain why “no Great Financial Crisis” does not mean the cycle is dead and why the dot-com bust proves it, we tie the private-credit deterioration back into the labor and AI-capex threads, and we get to the part everyone actually wants, the one actionable trade this setup points toward and the single instrument that has historically paid when this stage gives way to the next. Join us beyond the paywall, if you haven’t done so already for this weeks Weekly Wrap-Up.

Read more

Subscribe now

There is a number sitting in the market right now that has shown up only a handful of times in fifty years, and almost every time it did, what followed was ugly.

Not a dip. Not a healthy pullback. A bear market. The kind that takes a fifth off the broad index and four-fifths off the high-flyers.

The signal flashed last week. The strange part is not that it flashed. It is where it points, and how few people seem willing to look.

The seven-day signal nobody was watching

Over the seven trading sessions through June 25, the Dow rose 0.5%. Over the same seven sessions, the Nasdaq Composite fell 5.0%. That is a 5.5 percentage-point gap between the two indexes in a single week.

So what? Indexes diverge all the time. The Dow is thirty old-economy names; the Nasdaq is where the tech lives. Of course they pull apart.

Except they almost never pull apart this hard, this fast. Only about 1% of all trading days since the Nasdaq was created in 1971 have produced a trailing seven-day spread wider than the one just recorded. That is a once-every-few-years reading. And the company it keeps deserves attention.

When the gap has been this wide, the market was in a bear market within three months 66.9% of the time going back to 1971. Round it and you get the headline: a 67% chance.

Now give that number a baseline, because a number with no baseline is just a number. Across the whole history of the series, the market has been in a bear market only about 24% of the time. One-in-four. When this divergence shows up, the odds nearly triple. That is the entire claim. Not a prophecy. A conditional probability that moves from one-in-four to two-in-three when one specific thing happens. And that one thing just happened.

The cleanest illustration still makes tech investors flinch. In the ten sessions right before the March 2000 dot-com top, seven showed a divergence as large or larger than last week’s. In the bear market that followed, the Nasdaq lost almost 80%.

Don’t shoot the messenger. This is not a forecast pulled from thin air. It is the historical record of what tends to follow a rare condition. The condition is present. The question is whether anything underneath the market explains why it would persist, which turns this from a chart trick into a story about how the whole thing is wired.

Why the divergence isn’t random

A healthy bull market is broad. The boring industrials, the banks, the chipmakers, the speculative moonshots, all grinding higher together. When the generals march one way and the troops collapse the other, that is not strength rotating politely. That is money leaving the riskiest corner of the market for the safest one that still counts as stocks.

And which corner is the riskiest right now? The one that has carried the entire index for two years. AI. The Nasdaq is where the AI trade lives, and so is the tech-heavy slice of the S&P. When the gap opens with the Dow firm and the Nasdaq getting hit, you are watching the market quietly re-price the most crowded, most story-dependent trade of the cycle. The divergence is not random. It is the AI trade and the rest of the market starting to trade like two different animals.

Which would be a footnote, except this week the world’s most cautious financial institution put out a report saying almost the same thing.

When the central banks’ central bank rings the bell

On June 28, the Bank for International Settlements published its 2026 Annual Economic Report. The BIS is the institution central banks themselves answer to, the one in Basel that does not do hysteria. This time it named names.

The report flagged the bursting of an AI bubble and the collapse of opaque “circular” financing deals as among the top risks to the global financial system. Combined capital spending by the five largest hyperscalers is on track to clear a trillion dollars across 2025 and 2026. And in the BIS’s own words, disappointment in AI returns could trigger a sudden pullback in financing and turn the capex boom into a protracted investment bust.

The buildout propping up the market is, by the BIS’s own account, financed in a way where the same asset can be pledged more than once. That is not a sturdy expansion. It works beautifully right up until the financing slows, then runs in reverse just as fast.

So you have two independent signals pointing the same way. A fifty-year price divergence from the chart. The central banks’ own institution from the fundamentals. Same fault line, opposite ends.

If both say watch the AI trade, the real question is not whether the indicator is spooky. It is what would have to break for the 67% to become the base case instead of the warning. One mechanism turns this from a market story into an economy story, and most of the people quoting the number have not connected it yet.

CONTINUE READING: Below the paywall, the circle the BIS is quietly warning about drawn out in plain English, the one company whose filings show the loop better than any chart, the single data release this week that flips the probability and it isn't the Fed, why a drop now hits the real economy harder than the dot-com bust ever did, and the one condition that would tell you this is the real move and not a two-week head fake.
Join us beyond the paywall, if you haven't done so already.

Read more

That is the whole story of last week, compressed into one line. The Fed’s preferred gauge, core PCE, climbed to 3.4% year over year in May, the hottest reading since October 2023. Headline PCE hit 4.1%, the highest since April 2023. On a screen full of inflation data running well above the Fed’s 2% target, with a brand-new chair who just told the entire market he intends to break that fever, there is exactly one trade that makes sense. Rates should continue ripping higher.

They did the opposite.

The two-year Treasury, the maturity that tracks where the market thinks the Fed is going, had spiked to its highest level since February 2025 days earlier on the hawkish Warsh pivot. Then the hot number printed, and instead of breaking higher it slipped. The ten-year did worse, falling to a seven-week low. And the dollar, which is supposed to follow the front end of the curve through the interest-rate differential, did not sell off the way the textbook says it should. It held firm, parked just above 101 on the index.

So which is it? Is inflation the problem, the way the headline insists? Or is the bond market looking at something the headline isn’t?

When the price action contradicts the narrative this cleanly, the price action is almost always the one telling the truth. A hot inflation number with falling yields and a firm dollar is not a stagflation trade and it is not a hawkish-Fed trade. It is the signature of something else entirely, something the financial press covered as four separate, unrelated stories last week when it was really one story wearing four different hats.

Pull on the thread and it leads somewhere most people aren’t looking. Not to the Fed. To a preferred stock trading at a record discount, a private-credit market quietly slamming its exit doors, and a rocket ship that already round-tripped.

Here is what the bond market saw that CNBC’s homepage missed, and why the single most important number you can watch this week isn’t the jobs report everyone is fixated on. It’s the one that printed last Thursday and is already trying to tell you where this goes next.

CONTINUE READING: Below the paywall, the four moves that only fit one story, why a hot inflation print sent yields the wrong way, the GDP “beat” that turns out to be the weakest 2.1% you will ever see, the rocket-ship IPO that is really a thermometer for the whole AI buildout, the private-credit funds quietly bolting their exit doors, the 11.5% “savings account” that just broke its promise, and the one condition that would tell you this is a credit rotation and not a two-day head fake.

Join us beyond the paywall, if you haven’t done so already.

Read more

Subscribe now

Gold just broke $4,000.

Not drifted. Not slipped. Broke. The metal that spent the last two years as the only thing anyone trusted fell more than three percent in a single session and printed its lowest level since November. From a record above $5,594 in late January to a four handle in front of a three. And every wire headline on the planet reached for the same explanation off the same shelf.

Rate-hike bets. The dollar firmed, the Fed turned hawkish, so gold got hit. Clean. Tidy. The kind of sentence an editor can run without thinking about it.

There is one problem with that story. The bond market was open the same day, and it was saying the exact opposite.

The headline writers grabbed the easy answer

Give the consensus its due, because it is not stupid. Gold pays you nothing. It sits in a vault and looks expensive. When the market believes the Fed is going to hold rates higher for longer, the opportunity cost of owning a lump of yieldless metal goes up, and money that was hiding in gold has a reason to leave. That is a real mechanism, and on a normal day it is most of the story.

And the rate narrative had cover. The Fed left rates unchanged at its last meeting but tilted its projections toward a hike, and the futures market followed. Traders now price roughly a 68% chance of a hike by September, up from about 29% a week ago. New Chair Kevin Warsh keeps repeating that price stability comes first. So if you are an editor on deadline staring at a gold chart that fell out of bed, “hawkish Fed, strong dollar, gold down” writes itself. Reach for the shelf, grab the sentence, file it, go to lunch.

The dollar half is even true. The Dollar Index pushed to its highest level in more than a year, north of 101. A stronger dollar makes gold more expensive for everyone holding euros or yen or won, and that pressures the price. No argument there.

So the dollar is ripping. Fine. But here is the question nobody at the wire desk stopped to ask: if this is a hawkish-Fed, higher-rates story, then why were Treasury yields falling?

The bond market called the bluff

Walk it through, because this is where the easy answer falls apart on its own logic.

If the move were really about the Fed getting more hawkish, the front end of the curve is where you would see it. The two-year Treasury trades almost entirely on what the market thinks the Fed is about to do over the next couple of years. Hawkish repricing means the two-year yield goes up. That is the whole mechanism. That is the tell.

The two-year did not go up. Across the same session the long end of the curve was falling, and the ten-year was falling faster than the front, the kind of move that pulls the curve flatter from the top down rather than the bottom up. When the back end drops harder than the front, that is not the market pricing in a more aggressive Fed. It is the market pricing in less growth and less inflation out on the horizon. The two readings point in opposite directions, and only one of them can be the reason gold sold off.

You cannot have it both ways. Either the bond market is terrified of a hawkish Fed, in which case yields rise, or long-term growth and inflation expectations are rolling over, in which case yields fall. They were falling. So whatever knocked the legs out from under gold, “the Fed is about to hike” is not it. The market’s own pricing of the Fed says so out loud, and the wire desk quoted the one number that was moving the other way.

Which leaves a more uncomfortable question. If it was not a rate-hike trade, what was selling gold, copper, silver and oil all at once while the dollar caught a bid and the long bond rallied? What single thing does all of that at the same time?

CONTINUE READING: Below the paywall, the four moves that only fit one story, why gold falling is the opposite of the calm the headlines are selling, the oil drop almost everyone is reading backwards, the housing number that quietly confirmed it, and the one condition that would tell you this is a trend and not a two-day head fake.

Join us beyond the paywall, if you haven’t done so already.

Read more

Subscribe now

Silver got taken to the woodshed on Tuesday while gold barely moved. Same metals desk. Same screen. One down 6%, the other down a percent and a half.

That gap is not a rounding error. It is the entire story, and almost nobody on financial television connected it to the 10% wipeout that hit South Korea the same session, the bid that ripped through the dollar, or the quiet thing the Treasury market did at the exact same moment.

Here is the part that should make the hair on the back of your neck stand up. On a single day in June, stocks sold off, silver sold off, copper sold off, and oil sold off. And the dollar went up. And Treasuries got bought. Stocks and metals and crude going down while the dollar and government bonds go up, on the same afternoon, is not a coincidence and it is not noise. It is a signature. The market leaves that signature when it is reaching for one thing and one thing only.

What was it reaching for, and why is the crowd on CNBC still treating a falling oil price like a green light to buy everything that isn’t nailed down?

The answer is in four moves the market made Tuesday, the same four most desks are still ignoring. Below: why silver fell six times harder than gold, what the 10% crash in Seoul says about the AI buildout propping up U.S. growth, the SpaceX reading that should worry the bulls, and the one signal in the dollar and bond market that points where this goes next. Plus the exact line in the sand to watch this week.

Join us beyond the paywall, if you haven’t done so already.

The two-faced coin: why silver fell six times harder than gold

Start with the metal that gave the game away.

Read more

Subscribe now

Everyone spent the spring waiting for the dollar to die.

Go back to April 2025. The dollar index had rolled over hard, sliding from around 108 down to 97, and the usual voices came out swinging. The gold bugs. The Bitcoin crowd. A few genuinely serious macro names. The dollar is crashing, they said, and for a few weeks the chart agreed with them. The greenback was finished. The end of dollar hegemony was finally here.

Fast forward to today. The dollar index just printed its strongest level in over a year, sitting right around 101 as of June 22, having climbed steadily for weeks while almost no one was watching. Not crashing. Climbing. And the same crowd that was throwing a funeral in April has gone strangely quiet.

So which is it? Is the dollar dying, or is it the strongest it’s been since the middle of last year?

It’s both, depending on which dollar you’re talking about. And that distinction is the whole game.

Two Dollars, And Only One Of Them Is In The News

Here’s the trap most people fall into. Mention that the dollar is “going up” and somebody immediately fires back that they just paid four bucks for a coffee, so what on earth are you talking about. Fair enough. That person is measuring the dollar against goods and services inside the United States, and on that scorecard, yes, the dollar buys less every year.

But there’s a second dollar. That’s the dollar measured against other currencies. The euro, the pound, and above all the Japanese yen. That’s what the dollar index, the DXY, actually tracks, and it’s roughly 57% euro and about 14% yen by weight. These two dollars can move in completely opposite directions at the same time, and they frequently do.

Now, the natural reaction from most Americans is to file this under “not my problem.” The DXY is just a number on a screen. It doesn’t show up at the gas pump or in the rent. Why allocate any mental bandwidth to it?

Because that number is a wrecking ball. And the demolition is happening right now, in plain sight, to the economies the United States trades with. To see why a strong dollar is doing real damage on the other side of the planet, and why that damage has a return address, you have to follow what it’s doing to one currency in particular.

Why The Yen Is The Canary

Pick Japan, not because it’s special, but because it’s the cleanest example of a mechanism that’s quietly stressing economies all across Asia and beyond.

Start with one fact almost everyone forgets: oil is priced in dollars. When a Japanese refiner buys a barrel of crude from Saudi Arabia, the seller does not want yen. You can walk into a Tokyo dealership and buy a Porsche with a suitcase of yen all day long, because that Porsche is priced in yen at the point of sale. Try the same thing with a barrel of Saudi crude and you’ll get the same look you’d get slapping a fistful of yen on the counter at an American gas station and asking them to fill the tank. They don’t take yen. They take dollars.

So picture a refiner. The price of oil in dollar terms hasn’t moved an inch. But the yen has fallen against the dollar. What happens? The number of yen that refiner needs to buy the same barrel goes up. Same barrel, more yen. If their cash flow can’t cover the gap, they have one option: sell something to raise the dollars. Sell yen out of the bank. Sell treasuries. Sell gold. Borrow more yen and dump it into the market for dollars.

And here’s the feedback loop that makes this vicious. Every time they sell yen to buy the dollars they need for oil, they add to the supply of yen and the demand for dollars. More yen for sale, more dollars wanted. You know what that does to the exchange rate. The yen falls further. Which makes the next barrel cost even more yen. Which forces more selling. The currency feeds on itself.

That’s not a theory. That’s USD/JPY pinned near 161 right now, hovering at levels not seen since 1986.

Watch Them Fight The Tide And Lose

Japan’s authorities aren’t sitting still, which is exactly what makes this such a clean case study in futility.

Every time the yen approaches that 160 line, the Ministry of Finance or the Bank of Japan steps in to defend it. How do you defend a currency? You take your stash of dollar reserves and you sell those dollars to buy yen, the mirror image of the refiner’s problem. You add dollars to the market, you add demand for yen, and for a moment the yen rallies. Back in April, one record-sized intervention briefly yanked the yen sharply higher. And then, like clockwork, it drifted right back to where it started. Every single time. It’s like watching a man bail out a boat with a bucket while the hull keeps taking on water.

So they tried the other lever. On June 16 the Bank of Japan hiked its policy rate to 1.0%, trying to narrow the gap with U.S. rates and make the yen more attractive to hold. The result? The yen kept falling anyway. A rate hike that was supposed to prop up the currency instead read to the market as a flare going up, a signal that says we are getting desperate and we are running low on ammunition. The market faded it instantly.

Why are they running low? Because defending the yen means spending dollar reserves, and the whole reason the yen is under pressure is that the country is short of dollars to begin with. They are burning the exact resource they don’t have enough of. You cannot drain the pond to put out the fire when the pond is what’s on fire.

Then there’s the subsidy. With the war in the Middle East pushing crude higher, retail gasoline in Japan spiked to record highs, and the government stepped in to cap pump prices near 170 yen per liter by paying refiners the difference. Sounds like relief. But ask the obvious question: what are they paying the refiners in? Not dollars. Yen. Which still has to be sold for dollars to buy the next barrel. The subsidy doesn’t fix the dollar shortage, it just moves it onto the government’s tab, and that tab is enormous. Reports peg the program at roughly 600 billion yen a month, with the reserves backing it measured in mere months of runway.

Cracks in the dam, and a finger jammed over each one. A thumb here, an elbow there. It holds, until it doesn’t.

The Sea Peoples Problem

So back to the American who shrugs and says none of this touches him.

Run the history. Around 1200 BC, a confederation of naval raiders the texts call the Sea Peoples tore through the eastern Mediterranean, wrecking coastal cities one after another and pushing inland in a way that helped tip the entire Bronze Age into collapse. They finally reached Egypt, and the pharaoh managed to lure their fleet into the narrow waters of the Nile delta and crush it. Egypt won the battle. Everyone celebrated. The invincible empire had beaten the unbeatable raiders.

And then, over the years that followed, the Egyptian economy unraveled anyway. Why? Because the Sea Peoples had already destroyed Egypt’s trading partners. You can win the fight on your own soil and still lose the economy, because no empire is an island. It runs on the web of partners around it. Take out the web, and the center starves no matter how strong its walls look.

That’s the dollar. The dollar is the Sea Peoples. Right now it’s strong, and that strength is the wrecking ball methodically knocking over the trading partners the U.S. economy depends on. The American who says a weak yen on the far side of the Pacific can’t possibly reach his wallet is the Egyptian behind the pyramids, watching distant cities burn and assuming the empire is immune. It isn’t all-interconnected as a slogan. It’s all-interconnected as a balance sheet. Damage your partners long enough and the damage comes home.

This is why the dollar index touching 101 is not a nothing-burger. It’s the single most underrated story in global macro right now, and it’s barely getting covered.

What Comes Next, And Where The Real Work Lives

Watch three things from here. Watch the 160 line on the yen and how hard, and how often, Tokyo has to fight to defend it. Watch whether the Bank of Japan gets dragged into more hikes it clearly doesn’t want. And watch the dollar index itself, because the higher it grinds, the more strain it loads onto every dollar-borrower and every oil-importer on the planet. The pressure doesn’t stay in Tokyo. It travels.

Here’s the honest part. This post lays out the mechanism. It does not lay out the trades, the levels, the specific dominoes, or what a real dollar reversal would have to look like before it’s worth acting on. That’s deliberate. That’s the line between the free desk and the paid one.

Free subscribers get the map: the framework, the why, the mental model that lets you see what financial TV refuses to connect.

Paid subscribers get the territory. Every week, the full breakdown of which economies are closest to the edge of this exact dollar trap and why. The specific levels to watch on the dollar index, USD/JPY, and the crosses that matter, with the reasoning spelled out, not just a headline. The “lines in the sand” that would flip this thesis, so you know in advance what would prove it wrong instead of finding out the hard way. And the way these macro currents translate into actual portfolio positioning, the part that turns a good story into a decision.

The free post tells you the Sea Peoples are coming. The paid desk tells you which walls are about to fall, and when to stop assuming you’re safe behind the pyramids.

If you’ve been reading along and nodding, you already think like a paid subscriber. Make it official. Upgrade, and get the half of the analysis that actually moves the needle.

Facts > Narrative. Always.

Stand up for freedom, liberty, and free market capitalism.

Subscribe now

Here is the problem with going back to sleep.

The thing that actually drives the price of oil over the next six months is not whether the deal gets signed this week. It is whether the barrels that stopped flowing during four months of war can be turned back on as quickly as they were turned off. And the people who know this industry cold will tell you the same thing, quietly, every time: they cannot.

So the real story isn’t the rally. The real story is what the rally is hiding. And if that delayed mechanism plays out the way the energy desk veterans think it will, the consequences don’t stop at the gas pump. They run straight through the dollar, through the Japanese yen, through the Federal Reserve’s next meeting, and into a PCE inflation report that lands this Thursday and could rewrite the entire second half of the year.

What does the market think it knows that it doesn’t?

Read more

Subscribe now

There is a new chairman at the Federal Reserve, and on Wednesday he held his first press conference. The rate did not move. It stayed pinned in its range of 3.5% to 3.75%, right where it has sat since last year, and the committee voted to keep it there 12 to 0.

So nothing happened. End of story.

Except that is not the story Kevin Warsh came to tell. He did not walk to that podium to announce a pause. He walked up there to announce a renovation. A policy statement cut to 130 words from 341. Forward guidance, gone. His own dot, withheld from the dot plot. And the part he clearly wanted in the headlines: five brand-new task forces to take a fresh look at how the Fed does its job. One on communications, one on the balance sheet, one on data, one on productivity and jobs in the age of AI, and one on the inflation framework itself.

That is a lot of motion for a day on which the only number that touches your mortgage stayed exactly the same.

Now, the charitable read deserves a fair hearing, because plenty of serious people are giving it one. The Fed has been wrong about inflation for years, the argument goes, so a new chair tearing up the old playbook and bringing in fresh thinking is exactly what the institution needs. Warsh himself has called for “regime change.” Hard to argue a place that missed this badly shouldn’t review how it operates.

Fine. So review it. But here is the question nobody in that room thought to ask, and it is the one that matters: review it against what, exactly? What is it the Fed has supposedly been doing all this time, that five committees are now going to do better?

Read more