| I. | This Week's Story |
You know the feeling.
You're at a dinner, or a barbecue, or a group chat. And someone won't stop talking about how much money they've made in the market. Their account is up 40%. They bought the right stock. They tell you it's not too late to get in.
You go home that night and you do the math on what you'd have if you'd just bought what they bought. You feel a little behind. A little foolish for sitting in cash. So the next week, you put more in. Everyone's doing it. The market only goes up.
That feeling — the one where you can't stand to watch everyone else get rich without you — has a name. It's the reason markets go to extremes. And right now, almost everyone has it.
American households have more of their savings in stocks than at any point in history. The "dumb money" confidence index — which tracks what ordinary investors are doing — sits at 0.8 out of 1. Near the top of its range. Everyone is in.
But here's what should give you pause.
The people who manage the most money — the institutions, the pros, the ones with teams of analysts — are doing the opposite. Their confidence index is at 0.3. Near the bottom. They're heading for the exits while everyone else is still buying tickets.
And one fund tells the whole story. Last quarter, 40.7% of investors in Blue Owl's $36 billion private credit fund tried to pull their money out. The quarter before, it was 5.2%. In three months, the share heading for the door went from 1 in 20 to nearly 1 in 2.
This is what it looks like when the money starts to leave. Not with a crash or a headline. It starts in the places most people aren't looking… and by the time it reaches the dinner table, it's usually too late.
| II. | The Divergence |
| Retail ("dumb money") | 0.8 |
| Institutions ("smart money") | 0.3 |
| 0 · All Out | 0.5 · Neutral | 1.0 · All In |
Eighteen months ago, the pros and the public felt about the same. Cautiously optimistic. Both lines sat near the middle.
Then they split.
The public kept buying. Their confidence climbed toward the top of the range. The pros did the opposite — they sold into every rally, and their confidence fell toward the bottom.
The two groups now sit further apart than at almost any point in recent history. They can't both be right.
| III. | The Anomaly Score |
|
82/100
Dangerously Unstable
|
Up from 78 last week. The private credit crisis deepened, the smart-money divergence widened again, and three of our four readings moved the wrong way. |
| 0 · Normal | 50 · Unusual | 100 · Extreme |
|
93
Concentration
|
88
Valuation
|
65
Sentiment
|
82
Credit Stress
|
| Concentration Up from 91. The top seven stocks gained another 1.2 points of the index. The market keeps getting narrower. | Valuation Up from 84. Stocks are priced higher relative to the economy than at any point on record — past the dot-com peak. |
| Sentiment Up from 62. Retail keeps buying every dip. Institutions sell into it. The gap is the widest in years. | Credit Stress New this week. Private credit redemptions exceeded new investment for the first time ever. The money is leaving. |
| IV. | Where the Money is Going |
This connects directly to the private credit story.
According to S&P Global, corporate debt maturities jumped from $2 trillion in 2024 to nearly $3 trillion this year. Most of those loans were taken out when rates were around 4%. Now they have to be refinanced at 6.5% or higher. For some companies, that doubles their interest payments overnight.
Throughout 2025, corporate treasurers ignored this. They assumed rate cuts were coming. But the 10-year Treasury climbed toward 4.5% this spring… and the cuts didn't come fast enough.
Goldman Sachs and Morgan Stanley have already postponed bond offerings that were scheduled for March. When the banks that sell the debt are stepping back, that tells you something about how easy it is to refinance right now.
Apple and Microsoft will be fine. They have cash. But the companies that borrowed heavily and don't? They're the ones showing up in those PIK loan numbers… adding debt to the pile instead of paying it off. That share has doubled since 2022.
It's all connected. The private credit freeze, the rising defaults, the debt wall — they're not three separate stories. They're the same story.
|
$2.0T
|
$2.5T
|
$3.0T
|
| 2024 | 2025 | 2026 |
And here's where it spreads.
$21.3 billion in office loans come due before the end of 2026. Of those, 83.7% are already delinquent. And 92.7% have been handed off to workout specialists because the borrower can't pay.
Office vacancies are still at 18.5% nationally. That's near the highest ever. The work-from-home shift didn't reverse. The loans were taken out at 4% in 2015. They're maturing into a world where rates are 6.5%.
But here's the connection. The total commercial real estate maturity wall is $539 billion this year. And $550 billion next year. Private credit funds have been a major source of CRE lending over the past few years. And those same funds are now freezing redemptions and pulling back.
So the question becomes: if the lenders themselves are under stress… who steps in to refinance the buildings?
Meanwhile, something unusual is happening with central banks.
The Fed is cutting rates. But five other major central banks — including the Bank of Japan and several in Europe — are expected to raise them this year. That hasn't happened since 2004.
Why does this matter? Because when central banks move in opposite directions, money moves too. Capital flows toward higher rates. If the rest of the world is tightening and the US is easing, dollars flow out. The currency weakens. And all those foreign investors who have been buying US stocks and bonds start doing the math on whether it's still worth it.
Right now, US stocks make up 62% of the global stock market. That's an extraordinary concentration of the world's savings in one country. It only works as long as the money keeps flowing in.
If the central bank divergence causes even a fraction of that money to rotate out… the effects on US asset prices could be severe. And it would happen at the worst possible time — right as the debt wall hits and private credit is pulling back.
| V. | What Else We're Watching |
Three more things worth keeping track of…
For 40 years, the rule was simple: when stocks fall, bonds rise, and your portfolio holds steady. That's the whole idea behind the classic 60/40 mix. But it stopped working. Since 2020, bonds have failed to protect investors in 17 of the 19 months when stocks fell 2% or more. This March, both fell together again. BlackRock — the largest money manager in the world — just declared the 60/40 portfolio dead. In other words… the thing that protected ordinary investors for two generations no longer works. And most people holding a "balanced" portfolio have no idea.
They've started cutting how much crude they process, to conserve supply. Jet fuel prices in Asia spiked before crude oil did. Product markets tend to lead. Consumers don't use crude — they use gas, diesel and jet fuel. When those prices move first, it tells you the supply chain is tighter than the headline oil price suggests. That's an inflation problem hiding in plain sight.
The insider buy/sell ratio is down 17.5% year-over-year. The people who know their own companies best — the CEOs, the CFOs, the board members — are selling more and buying less. It fits the pattern of everything else this week. The people closest to the money are stepping back. We'll see who's right.
| VI. | $800 Billion in Circles |
Here's something strange.
Nvidia agreed to invest $100 billion in OpenAI. OpenAI then turns around and spends that money… buying Nvidia chips.
AMD handed OpenAI the right to buy 10% of AMD at a penny. In exchange, OpenAI agreed to run its operations on AMD chips. Oracle signed a $300 billion deal to provide computing power to OpenAI. To do it, Oracle is spending tens of billions… on Nvidia chips.
Do you see the pattern? The money goes in a circle. The chipmaker funds the startup. The startup pays the chipmaker. Round and round.
All told, there's now more than $800 billion tied up in these circular deals. And every time a dollar makes the loop, it gets counted again as revenue. It looks like booming demand. But a lot of it is the same money, passing the same hands, wearing a different name each time.
In 1999, telecom companies like Lucent and Nortel lent money to their own customers so those customers could buy Lucent and Nortel equipment. It made revenue look incredible. The stocks soared. Then the customers couldn't pay it back. Lucent fell 99%. Nortel went bankrupt.
The financial press has a polite name for this. They call it "vendor financing." There's a less polite name for what happens when it unwinds.
I'm not saying the AI companies are frauds. They're building actual technology. But when the demand you're paying for is partly demand you funded yourself… you have to ask how much of it is genuine demand. And right now, the entire stock market is leaning on the answer being "all of it."
That's a big bet. We'll see.
