soup question, is it?

This is how I look at it.

The concentration risks here are awe-inspiring.

So if you take a step back and someone said to me, why don't you analyze this software company? Forget about what it does.

It's a software company, an established software company.

And if it turned out that the company had thousands of customers, That would be great.

If it turned out the company only had two customers, you'd say, I don't want to invest in that because if something bad happens to one of those customers, this company is dead.

I

Right.

Or, or creates soft- creates software, which is another form of words.

So, uh, my thesis as to why they've done this is that, um, Dario Amodei and Sam Altman right now are petrified.

They are petrified because token maxing is done.

So for those of you who don't know, token maxing was basically people who did, who run software and enterprises were ordered by their companies to use AI 24/7.

Right.

So I, I don't know if this is the...

I honestly don't know.

This is a message, though, that I think is starting to sink in, right? People need to understand the S&P is maybe not as diversified as it was five, 10 years ago.

Not even close.

So couple of statistics InfoTech is 40% of the S&P right now.

No sector has ever even been close to this level.

Um, then add on top of it companies that are not technically in InfoTech, but you would think of as being techy, which would be, uh, I'll just name a few.

Google, Meta, Amazon, uh, Bookings.

You add them all up, more than 50% of the S&P is, is tech related, is tech.

... cloud revenue

Would you not say that in 2008 you did have a, a, an understanding of what was priced into the market, specifically that there was a lot that, that wasn't priced in?

Oh, absolutely.

I, I learned this because I re- I remember when, um, Bear Stearns's, um...

They had, they had, like, this fixed income fund that blew up in May of 2007, and what that fund had done was invest in subprime paper.

And I remember there was a sell side analyst who I was very friendly who, um, he worked at a firm that was very heavily involved in subprime paper.

He was a financial services analyst.

And so he asked me what was going on, and I said, "Well, why don't you just get in the elevator, go down two floors, and go talk to your own desk?" A- and what I realized was that w- w- w- the reason that the, that the equity people didn't know what was going on, because it wasn't an equity story.

[laughs]

Mm-hmm.

The way I sort of see the entire AI story is that Nvidia sells chips to hyperscalers.

70% of hyperscaler AI revenue is from Anthropic and OpenAI, and that equates to about 25% to 35% of their cloud revenue.

If you look at just Oracle, um, 50% of its $600 billion backlog is just from OpenAI.

So in a sense, the way I think about this is the entire AI ecosystem food chain is dependent upon the future health and success of Anthropic and OpenAI.

That's...

And, and by the way, that also means that the US economy hinges upon the f- the, the health and future success of Anthropic and OpenAI.

Yeah.

A lot of things go bad, and-

I, I don't think this is where the problem is.

[laughs] Um, the hyperscalers are making lots of revenue, but their free cash flow is, has deteriorated dramatically.

But that, even that is not the issue.

The issue is, to me, that s- something like 70% of all AI revenue from Microsoft, Amazon, Google, Oracle, is from Anthropic and OpenAI, period.

Just those two.

And then if you dig down even further, it's 70% of AI revenue, it's 25 to 35% of their cloud revenue.

... whose financials until they go public, are hidden to almost everybody.

[laughs]

But you know, if you compare this to '08...

So m- Michael and I both had the same thesis, which was underwriting, mortgage underwriting standards have deteriorated dramatically.

And then what was good about the thesis was every single month, securitization data came out showing credit quality of hundreds upon... and, and billions upon billions of mortgages.

And so you could see, you could actually see the deterioration every single month.

You had a, you had a, a data set that was incredibly robust that came out every single month and said to you, "You're right" [laughs] that, that...

So, so you had this reinforcement.

Yeah.

Yeah.

It's, uh, [coughs] it's AI.

Either it'll succeed or it won't.

But, um, I mean, I just take for example, uh, you know, the investment banks all reported the numbers were really great.

But, you know, why were they great? They were great because they're doing a lot of AI financing.

So it's not like, you know, if you bought Goldman Sachs, you're diversifying your portfolio away from AI.

Um, I, I, I think the biggest news is that a y- a year-- If we were sitting here literally a year ago, the entire debate on AI would be, "It's all, it's all great.

How big?

Mm-hmm.

But, um, there's no question in my mind that, that the business has become capital intensive-No moats.

And the other thing that's happening behind the scenes is that until now, um, companies have been charging for tokens way below the cost of tokens, and that's actually starting to change.

Microsoft's GitHub is now charging actually for the tokens.

And if you go on the Reddit boards, you're seeing that there's a lot of pushback.

Um, so it's gonna be very-

I mean, I, I don't know.

Well, let me address it.

Mm-hmm.

You know, private credit funds were originally created for institutional money, and that makes a lot of sense because you're talking about long-term illiquid loans, and institutions know what they're getting.

They're getting a higher yield in exchange for less liquidity, and that's fine.

Um, after private credit basically sold their funds to every single institution on planet Earth, they looked around and they say, "Okay, now who do we sell it to?" And they said, "Let's sell it to retail." The problem is that with retail, you have to create l- liquidity.

So what they did was they created mostly what I like to call the illusion of liquidity or semi-liquidity.

Now, all this was disclosed.

None of this is illegal.

I guess the q- the, the final question before I let you go here is, you know, i- if this occurs, if the credit cycle, it does occur here, uh, how bad would it be? Uh, what would that recessionary moment look like, and how much better would it be compared to, say, 2008?

Are there any risks below the surface that you don't think people are-- risks that there's asymmetric, uh, downside in terms of the attention they're getting? What are the risks you see that people aren't pricing into the markets right now?

I think the two biggest long-term risks to the market by far are related to AI and related to private equity and private credit.

You know, in terms of AI, the risk that I don't think is realistic is, is some nightmare scenario where you're gonna wake up one day and everybody's gonna stop buying Nvidia chips.

I, I-- that's not possible, at least not any time within the next year.

And the reason why that's not possible is just because, you know, last year, all of AI infrastructure spend was four hundred and fifty billion dollars, and this year, if you just look at Amazon, Google, Meta, and Microsoft, just those four is six hundred and fifty billion dollars.

I mean, these are crazy numbers.

So, and I'm not trying to justify them, but I'm not spending the money.

[laughs]

Why don't we just say where things stand? How are you doing?

Thank you.

Thank you.

I appreciate being here.

So in June, I was diagnosed with breast cancer, which breast cancer is not rare.

But in a man, it's quite rare.

It's less than 1% of all cases.

which is against sort of the sam altman

Yeah.

So, the government prevented the worst.

Um, and, um, they did as, about as good a job as they could do with one major mistake.

I, I, I think the biggest mistake the Obama administration made was that nobody went to jail.

That was a terrible, terrible policy error.

And by the way, to this day, no one knows how it hap- why that happened.

The fraud, and this is the biggest fraud in human history, was this; when you're a Wall Street firm, you have a relationship with subprime lenders from whom you're buying mortgages to create a securitization.

Yeah.

I speak with many of the characters that I followed back when I was writing the book, people like the hedge fund manager Steve Eisman.

But after the meeting, like, you know, guys thought, said, "What do you think?" I said, "What do I think? Somebody just told me how to make a gazillion dollars.

I'm ready to go."

Yep.

And it has only 1 billion in equity.

So it's levered 100 to one.

For bank analysts, the, the, the, the, the real determination of how profitable you are is the return on assets, because that's all the money you're lending out.

The leverage is just math, how much equity you have to have against those assets.

So the first question you ask is, what's the return on assets of a bank? So Citigroup, which has underperformed for years, has a return on assets of less than 1%.

JP Morgan, which is a great bank, has a return on assets that's probably something like one and a half, 1.6% of assets.

Yeah.

am I going to ask

you this question? People want to be me.

They want to predict the end of the world.

And I have a news for all of them.

The position of Steve Eisman is taken.

By an optimist.

By an optimist.

Steve, anywhere in the equity markets that you see unusual exuberance, just something that maybe it's a concentrated trade in and around AI or something that you'd really want to keep dry?

But they do think it kind of shrugs it off next year and the economy is back and growing at a normalized pace.

Well, I think that's very true if there's no trade war.

That's why I say the situation is very binary.

So look, if we reach deals with all these countries and there's no trade war, I'm very positive in the U.S. economy long term, and I would be very positive on the markets.

But if there's a trade war, chances are we go into a global recession.

And it reminds me a little bit about all the books I've read in life about World War I.

Nobody wanted World War I.

when the

Yeah.

And I think th- what they all miss is that the entire f- like I hinted at before, the entire financial system of planet Earth runs on Treasuries.

You've got a multi-trillion dollar repo market where banks lend to each other overnight.

What is it? What's the currency? Treasuries.

You're- you're a sovereign wealth fund in Norway and you need to park your money in some medium-term notes.

What are you buying? You ain't buying Chinese bonds.

You're buying Treasuries.

Okay.

but that's

my point.

There's trillions upon trillions of repos happen in the banking system every single night on treasuries.

So if there was an alternative to treasuries, I might be worried more about the deficit because I'd say, well, if we don't balance our budget, then people will sell our treasuries and buy something else.

But what else are they going to buy? They're not going to buy Bitcoin.

It's not big enough.

They're not going to buy Chinese bonds.

It's the

Mr. Eisman says the US is well-poised to succeed in trade negotiations.

I think people need to realize that in terms of the negotiation position that the United States has versus other countries in terms of changing the terms of trade, the US is in the best negotiating position, by far.

And here's where facts and details matter.

Um, the percentage of US GDP that comes from exports is only 11%.

That's about the lowest percentage in the world of any significant country.Uh, if we go country by country, China officially is at 20%.

That's if you Google it, but that's really not accurate because China ships so much to countries like Vietnam and Cambodia that skirt, uh, US tariffs.

It's probably the least, 30%.

Mr. Eiseman says Mr. Clinton was right about GDP growth, but dead wrong about it would mean for American workers.

Mm-hmm.

And I made speeches all over the country to various kinds of organizations and to lots of universities.

And I got to travel, lots of, lu- parts of this country that I had never really traveled before.

And what I saw in the, uh, big parts of the South and the Midwest is that, you know, not only did people lose their jobs...... because of what Clinton had accomplished.

But their communities had been obliterated.

I mean, literally obliterated.

You, you would go through towns where half of Main Street is boarded up.

Mm-hmm.

Hm.

So I, I think what President Trump is arguing, and you can disagree or, or agree, I have great sympathy for it, is that we really screwed a whole, half the country because of free trade.

Now, it ushered in an enormous bull market with fits, you know, with some fits and starts, but generally since 1994 it's been a bull, a s- it's been a, a tremendous bull market where people like me and others of my socioeconomic class have made a lot of money.

But the other half of the country got obliterated, and we didn't even retrain these people.

We said, "You're on your own.

Go learn how to code," and they didn't learn how to code.

So I think he's, he's trying to achieve two things.

So what, what, in what ways do you think that this moment is similar or dissimilar to '08?

What do you make of this market here? You're saying don't be a hero.

What I make of it is that...

I think part of the problem, I've been watching your show, everybody else's show, and people are getting on the show, they're almost in hysterics.

And I think the issue is that everybody of our social class took Econ 101.

And we were all taught the same thing.

Trade good, tariffs bad, trade war terrible.

And that was drummed into everybody's heads with such force that it's a paradigm that now you have a president of the United States who doesn't seem to accept that paradigm.

Just thoughts.