
Double double toil and trouble
Cauldron burn and AI Bubble.
Part One: Nothing But Tulips and Beanie Babies In Here!
A week ago, before* more recent events, I came across two comments about AI, the economy, and investing which made me question whether people understood AI, economics, and investing. Both seemed like fundamentally incorrect statements being made, though they made opposite claims.
- There is no AI bubble.
- AI is propping up the entire U.S. economy
In this post and the next, I want to explain why these two statements are misleading, while attempting to limit financial jargon.
I should mention that the recent news about the head of the chip-providing company Nvidia being caught dipping his oar into Chinese waters complicates the picture, but only a little. Nvidia as a company is part of the contribution of AI to the bubble and to the economy, but Jensen Huang’s replacement as CEO and potential—I don’t know—arrest for treason? (in any other administration) doesn’t necessarily detract from my economic arguments. It just makes me laugh.

There’s No Such Thing as a Stock-Market Bubble
That’s the title of a November 7th opinion piece in the Wall Street Journal, which I read so you don’t have to. I read WSJ for the puzzles. Seriously! They have this maddeningly difficult “Meta” puzzle on Friday where you use the basic crossword to solve clues to get the secret, “meta” answer. I still think the one using hexidecimals was unfair, although we got last week’s, where you had to remove a color from the longest clues to get… well, anyway, I like the puzzles.
As many do, I skew politically left, though often not as far as YOU do—flaming moderate is part of my brand after all. So I find it a good practice to broaden my view by reading articles written outside my perspective, provided they aren’t outright lying *FoxNews coff*. Plus, blue people often ignore or misunderstand basic business and economic facts, so here is the MBA radical feminist who loves spreadsheets to the rescue!
The WSJ op-eds can be fraught with peril because they don’t always use FactFacts. But that provocative title—there are no bubbles—was too tempting to ignore. What kind of argument could someone make?
Every time the market falls, the pundits come out with the same diagnosis: It was a bubble, speculators lost their minds, and somebody should have stopped them. This time the pundits are early. They decided artificial intelligence is a bubble before anything had fallen. Really? Why would rational investors suddenly lose their minds?
Jeff Yass,There’s No Such Thing as a Stock Market Bubble
Are We Rational Investors? Are We Something Else?
Here’s the basic argument by authorJeff Yass. Investors choose stocks under conditions of uncertainty. They place bets or weigh probabilities as they make choices. For example, during the early 2000s, if you bought Amazon for $100, you made a lot of $$. If you bought Pets.com, you lost your shirt. The market, he says, might have thought Amazon and Pets.com each had a 15% chance of succeeding, so each investment (bet) was a rational choice at $100. “Losing a bet doesn’t make it irrational… When the fall comes, it won’t prove anyone was irrational. It won’t even mean the prices were too high. It will mean the uncertainty resolved, and the small chance of an even bigger payoff that was holding prices up didn’t come in.”
In MBA school, we were often given textbook problems that worked like this. Suppose so-and-so is choosing between Option A and B, with prices of $X and $Y and probabilities of success N and M? Which should they choose? “It’s just math.” It’s not math. The question is where do the probabilities come from? Who the hell knows what that probability is? And is that price reasonable and based on rationality in the first place?
This idea of rationality emerges out of mid-1970s financial theories, some based on the Efficient Market Hypothesis (EMH), for which Chicago professor Eugene Fama won a Nobel Price in 2013. The idea is that prices in the market are rational because information across the market is dispersed across the market. The market as a whole uses information to converge to useful prices. But that doesn’t work at the individual level for a buyer or a stock. It also doesn’t deny that bubbles can exist since investors might be individually irrational, even if the market as a whole is rational. As Fama explained in a recent Financial Times interview,
In other words, for every Pets.com there is an Amazon. Individual stocks might prove to be dumbly priced, but on average in the long run the cumulative efforts of millions of people trying to outsmart the market means that prices are more often fair than not. … Efficient market is a hypothesis. It’s not a reality.

Financial students are further taught that, in investing, a company’s price reflects its underlying business. Why would a rational investor buy into a company whose underlying business is not good? But here’s the trick. What’s the value of their future? Investors use potential growth as a gauge, especially when looking for the really big score. Money today is chasing future money, investing NOT because a company will continue in the success they have had but because that success will multiply. Yass argues that is not irrational and that “rational investors” do not “lose their minds.” This seems like semantics: all investors are rational, even those who make bad decisions based on the uncertainties, and therefore no one is really a speculator. What does history tell us about that?
Exhibit A: Tulips
Ah, the 1636 Tulip Bubble, which I’ve written about before. The short version of the story is that a Dutch noble (or two) decided that tulips were so pretty that they’d pay above market prices for them and, as influencers, they convinced others to follow in their footsteps. Suddenly, people with money wondered whether they could themselves make a killing by buying AND selling the tulips—buy low, sell high—and the price skyrocketed. Some made a lot. Then, when the market crashed, many lost their shirts. This also happened with shares of trading in the South Seas, the Mississippi, railroad stocks, and so on.

Speculation is a fancy MBA term for greed. Beanie babies. Toilet paper in the late spring of 2020. NFTs. Scarcity drives prices, while at the same time people are driven by the possibility that they could get a big payout for little work. Is greed fundamentally irrational? You certainly can’t argue that it doesn’t exist.
If a price of an ordinary thing increases tenfold for no other reason than that other people want to make money from selling it, that’s a bubble. You can debate whether it’s rational or not, describe it in bland terms as resolving uncertainty, etc., but it is still a bubble. Saying that speculators are a mix of rational investors who chose well and rational investors who chose poorly does not make them something besides speculators. When financial media start talking about the “excitement” around an IPO, they don’t mean because the company is a solid earner. They mean get rich quick, if you can somehow time the buy and sell properly.
Exhibit B: Excessive Concentration
Here’s the other way we can know bubbles are forming: concentration. Most money in the market is from a small amount of institutional investors, such as the Big Gorillas of Vanguard and Blackrock. Institutional investors are typically responsible for an estimated 80% of the market (that was 2017, maybe more now). These Big Gorillas are not speculators, but they do tend to put money where there’s already money.
Many funds now also are indexes, baskets that reflect everything in the market, like the S&P. The S&P 500 is weighted by market cap, i.e., the price times shares outstanding. But when price reflects assumptions about future growth, and some start thinking that future growth is massive, the price far outstrips a company’s current earnings. That’s exciting! For the S&P, those increasing prices mean that the market cap of a few companies start to dominate the market as a whole. The top 10 companies in the S&P are 38% of the entire 500 companies. And that’s what institutional investors are increasingly buying. Now we have concentration on both sides: a few institutional investors as the buyers and a few firms as the sellers.

At the time, those few large market cap firms are becoming increasingly intertwined. For example, OpenAI recently agreed to pay Oracle $60 billion annually for computing services, even though its revenue last year was $13 billion and its running at a loss. Oracle’s stock price (market cap) went up 40%. OpenAI bought a 10% stake in AMD, a chip maker, while Nvidia, another chip maker, invested $100 billion in Open AI. Microsoft is a major shareholder of OpenAI, though Microsoft is a big consumer of cloud computing from Coreweave, which is another company that Nvidia has a stake in.
As a historian, this reminds me of when the emperors and kings in Europe used to marry off all their sons and daughters to each other. It’s not quite incestuous, but you do start to have a risk when cousins marry each other. They’re trying to keep all the power, money, and influence “in-house,” but there’s a risk when it becomes over concentrated.

Bubbles Don’t Apparently Exist, Says World’s 25th Richest Man
I always think arguments should always be weighed on their own merits. We should kick the tires of their logic, rather than dismissing the What because of Who wrote it. But Who can be relevant, so I’m going to include it as my last point. It turns out that Jeff Yass, who constructed this opinion piece, is a bit of a Wall Street celebrity. As one of the commenters on his articles put it:
from a Vampire Squid 2.0 outfit like Susquehanna no less…
Yass made his fortune running a trading firm called Susquehanna International, bankrolling his stake from his winnings as a professional poker player while in college. Yass is reportedly the richest man in Pennsylvania, 25th richest in the world. Vampire Squid refers to what Rolling Stone journalist Matt Taibibi called Goldman Sachs in 2009, “a great vampire squid wrapped around the face of humanity, relentlessly jamming its blood funnel into anything that smells like money.”Sounding more familiar? Yass bankrolls MAGA causes and school choice… ding ding ding.
It’s not that uber-1%-ers can’t make valid economic arguments. But if the argument is that investors need not worry about what’s happening to their money in the market (because the market is a rational place where uncertainty gets resolved) and when that argument is made by someone who has made $78 billion correctly resolving uncertainty in his favor at the expense of other investors who played the probabilities incorrectly, that does tend to make me skeptical about this parsing of the word “speculator.” It’s all about uncertainty and probabilities until it’s my 401K holdings at Vanguard, at which point, I would like the fund managers who are choosing on my behalf to avoid placing risky bets and I would prefer it if the S&P 500 was not suddenly so heavily dominated by a handful of these unresolved uncertainties.
It’s all fine and good if you are a billionaire gambler, who can afford to lose a few million on Groupon if you strike it rich on Facebook (both debuted about the same time). But if 38% of the S&P 500 index, the cheap, simple, diversified fund that makes up a lot of people’s financial holdings, is based on around a few tech stocks, all of which are driven by a single thing—AI… well, then, the consequences of uncertainty resolving itself if AI does not pay off will become large.
I don’t know if this is a bubble, but I do know greed when I see it.
Next post: Part Two–Is AI Driving the Economy?


