Andrew Ross Sorkin on the 1929 crash

Summary:

1. 1929 wasn’t just a crash. It was a policy disaster.

The 1929 plunge was just the first domino.
What turned it into the Great Depression was a series of bad decisions: tariffs, bank failures, and letting leverage unwind violently.
25% unemployment, ~9,000 bank failures, and an entire generation that swore off stocks for life.

The ’20s were full of hype around automobiles, radio, and telecom. RCA was “the Nvidia of its time.” Everyone thought the future had arrived and price didn’t matter.

2. Today’s bubble is built on AI plus leverage

Sorkin’s point isn’t “AI is fake.” It’s: the economics haven’t caught up to the story.

Big Tech is plowing insane amounts into data centers and Nvidia chips.
Meta is spending almost all of its cash flow on this build-out.
Around that you’ve got real estate, construction, and especially energy companies borrowing heavily to power and house all this hardware.

Then you get circular financing like:

OpenAI “commits” $100B+ to Nvidia.
OpenAI doesn’t have $100B.

So Nvidia effectively finances OpenAI… to buy more Nvidia.

That’s not necessarily fraud. It’s just leverage stacked on top of narrative. When things break, it’s not the 20–30% price drop that kills you; it’s owing 3–4x that with borrowed money.

3. The scary part: we don’t even know where all the leverage is

In 1929 and even 2008, most credit ran through banks.

You could sort of see the risk.

Now?

Huge chunks of lending happen in “shadow banking” and private credit funds.
Pension funds hand billions to a manager for 10 years. That manager quietly lends into this AI arms race.
There’s minimal disclosure. We don’t really know what’s inside these vehicles or how they’re valued (6:15 in the interview)
And those private funds themselves often borrow from the banks. It’s all looped together.

If the cash flows don’t show up, you don’t just get a bad quarter. You get a credit crunch.

4. Nvidia is basically holding up the growth stats

Jason Furman (Harvard) ran the numbers in a recent study:

Strip out data-center / Nvidia-related capex, and U.S. growth is roughly flat.

So:

AI is real, like radio in the 1920s or the internet in the late ’90s.
But just like the dot-com bust, the technology can be fine while the pricing is insane.
The AI boom might be papering over deeper problems in the real economy.

5. Tariffs, Chinese EVs, and the American dream

Layered on top of all this:

Tariffs are now treated as permanent by a lot of business leaders. The Overton window has shifted.
Chinese EVs (BYD etc.) are, by Sorkin’s account, cheaper and better. Even Elon basically admits it — and says without tariffs, the U.S. auto industry is in trouble.
We’re choosing which industries to prop up and which to abandon, while simultaneously moving toward a services economy that AI can directly attack.

Meanwhile, most people don’t feel like they can get ahead through the classic “work hard, save, invest” route.

The American dream gets sold as a lottery ticket on TikTok instead of a 30-year compounding plan.

6. Inequality, tax code games, and philanthropy as a tax shelter

Sorkin also goes straight at the structure:

Estate tax loopholes that let huge fortunes pass untaxed.
Capital gains treatment that’s supposedly there to “incentivize investment” for people who would invest anyway.
Real estate and depreciation games.
Philanthropy used as a way to avoid ever realizing gains... effectively letting billionaires decide where untaxed capital goes forever.

His view: you could raise more money cleaning up the basic tax plumbing than by bolting on a wealth tax.

7. So what do you do as an investor?

The punchline is very “Flank”:

Over the last 100 years, it has paid way more to be a professional optimist than a professional skeptic.
If you have a long horizon, broad market exposure (indexes, not YOLO calls on the AI flavor of the month) still makes sense.
If you’ll need cash soon, you probably want 10–20% in something liquid so you’re not forced to sell into a crash (or follow Howard Marks' INVESTCON).
Time in the market still beats timing the market...but pretending there’s no bubble or no leverage is how people get wiped out.

This is why, as Flankers, we obsess over:

Real cash flows vs. narrative.
Balance sheets and hidden leverage.
Who’s actually financing whom, and on what terms.
Focus on what we can understand. Duolingo making learning more accessible has nothing to do with M2 Money Supply, Tariffs, Reshoring manufacturing, etc. We don't need to play 3D chess as Buffett style investors

A great Buffett to close: "I don't look to jump over 7-foot bars: I look around for 1-foot bars that I can step over."

Let's keep looking for those 1-foot bar, friends. Thanks for reading and being in our community... happy Sunday!