tl;dr: Listen to AI is Changing Markets, Not Just Companies | Treussard Talks (E28) on Apple Podcasts, Spotify, and YouTube.

On September 7, 1982, Joe Strummer leaned into the microphone at the Orpheum Theatre in Boston, in front of a few thousand of his closest friends, and said the following before launching into a fiery rendition of "Know Your Rights:"

You already know I quote Joe Strummer too often for it to be a healthy habit.

But here we are again.

The latest episode of Treussard Talks spells out how AI is impacting markets before our eyes. And yet, what's pushing and shoving investors around as part of the process feels very deeply human indeed. Outside of the Iran war, this is "the market theme" of 2026. Give me five minutes and I'll walk you through it.

There are very few things less convincing to me than the headline “Person who predicted previous catastrophe is warning about a potential new plague.”

Too many things cognitively wrong with that, starting with “even a broken clock is right twice a day.”

And yet, there has got to be a reason why my first boss Alec Crawford, who was Head of Risk Management at a multi-billion-dollar family office back in 2008, had every one of us read Charles Kindleberger's "Manias, Panics, and Crashes: A History of Financial Crises" when we showed up to the office for the first time.

Learning how past stories have connected helps us have a shot at connecting stories in real time, as we proceed through life.

This piece does that.

We called this week's episode "AI is Changing Markets, Not Just Companies," with econ student Kiara Galvin serving as our Teaching Assistant for the day.

This is NOT a "new plague" piece. I don't know what happens next, and you deserve better, anyway.

Let me take you to the Situation Room.

Earlier this year, we talked about something called SaaS-mageddon in Gradually then Suddenly.

This is one of the figures from that piece, updated through earlier this week.

The short version is that software companies have done very poorly this year.

Which is a big deal, because for a long time, software companies were market darlings (in public and private markets).

The reasons are simple. In the modern era, software companies have benefitted from scale (build once, sell infinitely), a subscription revenue model (clients pay over and over again), and a technological moat (some people know how to code, others don't). Put together, that's a lot of pricing power.

AI changed all of that — or at least, that's the market's reaction to the technology. The moat is crumbling every time someone fires up Claude Code to create an alternative to a legacy SaaS product, and that's threatening the legacy players' pricing power "on every floor of the building."

The narrative has made sense, as far as you can take those things, and the market has by and large supported the thesis.

In short, meet the "losers of the AI trade"… so far. Nice, clean, fits on a bar napkin (or a slide in a pitch deck for an AI-minded hedge fund). We're going to call this "the shorts."

The longs have been tougher. OpenAI is not public. Anthropic not public.

Sure, you need chips for AI, so Nvidia and AMD became "Artificial Intelligence stocks." But how much of a couple of names can you own without starting to look "concentrated" (aka, fresh out of good ideas)? So people have looked beyond "Big Chips."

The hyperscalers (the Metas, Amazons, Googles, and Microsofts of the world) have made some sense but are messy businesses causing all sorts of doubt. In market parlance, they are not “pure plays” on AI. What happens if people stop "googling stuff," wouldn't that be bad for Google? What if Microsoft's office software gets displaced by Claude Cowork? You get the point. Who wants messy uncertainty in a world desperate for comforting narratives…

So we've had "other chips." Chips down the supply chain. Memory chips in the US, sure. But also chips companies outside of the US. Especially in Korea. We talked about that last time, in The First Rule of Fight Club.

And this is how you end up with this headline on Bloomberg: "Why South Korea's Stock Index Is More Volatile Than Bitcoin" (Vishnoi and Cha, July 31, 2026) and the graph that goes with it.

Things are weird when a national stock market becomes more volatile than crypto.

The reasons are mechanical — analog, even — and they start with leverage. (And yes, a wall of money chasing a niche theme in a relatively small and concentrated market.)

It turns out the Korean market saw a rash of leveraged ETFs (funds that functionally borrow money to make bigger bets) hit the scene just in time for this. That's how you end up with this figure from last time.

Leverage is the classic lever by which good outcomes feel great and bad outcomes can cause ruin.

As luck would have it, leverage wasn't limited to retail funds in Korea.

I know, shocking, right?

Enter Situational Awareness… a hedge fund started by an objectively brilliant former OpenAI employee by the name of Leopold Aschenbrenner.

The shortest version of the story goes like this.

Aschenbrenner writes a piece called "Situational Awareness," spelling out his case for why AI is about to change everything.

People who are really into AI think this guy has "got it" (the plot, I assume).

Leopold is rapidly convinced that running a hedge fund beats working in tech.

And so, he launches an AI-themed hedge fund, named for the essay.

He does all the things that we've discussed above.

He shorts companies that he deems to be doomed by AI, and invests on the long side in the companies that he deems to be positioned to do really well as a result, and he levers up the investments — as one does in the hedge fund space — using borrowed funds from investment banks (largely headquartered in New York) whose business it is to lend to hedge funds, which we call "prime brokers."

It all goes well for a while. Very well indeed, including returning more than 400% recently. As you will recall, leverage will do that. It will turn good outcomes into great ones.

Until the machine goes into reverse. Which it did, in late July. (Go see the graph above for SK Hynix, to see what I mean.)

At which point, Situational Awareness started experiencing investment losses that chewed up enough of the investor capital over a matter of days that their prime brokers started getting anxious.

And when those people get anxious, they do what they have to do to protect their balance sheets.

It starts with calling you for more cash, which is what we call a "margin call." And if you can't come up with the cash, the next thing that happens is your positions in suspected "AI winners" and suspected "AI losers" get liquidated to take down the risk and raise the cash.

That's pretty much the worst-case scenario if you're in Aschenbrenner's line of business. And I'll spare you the details here — since you can hear more on the podcast — but that's how you reportedly start the week overseeing $45B of assets and end it with $10B.

Why do I tell you this?

I am telling you this because all of these stories are deeply related, and the commonality is that markets are trying to make sense of what AI means.

For companies. For the economy. For people.

That's a big job… It's also one that is starting to show signs of dysregulation. Things like national markets being more volatile than crypto, or fancy hedge funds suffering violent AI-trade reversals in a matter of days.

In and of themselves, these are not small events. And in this case, they fit into a much bigger set of questions and circumstances.

Like I told Kiara on the podcast, we squarely don't know what it means to be "right about AI" or "wrong about AI."

In fact, someone taught me this Wall Street adage: "Being early and being wrong often end up looking like the same thing in the end."

Having started my career in risk management in the face of a major market dysregulation, I can tell you one thing: trying to avoid existential risk along the way is a very solid place to devote time and attention.

A large part of the job is: "survive really bad outcomes."

And that may be the most important lesson for investors right now.

Special disclaimer

This piece is educational commentary. It is not investment advice, and it is not a recommendation to buy, sell, or hold any security.

Every company, fund, index, market, and security named here is named for one reason: to illustrate how markets have been behaving. Not one of them is a recommendation. Not one of them is a forecast. Not one of them reflects a view on what any reader should own, avoid, add to, or sell.

Nothing here is a prediction. Where this piece describes returns, asset levels, drawdowns, or price moves, it is describing what has already happened. Past performance does not predict future results. That an investment returned more than 400%, or that a fund's assets fell sharply over a matter of days, tells you nothing reliable about what either will do next.

Nothing here is personalized. It does not account for your portfolio, your concentration, your leverage, your tax situation, your liquidity needs, your time horizon, or your tolerance for loss. Speak with your own adviser about your own circumstances before acting on anything you read in a newsletter — including this one.