The Next Winners in AI Aren’t Chipmakers

Dear Reader,

The AI trade is shifting.

I’ve been waiting for this.

It’s moving from enablers to adopters, and we’re going to watch it play out over the next several months.

Who Has Won So Far

Up to now, the benefits of the AI trade have gone mostly to the enablers.

Power stocks.

Semiconductors.

Micron, Nvidia, Broadcom.

The companies that make AI adoption possible.

They’ve done very, very well.

That’s starting to change.

Sophisticated investors are asking who benefits most from here, how the ROI plays out, and how the economics work.

I think that’s part of what’s been shaking up this market and driving the correction we’ve had.

As the market finishes separating winners from losers, what keeps profits rising next year is how the adopters handle the rollout and the return on their investment.

The Timeline

Think about the timeline, and I’m thinking about my own business here too.

In the fourth quarter of 2025, AI really grew up.

It was a step function in ability.

That’s when we started using it aggressively, and we became very heavy power users.

All through this year, companies have been experimenting with how to use it.

I believe next year is when the bigger organizations start filtering through and changing their processes.

When you run a business, you’re always going for three things.

Faster, better, cheaper.

Do something faster and you make it less expensive.

Do something better and your return goes up and your customers are happy.

Do something cheaper and you lower your investment.

That’s how you stretch return on investment.

Who Benefits Most

So who’s best positioned?

In my view, it’s the adopters with oligopoly-like pricing power, proprietary data, and a physical or regulatory moat.

They adopt AI, and their margins expand up and down the income statement.

Here’s the list I’ve written down.

Insurance, underwriting and claims.

AI is built for a business like that.

Large banks and asset managers.

AI is built for that too.

Healthcare payers and scaled providers.

Same story.

Exchanges and data franchises.

I’ve been thinking about recommending and buying FICO.

And asset-heavy logistics incumbents.

Think FedEx and UPS.

They have paper-thin margins anyway, so anything that makes them more efficient adds a lot.

Not a business I would normally buy.

But those are the ones that seem most likely to benefit from AI adoption in the next year.

Where I’m More Skeptical

Professional services is the group I’m most skeptical of, and the story isn’t settled.

There’s a lot of debate over whether firms like McKinsey do anything.

I don’t know how that works out.

Those stocks are trading cheap, but I can’t figure out what they morph into over time.

That’s too hard for me, which is why I’ve been avoiding them.

Then there’s horizontal software.

AI saves time and cost, but it can also erode a software company’s moat.

I don’t believe that about the Microsofts of the world.

I thought that idea was laughable earlier this year when Microsoft sold off with everything else.

I think people feel safe using AI inside an established software environment like that.

The Bottom Line

Everybody has the enablers figured out.

The Nvidias, the Broadcoms, the Microns.

That’s all baked in.

Now the argument shifts, and we start to tease out the real return on investment.

There are going to be some great winners and some punishing losers.

I look forward to helping us avoid the losers and get toward the winners.

And for my #1 AI winner on the market today, I’d have to say it’s this one.

Have a great day.

I’ll see you tomorrow.

“The Buck Stops Here,”