Episode summary
Chip Stocks Sell-Off: Why the AI Build-Out Just Went From 3 to 5 Years
Summary
Dylan Grice argues that last week’s chip-stock sell-off did not mean the AI trade was over; it reflected a market repricing of the timeline. He says the broader AI build-out is still intact because the economy remains strong, the jobs report was solid, and the recent weakness was driven by bond yields, systematic selling, and a cascade in fast-moving tech names rather than a true change in fundamentals.
His core claim is that many semiconductor stocks had been valued as if the AI infrastructure boom would be finished in about three years, by 2030, when it may now take five. That stretches expected earnings out over a longer period, which hurts near-term per-share results but can make the long-term earnings stream more durable. Grice says this matters most for companies without strong competitive positions, because they may capture profits only during the first few years of the build-out and then lose them later.
He contrasts those names with businesses that have durable moats, which he says can benefit from the same correction by giving long-term investors a chance to buy quality companies more cheaply. He frames the sell-off through Warren Buffett’s “Mr. Market” lens: when the market turns manic or panicked, disciplined investors should use the volatility rather than fear it.
Key takeaways
- Grice says the AI trade is not over; the sell-off was a valuation reset and a fast-money cascade.
- He argues the AI build-out may take five years instead of three, which stretches earnings and lowers near-term per-share results.
- He says companies without dominant competitive positions are most at risk because their profits may not persist into years four and five.
- Grice points to strong-moat businesses like Amcor as potential opportunities if the correction pushes prices lower.
- He recommends using “Mr. Market” panic as an opportunity rather than taking its mood swings as guidance.
Notable quotes
“I think the nature of the AI expansion or revolution has changed a little bit. The long-term picture, as I said on Friday, is still very much intact”
“there's a a near zero chance of the Democrats banning data centers. That's not how our system of government works, okay”
“Instead of a 3 year build out, for example, we're not talking probably about a 5 year build out”
“So if Micron was going to make, you know, $5 billion in earnings over the next 3 years, for example, it's now likely going to make that amount over the next 5 years instead of 3 years”
“what Warren Buffett says is you're free to listen to Mr. Market but you're a fool basically if you take advice from Mr. Market”
How the episode opens
Good morning. This is Dylan Jovanov with Behind the Markets. Happy Monday.
Today is Monday, November 9th. Today I'd like to talk about the most common question that I have been asked all weekend. Is the AI trade over?
What's going on in the markets? The election happened. First it went up, then it went down.
It's a crazy week. Oh my goodness. Then the jobs report came out strong, bond yields went up, markets sold off some more.
All the things, right? And I've actually seen some of our competitors very quick to take advantage of the confusion. You know, one of them in particular, won't mention their name, RIP AI.
It's all over, you know, and they're being very very dramatic. And look, here here's my take on this. I know I talked about this a couple times last week, but look, you do have to ask yourself when you have a week like we had last week, what changed?
You know, what fundamentally changed in the past week? You know, obviously, what changed in the past week? And you know, look, I think a couple of things changed.
I think the nature of the AI expansion or revolution has changed a little bit. The long-term picture, as I said on Friday, is still very much intact. The economy is still growing very very strongly, as I talked about last week.
I mean, really great job market now. It's really starting to show up in the job numbers, as we saw Friday. It's really starting to show up in US economic growth.
You're starting to see that thing. You're starting to really really see it. But what changed?
I mean, some friends of mine are like, the Democrats are going to ban data centers. And you know, this is something we've talked quite a bit about.
From the episode’s captions: roughly the first two minutes, lightly cleaned. Keep watching on YouTube →
Topics
- ai build-out timeline
- chip stocks sell-off
- bond yields
- competitive positioning
- data center policy
Companies and tickers mentioned
- MU
- AMKR
- ARM
- Micron Technology
- Amkor Technology
- Arm Holdings
Watch the episode
Watch “Chip Stocks Sell-Off: Why the AI Build-Out Just Went From 3 to 5 Years” on YouTube
Summary, takeaways and quote selection by Octagon. Quotes are short excerpts; the full conversation is in the episode.
More from Behind The Markets Podcast
-
Nvidia CEO Out: 4 Reasons I’m Not Panicking About Nvidia -
The Mag 7 Sell-Off: My 3 Secrets to Picking Stocks -
Morgan Stanley’s AI Winners List Is Full of Bank Stocks -
Google Stock: Why AI Mode Earns Less Than Search -
Q3 GDP: Data Centers Up 127%, But 62% Say the Economy's Bad -
AI Debt Bubble? Wall Street Is Done Funding Bad AI Companies
Behind The Markets Podcast, this episode and its artwork belong to their owners. Octagon is not affiliated with the show; this page is an independent summary.
