AI Is Becoming a Stock Picker’s Market

Chris Hussey: This is The Markets. I'm Chris Hussey, and today is Wednesday, September 30th, and I'm here on the Goldman Sachs trading floor with Pete Callahan, who is our US technology, media, and telecommunications sector specialist within FICC and Equities.

Pete, thanks so much for joining us again on The Markets. 

Peter Callahan: Thanks for having me.

Chris Hussey: All right. Let's get to it, because we had kind of a meh summer. When you look at the 10 biggest tech stocks, half of them were up, half of them were down. Is the AI trade getting long in the tooth? What are customers telling you? 

Peter Callahan: Yeah, listen, I think tech and AI story is still intact, but you're right, it's, like, nuanced now, right? There's a lot more dispersion, which could be euphemism for bad market breath, right? But what you're starting to see now is much more of a stock picker's market. We're kind of year three, year four of the AI cycle. There's a lot of things to be excited about. There's also a lot of questions that investors are asking.

And so I do think the trend for tech is still higher. I still think earnings growth, valuations, and thematic tailwinds are on its side. But to your point, we can kind of go through these lulls after a really good start to the year, call it Jan through June. We've kind of been in consolidation mode this summer, and I think it's a fair thing as investors sort of wrestle with the handful of AI puts and takes down the stretch of the year.

Not to mention, of course, rates and oil and macro, and all of those are parts of the conversation today that were a little bit less so in the first half of this year. 

Chris Hussey: Yeah, I want to get into the macro, but before we do that let's keep exploring that AI trade for a little bit because is the complexion of the AI trade starting to change now? Originally this was an AI infrastructure trade. It started with the semis. It went into the hyperscalers. Is it now making the move into applications, things like agentic AI? How do you look at it? 

Peter Callahan: Yeah, it's a good question. I mean, let's just start. We had our big flagship technology conference early in September, and it's kind of hard to walk out of that conference and not feel good about the state of AI and the state of tech.

And so I kind of, like, that's the backdrop, and that's a good constructive backdrop. But you're right. For the first couple of years when people said AI, they just said semiconductors, kind of CapEx beta. We want to be picks and shuffles. In the last couple of months, you've really started to see more focus up the stack, some of the scaffolding at the infrastructure software, security software layer.

Some of the tools and companies that help organizations adopt and deploy AI, that's what you're starting to see. So yes, agentic commerce is part of it. Cybersecurity's been a piece of that. Some of these data infrastructure stocks have been a piece of that. So, I would say the trade has broadened out.

And in some capacity you could say, okay, well, it looks flatter and broader, but I think that broader breadth of the trade you could argue is a good thing, even if it creates a little bit more of a stock pickers rather than a sort of a single thematic narrative underneath of it. 

Chris Hussey: All right. Terrific, Pete. You mentioned Communicopia and Technology, our big conference that we had in September. Did you hear from a lot of people there that these types of AI implementations are working, and that you're getting those types of productivity improvements that people are looking for?

Peter Callahan: Yeah, you are. Listen, there's a bunch of soundbites that are coming out of corporates these days around productivity, sort of new revenue opportunities, things like that.

It's just the measuring stick, right? Because we're spending trillions of dollars on CapEx. And so I think the market writ large is getting what they want from a productivity narrative, from a use case narrative, now that's maybe emerging on the consumer side with the rise of these personal assistants.

But it's always about the measuring stick of how much is going into the ground versus how much companies are getting on the back end. And so of course, with three Q earnings around the season, we'll get this sort of numerical measuring stick in how companies are benefiting from generative AI.

Chris Hussey: Okay, Pete, expand on a little bit on the AI build-out here. What are investors most worried about there? Are they worried that we're going to run out of money and stop doing it? What exactly is the risk that they're focused on?

Peter Callahan: Yeah, it's a good question. I think most of it would be more about the linearity of the deployment. And so I think investors believe this theme is a decade long mega trend. And so I think investors feel pretty good about the end point, but the linearity has gotten a little bit more, uh, uncertain, right?

And that's a comment on sort of the deployment and the physical sort of standing up of these data centers and turning them on. And so I think that's the part of the conversation that's gotten trickier. And again, not to come back to it, but there's just a lot of noise when you look where the rates market is or the oil market is.

 and so there's just a lot more considerations that investors have today versus kind of that 2023, 2024 cycle of, of generative AI 

Chris Hussey: Okay, Pete, enough about AI. Let's move on to the broader tech space. Let's talk about software. That was a group that really got taken out earlier in the year. Had a nice summer though, but the last month it's been under a little bit of pressure. Where does software stand today?

Peter Callahan: You're going to catch me using the word nuanced again, dispersion, right? Because part of software is that cybersecurity story, that data infrastructure story, and that's become a really big theme for investors to latch onto, and you can kind of call those inference plays, if you will. The sort of core SaaS and application layer of software, the market's still wrestling with that stuff.

That's the market where the market's still trying to figure out product market fit and pricing and where they sit alongside that new model layer that's emerged over the last couple of years. And so, it's really kind of a dispersion trade within software, but that's a lot better. I mean, the first half of this year was kind of correlation one type of price action, and that's not great for anybody.

And so, it's nice to see dispersion. You're starting to see stock picking come back to software. But the trend has been higher. The group's treated much better over the last couple of months and I still think positioning's a tailwind in that subsector. 

Chris Hussey: So, if you're a software company that can help implement AI or keep AI safe, you're in good shape. But if you're a software company that has to use AI in order to make their software better, not so yet. 

Peter Callahan: Yeah or deliver that end product to the customer. That's where the market's trying to figure out how that story plays out over the next couple of months and quarters. 

Chris Hussey: Okay, let's cut the macro. Tech and rates have always been closely correlated. What's going on today?

Peter Callahan: Yeah. Listen, rates have definitely become part of the conversation and a consideration for tech investors. This year you've seen NASDAQ multiple compress about 20%. Now, there's a lot of things that drive that, but I think certainly rates is a piece of that. We now have 10-year yields certainly north of five and a quarter, right?

So I think that's been a piece of this story. Now, tech still generated great returns this year, and why is that? That's because the theme is intact, and the earnings growth is intact. And so we look at rates, and we're obviously careful if you sort of lose the long end of the curve. But from a, is it keeping people out of tech? No. But is it putting more pressure on earnings to do the heavy lifting rather than multiple? Certainly, and I think we've seen that this year. 

Chris Hussey: Now, you mentioned this about the earnings and the PE multiples having come in. Is that a symptom of healthy skepticism in the marketplace around earnings, that people just don't, they don't think the earnings are going to be as strong as we think? 

Peter Callahan: Yeah that's right. I think part of it's the rate story, and then of course, part of it is how long can we compound at this 20% plus earnings growth story, right? It's been a great stretch. If you get the productivity from AI in the coming years, you see corporate margins stay healthy, I think that story stays intact.

You climb that wall of worry on 2027 and 2028, and that's probably the path higher for the tech sector. 

Chris Hussey: All right, great. Let's put a bow on it. What's the trade? 

Peter Callahan: Yeah, I got two things in focus, right? One, of course, semiconductors. We talked about it, but that's a group that hasn't gone anywhere for a while.

Multiples have compressed, and we kind of talked about this sort of personal assistant, consumer AI story starting to pick up over the last couple of weeks. So I'll be watching for semiconductors into 3Q earnings to hopefully find its footing again.

And then the other one, consumer services and consumer experiences. Consumer stocks had one of their biggest months of underperformance in over 10 years in the month of September. Of course, rates and oil weigh heavy on the group, but you could argue that's an opportunity to climb a wall of worry down the stretch of 2026. And that'll be an area I would start to reengage and sharpen pencils, and find some good idiosyncratic ideas.

Chris Hussey: In the tech world, does that mean e-commerce? What do they mean? 

Peter Callahan: Yeah, exactly. Internet services, commerce, and then some of the live events and experiences that we have in the sector. 

Chris Hussey: All right. What are you watching for in the month ahead? 

Peter Callahan: Yeah, of course. So early month macro data, right? We have a Fed meet at the end of this month. Right now the market has about a 40% chance of a hike priced in.

So we'll see if the macro data tilts that one way or the other. We'll kick off corporate earnings in the middle of the month. Of course, all about finding that return on invested capital vis-a-vis AI, both in the tech sector, but of course outside of the tech sector as well, increasingly important. And then finally, we have midterm elections some 30-odd days away. That'll sort of be in the background as we head into earnings season, and so that'll be something to watch as well. 

Chris Hussey: Pete, thanks so much for joining us. Really appreciate it. 

Peter Callahan: Thanks for having me. 

Chris Hussey: That does it for this week's episode of The Markets. I'm Chris Hussey. Thanks for listening.