Executive Vice Chairman, IBM Common Stock
Former NEC Director Gary Cohn: The market would be floundering without the AI energy trade
Gary Cohn, IBM vice chairman and former Trump NEC director, joins 'Squawk Box' to discuss the latest market trends, what's driving the tech selloff, state of the AI boom, state of the economy, and more.
About Gary Cohn
Gary Cohn, vice chairman of IBM and former director of the National Economic Council under President Trump, has appeared on multiple CNBC programs in recent months to discuss the economy, Federal Reserve policy, and the impact of artificial intelligence. In a June 2026 appearance, Cohn said that without the AI and energy sectors, the stock market would be "floundering" and described the two industries as "intertwined." He stated that he believes computing capacity will be overbuilt and will become a commodity, with companies purchasing it from the lowest-cost provider. Regarding AI's effect on employment, Cohn said he is "in the camp that this time is the same" as past technological advancements, arguing that such innovations historically have not led to the "demise of human capital" but have instead grown GDP and created more jobs. Cohn also commented on economic policy and inflation. In June 2026, he said that if a deal to open the Strait of Hormuz is signed, oil prices would "not... fall like a rock overnight" but that a change in psychology could lead to lower prices over time. On the Federal Reserve, Cohn said that new Chair Kevin Warsh "will remove himself from the political pressure" and "do the right thing economically," adding that the Warsh Fed "will look different than the Powell Fed" with less forward guidance and data releases. In July 2026, when asked about IBM's software business, Cohn affirmed that the company's software is "not being disrupted by AI," and noted that companies are beginning to evaluate the return on investment of their AI spending.
Transcript (16 segments)
B
Becky0:09
All right. From Kevin Warsh, his first Fed meeting as the chairman to the Iran deal. Impact on energy prices. We've got a lot to talk about. Joining us right now is IBM's vice chairman, Gary Cohn. He served as director of the National Economic Council in the first Trump administration. He also happens to be the vice chairman at IBM as well. And that's one of the things I want to talk to you about today, Gary, watching what's happening with the markets this morning. I think these are questions about AI and where we go from here. Why don't we start with that this morning.
G
Gary Cohn0:40
Look, Becky, first, thanks for having me. I think it's an interesting question, one of which we've talked about on this set before. We have this really unique phenomena going on where we're taking companies that historically have been IP heavy and asset light, meaning that they own something that generates enormous amount of cash flow. And they've used that cash flow to acquire other assets that generate cash flow. And they've used it to grow and create dividends for their shareholders. We now have those exact same companies that have worked for decades to become asset light, and they have chosen to become asset heavy, asset heavy to the point where they're issuing new stock, they're issuing new debt, and they're no longer cash flowing companies. You know, we have to sit back and look at this and say, look, does this make sense to take companies that generate huge amounts of cash flow now that do not generate cash flow? And then we also have to evaluate how are we valuing those companies? We knew how to value them as growth cash flow generating companies. How do we value them as asset heavy companies that are now taking on debt, that have a higher debt service charge, that never had a debt service charge before, that will grow potentially in the future, but they're investing today for growth in the future. I think that may be a different valuation metric.
B
Becky2:01
So they've gone from super high growth companies to not utilities, but something that maybe requires a different overall measure.
G
Gary Cohn2:09
Well, we don't know if they become utilities. You know, the question is, does this whole AI infrastructure and the LLMs themselves become more of a utility infrastructure play? We don't know because it's evolving out over time here. We also don't know what the pricing is going to look like. What are LLMs going to be worth? What are you going to pay for an LLM? Is it a commodity? Is someone going to come in with a newer LLM and undercut pricing? Are you going to buy the software separate from the compute? Are people going to compete for compute? We see that now. We see the Chinese coming in and undercutting the price of compute pretty dramatically. So all of these questions have to be asked. And what's really important behind the scenes is if you strip out AI and energy from the stock market this year, we have a down market. The really the two businesses that are driving market and market value is AI and energy. And they're both intertwined with each other.
B
Becky3:07
Okay. There's a lot to dig through on that. Let's start with just where the huge declines are today. Most of them are chip stocks, not necessarily the Alphabets and the Metas, although we did.
G
Gary Cohn3:19
See, although they may have been yesterday, that might have been yesterday's move.
B
Becky3:22
We see them moving some of those things not 9% declines like we're seeing with some of these stocks today. But you did see what was it, three and a half 4% move lower for Alphabet and some of the others. Guys, if you can call up the Mag-7. Let's take a look at Meta. Let's take a look at Alphabet. Let's take a look at some of these stocks that are. Yeah, here we go. Apple, Microsoft, Meta, Tesla. So yesterday we saw some pullbacks from these but not the same magnitude of some of these issues. What I did not realize is if you stripped out AI and stripped out energy that the market would be down for the year.
G
Gary Cohn3:53
Torsten Slok put out a piece on that last week. You know, very interesting to really break down what's driving the market. And look, we're always like this. You know, last year every guest sat here and talked about the Mag-7, what the market would look like without the Mag-7. You know, maybe the Mag-7 is going away. We had FANG before that. We've had BRICS before that. We've had lots of different clarifications of what drives the market. Right now it is AI energy trade. Without the AI energy trade, the market is sort of floundering.
B
Becky4:19
What you do find, we had Philippe Laffont here earlier today, and what you do find is that people are trying to chase the money that those big spenders are putting out into the infrastructure, right? You got to go down to the next level or maybe two tiers down to figure out where that money is eventually going to flow. Those could be the plays for what comes next. But are you questioning the idea of AI spend overall? Because I think that would be the issue. That could really be a major, you know, I don't know if you want to call it a hiccup or something much larger than that. If those companies decide that they're not going to continue to spend like they've been doing.
G
Gary Cohn4:56
Look, I think those companies are going to continue to spend. I think they're committed. They want to be there. They want to be a dominant player. They want to have huge market share. The question is, and the big, big question to me is, are we building out more infrastructure than we could ever use? And so just because you want to be a dominant player, that doesn't mean that all dominant players make money. You know, we went back to the fiber optic days. All of the big three companies that laid all the fiber optic cable never got to monetize that cable. WorldCom, Enron, Global Crossing, they never got to monetize that infrastructure. The infrastructure was valuable. It was valuable to the next buyer that bought it at a deep discount. So I think we have to ask ourselves, in this period of time, everyone's trying to capture market share. Everyone's trying to drive value, everyone's trying to be there. But what is there going to be? And on top of that, you know, while we were all looking at those stocks going down yesterday, the White House signed an executive order on quantum computing. So the reality that quantum computing is around the corner in coming, we can't dismiss that. And that's going to need a whole different set of infrastructure. It will, yes, it will leverage upon the AI infrastructure, but the quantum infrastructure and quantum data centers are going to be different than the traditional data centers being built today.
B
Becky6:13
I'm trying to put the comparison of WorldCom with one of the Mag-7 from today.
G
Gary Cohn6:20
I didn't mean to throw. I didn't mean to throw you there.
B
Becky6:22
No, I mean, that's hard to do. I understand. And if you look at like what happened with Cisco and how the valuations never came back to those levels, I get that. But is that kind of your suggestion of what we might.
G
Gary Cohn6:34
Be seeing? My view on WorldCom and Enron and Global Crossing is these were really important companies at the time. They talked about fiber optic cable and connectivity and video over internet, voice over internet, how it's going to dramatically change the world we lived in. Guess what? They were completely right. What they were doing was going to dramatically change the world we lived in. But the way investors were valuing those companies is they're saying, okay, you can go out and borrow as much money as you want today. You can go lay as much fiber optic cable as you can. And we're going to take the forward value of your contracts and discount them back and put a multiple on them. Unfortunately, the forward value of those contracts wasn't what they had put into their models or what they had projected. So you're building a data center today. You had someone on yesterday talking about a data center with a 20 year takeoff agreement for power. What is the value of compute 20 years from now? Do we need as much compute as being built? I happen to believe we will have overbuilt compute by that point in time, and compute will become a commodity, and companies will buy compute from the lowest cost provider like every other commodity.
B
Becky7:39
All right. That makes a lot of sense. It's interesting. And it's definitely, you know, some days the market thinks that, some days it thinks the opposite. And that's where you see the swings in the ebb and flow of the market.
G
Gary Cohn7:50
But that makes sense to me that we're seeing the flows because we're all trying to figure out where this goes, how we