Kaplan on Raising Equity for Tech, Fed Decision

We begin this hour with tech leading stocks higher following a trio of major financing deals. One bank sitting at the heart of all three of them. Goldman Sachs co leading Meta's $98,000,000,000 equity raise, structuring a $40,000,000,000 SpaceX debt agreement, and taking a top banking role in Anthropic's mega IPO. The Goldman Sachs vice chair, Rob Kaplan, joining us now for more. Rob, how big is the bonus pool gonna be? Never mind that. Congratulations. Thank you. How busy have things been? Things are are extremely busy, and I think we're, and and I am I am and we are very optimistic about the promise of AI. Having said that, I think what you're also seeing even with these large transactions, you're seeing in the last few weeks, the market started to differentiate based on business models and on capital structure. I think that's a healthy development. What what I mean by that, I think you're gonna have to have enough equity and limit the amount of debt based on the business model so that the model works in a downside scenario. And I think you're starting to see the market police that, and I actually think that's a healthy development. Is it fair to say that access to capital has become an increasingly uneven? It's... The market is doing what it does in most industries. It's differentiating based on promise of the business model, sustainability, and then, again, it's looking at how much debt versus how much equity. And I think it's starting to charge more if you are, you know, less than investment grade, single b. And I think that's what the market normally does, and I think it took a while for that to happen here. But I think it's a welcome development to see this scrubbing and differentiation. It's still more challenging to raise capital. There's enormous amount that needs to be raised even for investment grade companies. The concession has widened a little bit, but I think there'll be good access. The market will just do what it typically does. It'll make choices, and I think that's a good thing. Just to underscore what you're saying, are the hyperscalers gonna have to sell more equity in order to raise more money in the debt markets going forward? I think they've decided to, monitor their capital structure and their debt to equity. And so, yeah, I think they've decided to make sure that they've raised enough equity so they don't get, too highly leveraged. And I think that's a responsible decision and I think a prudent one. We've been talking about what almost feels like a race to get in before the window closes for financing. And there have been a lot of questions about how long the pipeline is before that financing window closes. From your vantage point, how how narrow is that window versus how broad is the investor appetite? Listen. There's there's two trends going on. There's the infrastructure boom. Any... Anything that creates compute, that's where... That's the capital debate we're having is is about that. Then there's the adoption boom, model development. That's the most most recent... Some of their most recent announcements. Honestly, a little less capital, but very dependent on enough compute. And I actually... We think there's still gonna be very good access for both. But, again, the market will differentiate, and increasingly, it will segment different types of models. Are you using the compute for your own purposes to monetize your own customer revenue, experience so that you you can control it? Are you a merchant that's selling it to others? How dependent are you on compute pricing? We think there's gonna be a big ramp up continued in token demand and demand for compute, but the pricing for certain types of tokens will probably decline, and we think that's a healthy natural development. And I think then you look at models and reframe and say, how are... How... Which are they most dependent on? How speculative is that? And then how much how much equity? How much debt? What's the right capital structure? A big question is how much adoption there's gonna be across the broader economy and how companies are deploying it. And as Microsoft offers 60,000 buyout offers to people, there's a question of how much the efficiency is gonna have to come with layoffs or cutbacks of human workers. What's your experience at Goldman Sachs in this area? Well, here's what... And I and I can tell you talk with clients in a range of industries and including small business and entrepreneurs. AI, we're probably in earnest a year into in earnest AI, use case development. In other words, we... Most companies started in earnest trying out 15 to 20 use cases maybe a year ago. We've learned an enormous amount. Some of the cases will work, some won't. We're refining it. We're learning how to use tokens, how to segment them. And I think even though many companies say we haven't yet seen tangible benefits, I think, actually, you're gonna see more and more benefits realized over the next year as companies learn, and that's been our experience. And the other thing we're seeing is it's giving CEOs more confidence to say, we're gonna have 30 to 50% higher revenues. We're gonna redeploy some of our people, or hire new people to cover more clients, maybe, less of some other internal functions. And because of that, you know, five years from now, we're gonna have 50% more revenue with the same number of people. Maybe not less people, but a different mix of people. And, you and I were talking off camera. The ability to cover a client, build a relationship, have a face to face communication, those old fashioned skills and flexibility are are gonna be at a premium, and I think that's probably gonna be, in in many ways, an exciting development. But if you view your career very narrowly, I only do this or only have a certain function, I I think you're gonna be more challenged. You're gonna have to view yourself, more broadly, and, and, I think we're seeing that migration happen. How do you view the political pushback we've seen in terms of AI? Well... So there's there's two parts of that. There's number one, on the infrastructure side, there's the moratorium in in number of states, including the one I live in Texas. But but I think that's a little bit more trying to create a more level playing field on behind the meter power, sharing of tax revenues, water usage, etcetera, and I think trying to regulate these deals. K? So that's one kind. I think that will get resolved. And then the other the other, concern is, are we gonna destroy jobs? Are we gonna have new jobs? I think we're more optimistic on that. But, also, what are some of the more, dangerous elements of some of these models, and how much should there be self regulation? How much should there be a third party or government regulation? And I think those debates will continue, and and I think that's where you're seeing a different kind of concern. They're separate concerns, but sometimes they get conflated and mushed together, in a broad concern, but I think they're really different type of concerns. And the Federal Reserve's gonna monitor this as well. You've got experience in that committee. We were joking earlier. Could you describe day one of a two day meeting? What time do you wake up? What time do you all get together? What is it... How does it go? Get there you get there on... The meeting begins on Tuesday, but you get there on Monday. And, normally, all the presidents, we would normally meet Monday afternoon. Away from the board? Yeah. Away from the board. You would have probably discussion of maybe about bank supervision or some other topic, probably have dinner together Monday night. Meeting starts Tuesday morning. Day one, as you know, is a review. We all go through our views on economic conditions. And then day two, we all go through our views on on, what's the right monetary policy. The issue now, and and you've heard me talk about this, is part of the, this inflation broadening. How much of it is supply? How much of it is demand? How much of it is cyclical? And how much of it is structural? And I think the thing that's challenging here, we just talked about AI. If it's AI related, it's strong. If it's defense spending related, very strong. If it's housing, autos, serving low moderate income consumers, sluggish. And so there's a more mixed economy than the headline suggests. And the other problem is if you if you could undo the diesel and oil spike, just that one thing, I think, some of the bleed and the broadening of inflation is because of that structural change, not because of excess demand. Now there's no right answer on this, but that's the debate they're having, and that may be one of the reasons. If they skip October, that's the reason they would skip. It's not that they won't keep hiking, but they wanna make sure they understand these crosscurrents. Do think they've got the luxury of time of waiting? I I think they've got the luxury of waiting one meeting. I think, you you know, I've talked about there's there's strategic errors and there's tactical errors. Strategic error is you you delay in a way that it takes a year or two to to to fix it. That probably is true of what happened in '21 and '22. I think that was a more strategic error. If if you're meeting late, you'll live. And I think what they don't wanna do is make a mistake where they they hike too far, too fast, and we... Lo and behold, we look up next year because the... Because diesel's not resolved. You haven't made a dent in inflation, but you have slowed further a bunch of segments of the economy that were already sluggish, and and maybe you're struggling with both ends of the dual mandate. And that's the balancing of the debate that they're having. Can we finish on communication? I don't wanna put you in an uncomfortable position, but do you think that the troika is alive and well? I don't know. I can't I can't tell exactly. And, you know, honestly, I will just speak for myself. If I walked into the October meeting and it was clear to me that the chair would like, in this meeting, would like to raise rates, I would not resist. I would go along with it. I'd prefer to skip October for the reasons I just said. But if you know you're gonna raise in December, I think I think people would go along probably with October, but there wasn't the stage management to set it up. Do you think it's necessary? I think I think in an economy that's leveraged over a 100% debt to GDP as opposed to the Greenspan economy, which was in the fifties. Okay? When the economy is this highly leveraged, I think a little bit... You don't have to forecast what you're gonna do, but a little bit more stage management, clarity about your reaction function, and get a little bit of the premium and uncertainty out of the treasury market is valuable when we're this highly leveraged. It's a different economy than the one Greenspan, presided over. Rob, deeply thoughtful. Good to see you, sir. Appreciate it. Thank you very much. The Goldman Sachs vice chair, Rob Kaplan.