Moody's Talks - Inside Economics - The AI Series: A Question of Timing

Episode Date: August 25, 2026

Capital Group economist Darrell Spence, who has covered the U.S. economy for more than three decades, separates AI hype from the evidence, explaining how much growth the current capital-spending boom is actually buying, where productivity gains are most likely to show up, and who benefits when they do. He also looks to earlier investment cycles for a sense of how long this one will last. The spending is real, but Spence makes the case that enthusiasm and economic resilience are not the same thing — a distinction that matters more as consumer spending softens, inflation persists, and tariffs push up costs. Check out the link to Capital Group's 2026 outlook: 2026 Midyear Investment Outlook View our latest articles and research on AI- Is AI a Miracle or a Mirage? Questions or Comments, please email us at [e-mail]. We would love to hear from you. To stay informed and follow the insights of Moody's Analytics economists, visit Economic View. Hosted by Simplecast, an AdsWizz company. See pcm.adswizz.com for information about our collection and use of personal data for advertising.

Welcome to Inside Economics. I'm Mark Sandy, the chief economist of Moody's Analytics, and I'm joined by my two trusty co-hosts, Marissa Dina Talley, Chris Trudis. Hi, guys. Hey, Mark. Hey, Mark. Hi, Chris. It seems like we've been doing a fair number of podcasts recently. Never enough. Never enough. Three this week, right? Yeah, yeah.

And this is the second podcast in our AI series. And we have a, as everyone knows, our first guest was David Autor of MIT. The second guest who's with us right now is Daryl Spence. Daryl, good to see you. Good to see you. Thanks for having me. Absolutely. Daryl is an economist at Capital Group and a great asset management firm.

And just to the sake of disclosure, did you know my, Darryl, did you know my daughter it works at the Capital Group? Absolutely, yeah. What, does everyone know her? Can I tell her that everyone knows her a Capital Group? We're glad to have her. I can tell you that much. Oh, boy, I'm going to let her know that.

Yeah. I think this is, I know there's different ways of describing the portfolio manager, but I'm going to call her a portfolio manager out of London, and she's been with Cap Group now for a number of years. Great, great institution. And we had Jared Franz on from Cap, Capital Group not long ago. Do you work with Jared? I work very closely with Jared.

Yeah. We both cover the U.S., and he is the true AI guru when it comes to knowledge about the technology. But I've been working with him for, I think, over a decade now. Oh, is that right? Okay. Well, it's great to have you aboard. And how long have you been at Capitol Group? I just passed my 34-year anniversary.

Well, congratulations. I've got you beat. I'm 36 years. Wow. It's a rarity. Yeah. Every day I count my lucky stars that it worked out the way that it has. Has you been a CAP group for, you said 34 years, my goodness.

34 years, yeah. Same kind of job as you have now, economist, or? Yeah, I have the shortest bio, I think, of anybody here probably as I went to undergrad. I did major in economics, and then I came here. I knew nothing about certainly capital group. I knew a little bit about investment management and then just spent the next 34 years trying to learn the role and how to apply. economics to market analysis and to investing. And I think in a way it was fortunate that I came in as a blank page because I was able to learn what I needed to learn while actually doing it.

But this is a tremendous place. I'm not going to turn this into a commercial for capital group, but just the fact that I've been here for 34 years is probably a testament too. It allows you to really explore what interests you. It allows you to exercise your creativity in the role, even though perhaps that's not often a word that people, associate with economics or investment analysis, but I do think it's important when you're trying to find different ways to look at the world. It just allows a lot of autonomy. So I have found that it's just been a tremendous place to grow and work with just amazing, amazing people over the past nearly three and a half decades now. Yeah, it's pretty amazing. Are you in Southern California?

Yeah, I'm in Los Angeles office, which is the main headquarters for Capital Group. Yep, it is. I keep saying Cap Group. You don't say, cap group. You say capital group, right? Well, I mean, when we're shortening it, yeah, we'll say cap group when we're referring to our own company, but yeah, out in the broader world, it's supposed to be capital group. Yeah, great, great organization. So let me ask, I was reading some of the work you've done, it's up on the web. Do you do explicit forecasts? I mean, like in our world, We do, you know, a gazillion real GDP, inflation, interest rates, demographics, lots and lots of things.

Do you do those kinds of forecasts as well, put pen to paper? Yeah, I mean, probably not as explicitly as other people do. You know, are we putting together the four-quarter GDP forecast by component table once a month or once a quarter? Not really. And I think part of it has to do with the nature of our investment process. We are still a very bottom-up firm, and obviously economics is a top-down discipline. So I think if you actually saw the day-to-day here of how we try to take economic analysis and make it applicable to our investors, it would look a lot different than what you might see elsewhere. You know, as a bit of a background, we sit with our analyst and portfolio managers.

We're on the same floor with them. We intend all of their investment calls. We even travel with them on off-sites and sometimes to company visits. And the idea is to really be in the flow of their conversation, to listen to what they're talking about, the questions that they're explicitly or implicitly asking. And so we can kind of respond to the moment. We're not doing macro over here and lobbying it over the wall

and hoping that they can kind of pick it up and know what to do with it. It's very much a collaborative process. And what that's also led to is the nature of the way we do economics is very tailored to our investment process. So it's less about what's GDP going to be this next quarter and the quarter after that and more about what is the market misinterpreting? What is it not appreciating? Where do we think it has it wrong?

What's it going to mean for different industries and companies? And one of the areas that I've taken analysis over the past, well, 34 years, but this is probably newer over the past 15 years or so, is to really try to make concrete linkages between macro developments and company and industry fundamentals and define correlations that are interesting between global commodity prices and industrial production and caterpillar sales, things like that,

which is very unique, I think, to our process. But it's also a lot of fun to try to find these things and these hidden relationships that maybe the rest of the market hasn't totally gotten or found out themselves, and then tell our analyst and our portfolios and managers about it. So, you know, what we're trying to do here is if you think about what an investment firm is trying to do when the analyst and the portfolio managers are trying to do, they're trying to put together the pieces of a puzzle that will tell them what the future is going to look like,

and you're never going to have all of those pieces. And what we're trying to do as top-down investors is to add a few more pieces to that puzzle, and a lot of those pieces are coming from the bottom-up analysis that our analysts are doing. So again, it's very tailored to our specific investment process, which lends itself a little less to doing a lot of pinpoint forecasting, a little bit more about what is priced into the markets, do we have a vague sense of what their forecast is or the market's forecast is as it pertains to inflation or growth or the Fed or whatever, and where do we fall on what side of that kind of vague forecast do we fall?

And if we are right at the end of the day, how will the markets eventually adjust if they come around to our view? So I apologize. That was a bit of a long-winded thing, but I think it is different here. Well, I can tell why you've been a cap group for 34 years. Yeah, I can feel the enthusiasm. But it sounds like a fantastic job, a great job. I tell my daughter that every day she's got the best job on the planet.

You know, long equity. Who does long equity? I mean, just an amazing place. So I guess AI, our official intelligence, is, kind of right in the nexus between the top down and the bottom up, right? I mean, there are very few kind of forces, events, dynamics played out that have such important implications both for the macro economy, you know, what's going to happen in aggregate around the world, but also what it means for investing, both in the equity market and the debt market.

I mean, the nexus there is pretty clear. Yeah, and there's just, there's a lot of. reasons to be excited and there's a lot of reasons to be concerned. And there's just a lot of different places you can take the conversation whenever you talk about AI. I mean, the first one from a macroeconomic perspective is just the impact that it's having on growth. Everyone kind of has their own definition of what's AI-related spending within the economy. And I have mine. It's essentially computers and peripheral equipment software and data centers as part of the CAP-X story. And if you look at the past four-quarter, you know, we had 2.1% GDP growth, but if you get rid of that category, which is depending upon,

again, how you define it, 5 or 10% of GDP, that drops down to 1.4%. And I'm sure you've seen the statistics that show that, you know, in a quarter or two, AI-related CAP-X has actually added more to growth than consumer spending, even though consumer spending is 70% of the economy, and AI-related CAP-X is less than 10. So all of those are very complicated, wonky economists, ways of saying it's growing really, really fast. So obviously, the economy is kind of becoming a little bit dependent on this spending. So if that went away, obviously growth would slow and there would be implications. And then there's the whole financial side of it. I don't know if can I just ask there on that one, on the investment side? Because there's a number of different

channels through which AI influences the broader economy. So far, you're just focused on the build out, the infrastructure, the data centers, and all the other. investment. But, and you said 2.1% GDP and you said 1.4%. That's the buildout. But that overstates the case, doesn't it? I mean, because of all the imported electronics that are coming into the country to fill those data centers, that's an offset to that 1.4. No? Yeah, absolutely. And then there's yet another offset on the other side is if we're just measuring software, computers, and data centers, we're not measuring the power build out, a lot of the other stuff that goes along with that, too. And I haven't dug so deep into the data to know which one of those is the larger one.

I suspect the drag from imports is a little bit bigger than the affiliated stuff that we're not capturing in the actual AI investment numbers. But yeah, it definitely is a little bit less than that, but it's still, no matter how you slice and dice the data, this is a very large contribution from what is not a huge sector of the economy, right? now it's out punching its weight if I could put it in more simple terms and I think you know when you look underneath the GDP numbers take out look at investment of everything that's not those categories of AI and it basically contracted for five quarters and it finally turned negative or excuse me finally turned positive just barely in the second quarter of this year so you can make the argument that it's I don't know if it's crowding out other investment if it's just

that the actual economic environment outside of AI isn't robust enough to incentivize companies to invest. But again, investment, business investment outside of those three AI spaces has not gone anywhere for well over-y. Oh, I see. So you're not talking GDP. You're talking overall investment. You're saying, hey, if I, if you take out the AI, AI plus non-AI is leading to all this growth. But if I took out AI, the non-AI investment spending would actually be negative, is what you're saying. It was negative on a quarter to quarter basis for, I believe it was five quarters in a row until just the second quarter and it turned slightly positive. And then the other numbers I mentioned, the 1.4 is GDP without that AI cap X. Not considering all these caveats, the imports.

Not considering all the cabots. Exactly. Exactly. You were going to go, the next thing you were going to go to another channel, I think, in terms of, the financial market effects, were you going to go there next? Yeah, I mean, a lot of this is, like I said, there's so many aspects of it is there's the financing of it. Will the demand actually materialize? What happens to the labor market? And as I think about this kind of broadly and watch the way this is playing out in markets,

you know, I kind of have this feeling that there's a tension between two different forces here. You know, for AI to justify the revenue expectations, the capital commitments, all of the earnings forecast, etc., it really has to be a truly transformative technology. And it very well may be, again, I am not the expert in AI to answer that question. But by truly transformative, I mean that it has to be adopted largely by enterprises broadly. and they have to be willing to spend a lot of money on the technology, and presumably the reason they would do that is to save cost on labor. So if that happens, and it happens on the time frame that the AI enthusiast

suggests that it's going to, or maybe even has to, to justify these very, very optimistic forecast about what the next few years are going to look like, clearly that's going to lead to labor market disruption, and that's going to require some type of efficient and effective redistribution system that, at least in the U.S., we really aren't set up to do. Now, the alternative is maybe it's not as transformative or not as transformative on the time frame that the AI enthusiast and maybe the markets expect it to be. Obviously, that's a better story for the labor market, but it would clearly call into question some of these optimistic forecast and revenue expectations and capital commitments. And obviously that would play out in the market. So what's the lesser of two evils there?

You know, when the Internet bubble burst, I think that proved that you can have a sector-specific bubble unwind without a lot of economic damage. The recession, as you know, that occurred back in 2001, too, was actually a very, very mild recession, despite the fact that the equity market writ large was down substantially. If you remember the S&P 500, I believe it was down over 50 percent in the NASDAQ. close to 80% if not a little bit more. You know, if that played out this time around, would it be the same? Hard to know because it seems like all of the CAP-X activity that we're talking about is real economic activity.

So the notion that if the AI cycle unwinds, that it would largely be a market phenomenon, I think is probably a little bit optimistic. I think it would translate into economic weakness a lot more than the Internet bubble did when it unwound in the late 1990s and early 2000s. And the one thing I mentioned earlier, the Wall Street Journal article, there was also a Financial Times article yesterday about a lot of the commitments that companies have made that aren't on the balance sheets now. You know, leases that they've agreed to in 2029, 2030, which accounting regulations don't require them to recognize until they start. The Wall Street Journal had it at a trillion and a half dollars. So it's not just the money and the CAPEX that we're seeing put into the economy right now. it's what companies have committed to starting three, four, five years from now.

And again, this could all work out great. It could be transformative in the way that the market expects, but that's kind of a long-winded way of saying there's a lot of expectations about a lot of money being priced into all of the AI-related stuff right now and a lot of real economic activity that's going toward anticipating this being, you know, pervasive throughout the economy and pervasive throughout business and pervasive throughout life, it very well may be. A lot of it also has to do with time frame. Will it do it on the time frame that the market and everyone kind of expects it to or needs it

to to justify current valuations? You know, the same is true about the internet. I mean, it's not like the internet went away or didn't become a thing just because the bubble burst. I mean, when the power goes out, I'm not mad because I'm sitting in the dark. I'm mad because I can't go on the internet, you know? So it's still here. It's just that the expectations and the time frame that markets and individuals thought that it would occur on didn't actually come to be met. And it just took a while for that to play out. Okay, okay. Let's slow down. There's like a gazillion things to unpack there. So we're going to have to roll it back and unpack them one at a time. But let me just first say,

do you actually talk to Jared? You guys like seem to be on different planets when it comes to, you know, he's much more optimistic, if my memory is correct, you seem pretty lugubrious, but we're going to go back and unpack it, but Jared seems a lot more optimistic. Do I have that right? You guys have different perspectives on this. Jared is much more optimistic, and this is when the conversation expands even further. He is more optimistic. Don't do that. Don't go anywhere else because I'm taking you right back. I just want to know if you guys are on the same page. And it's okay, because Mercer argues with me all the time. Chris never agrees with me.

I'm just trying to find, I'm just curious if you guys have a disagreement on this. Yeah. Jared is a, and Jared, he is, as I said, he is the AI guru in terms of the technical aspects of all the technology and learning a lot about how it works and the different models, et cetera. I listen to him a lot. He's a very big productivity bowl when it comes to, AI. I don't disagree that it will probably be productivity enhancing. I think where we probably

disagree with is on the timeframe. Okay. God. You know, people have looked at productivity statistics recently over the past two quarters and said, hey, it's been growing, you know, two and a half, three percent. That's proof that AI is playing through. You know, keep in mind, there's a huge pro-cyclical component to productivity growth. It's the short-term buffer to accelerations and decelerations in economic activity. So you get a modest pickup in activity. Firms don't just instantly go out and hire a bunch of people. They just boost productivity a bit. And the same thing when things slow down, they don't can a bunch of people. They just reduce productivity a little bit. And so if you look at kind of the correlation between accelerations and deceleration to GDP, there's a high

correlation with productivity growth. And we had a bit of a minor, uneven but minor acceleration in GDP and productivity picked up. And in the most recent quarter, it slowed. down and guess what? Productivity on a year-over-year basis went from high twos to two and a half or something. I think a lot of its time frame. I think maybe there's a little bit too much optimism being given, or AI is getting too much credit for recent productivity growth increases than perhaps I think it should be getting. But that place is the consensus, Daryl. I think I think the consensus is that, you know, it really has not added. Okay.

Some business as usual may be kind of productivity games, but no juice. I agree. And the survey seemed to support that. It's like a lot of companies are giving it to their workers, and it's kind of improving individual worker productivity in different ways, depending on the individual. But to really boost productivity over long periods of time, it has to change business processes.

There almost has to be someone in charge of figuring out how to implement it across an organization and across workflows. Yeah, I'm just saying you sound like I'm selling AI stocks. He sounds like I'm buying AI stocks. I'm just saying. I'm just saying. I know timing is everything. But let's roll the conversation back a little bit.

Okay. Because where we started was current growth, and you went right to all the investment spending that's going on, the infrastructure buildout. The one thing you didn't mention, and I'm just going to fill in the blank, what about the wealth effects on consumer spending related to the AI stocks? Do you feel like that's a major contributor to growth as well, or not so much? I don't. Really? Oh, interesting.

Okay. I am not a huge believer in wealth effects, and I have a very simple explanation for why that is. If you go back and you look at, you know, 70 years of income growth and 70 years of spending growth, they are very, very closely correlated. The big exception is obviously during COVID where income went way up because of all of the government transfer payments, but nobody could spend anything because they were locked down. And then the reverse happened when we came out of COVID and the saving rate came down and all of that. But I think it matters at the margin. I think home prices might matter at the margin. I think things like the alignment of the planets might matter at the margin.

But at the end of the day, it comes down to income growth. And just the correlations there are really, really tight. And if I thought there was, you know, strong evidence of a wealth effect, you would see divergences between income and spending. And every now and then you do, but it's a tenth of a percent or two, tens of a percent, it's not really meaningful. So I'm not saying it doesn't matter at all, but if you gave me two things to pay attention to, and one was the equity market, and one was the job market, I mean, at the end of the day, it would always be the job market. Even if you told me equity, you're going to go. Yeah, yeah, yeah, yeah. I don't think we did.

I know it's kind of stating the obvious, but there's a possibility that maybe I am underestimating the wealth effect, but I'm not totally seeing it in the data yet. maybe maybe this drop in the saving rate is indicative of people spending that wealth because obviously that doesn't show up in the income. But I think there's a more logical explanation for that and that is inflation and energy prices and people trying to keep their consumption basket unchanged in the midst of that. But, you know, that could that could be the wrong interpretation. I admit that. But I generally don't think it's a huge factor in the economy. Got it. I totally disagree, but that's okay. But the,

irony in all that is, I think we still land in the same place in terms of the actual contribution of AI to GDP because I think you're way overstating the effects of investment because of the trade balance effects, the imports. So we probably land in the same place. It's a big deal for growth, but we're getting there in a very different way. But let's not dwell on that. But I will, I hate to lose argument. So I'm going to send you a paper I just wrote so that you can take a look and you can you can poke holes on it. But here, let's go on to the outlook and you kind of, I'm going to frame this just to restate it so the listener can get their minds around it. The way you framed it was, you know, there's kind of two,

there's, I guess there's three scenarios. One is the baseline scenario, you know, what you think actually is going to happen with AI and how it's going to play out. You didn't actually talk about that, and I want to come back and hear what you, what you think this is actually going to play out. But then you talked about the other two scenarios, one where, uh, AI comes on quickly, significantly raises productivity and leads to a job, significant job displacement. And you talked about needs for policy around redistribution, that kind of thing. And the other scenario on the other side of that, almost the polar opposite, is that AI doesn't generate those big productivity gains. It's kind of a bust. And the result is

the markets, the equity in the bond market, are over, uh, overstating the contribution, and there's going to be some type of sell-off or correction or sell-off in those markets, and that has some implications. Do I have that roughly right? Is that framing roughly right? Yeah, I mean, the two extremes are not going to be either one or the other. Yeah, right. Reality is that you fall somewhere in between. Of course this technology is going to be around. Of course it's going to be utilized.

The question, really, in my mind, is on timeframes. And our expectations about how quickly it can be monetized, adopted, et cetera, realistic, you know, because again, the internet is the same example. The internet was a thing. It is still a huge part of our lives today, but things got a little bit ahead of themselves back in the late 1990. I think the big difference from an investment standpoint this time around is the profitability of a lot of these companies just is just through the roof, as you know. And so, you know, you had some kind of funky companies back in the internet bubble that probably didn't have a really good chance of survival at all. They had weird businesses, but they had a W-D-

or a dot com in their business strategy and the market just took it up. It's a different world right now. I mean, these are truly, as you know, cash flow generative, very, very profitable companies. The question is whether or not the ROI on a lot of this CAPEX spending comes back to them. Again, timeframes a very important part of my thinking about this, not because I'm not thinking about it from an economic growth perspective or a productivity enhancing perspective, but I also have to think about it from an investment perspective. And around here, Ghealer, and go ahead. No, I was just going to say, so for the next year two or three, how do you think this plays out? I mean, what's the most likely scenario?

I worry more about the slower adoption. And ultimately, that comes back into pricing. We had CSR Capital Strategy Research is the unit that I work with in Capital Group, and we do macro research and currency research and political and accounting research. And we had an off-site recently, which was dedicated to, guess what? AI, surprise, surprise. But what we learned about adoption, and again, this technology is so new. There's not like really good data out there, but that adoption is probably going a little bit slower than perhaps people thought it was. And firms are starting to bring it in.

They're realizing it's not just as simple as giving the tool to somebody and all of a sudden productivity just blossoms. There needs to be a more proactive effort around it. And it's also expensive. I mean, it can get pretty expensive, too, if you're really trying to incorporate it and if you want to protect your data and do things like that. So I worry it's going to be a bit of a slower rollout than perhaps the market expects. You know, and then when I think about it from an investment perspective, too, I mean, you know the concentration numbers in the S&P 500. You know, that the top 10 companies are now 40% of the index. You know, is that a prudent portfolio for most investors?

I just don't think it is. I don't know if there would be too many financial advisors who would sit down and say, yeah, I really think you should have all of your eggs in this particular basket, not only in these handful of companies, but companies that are all exposed to the same thing. And so I do worry a little bit that when people look at the S&P 500 or passive investing, and full disclaimer, I work for an active investment manager, and I believe in active management.

But people are believing that perhaps investing in an index because of the diversive that supposedly it provides is actually safer when, in fact, you are very, very, very exposed to this theme right now. And it very well may play out, and I'm almost certain it will play out over a longer time frame, but will it play out enough, again, for these financial commitments to be realized and get a return on investment on them? And I started to stray there for a bit, so I kind of forgot where your question was going. Oh, no, no. So I'm going to bring you back because I'm going to try to pin you down. I know you investment managers, you don't like to be pinned down, but I'm going to, I'm still going to try to do it. So you, I asked, I kind of framed it as here, you've got,

you've got two scenarios on other side of the distribution of possible outcomes. And that's oversimplifying, obviously, because of timing issues, but broadly speaking. And then I kind of framed it as there's a baseline somewhere in between. Because are you saying your baseline is that there will be some kind of correction sell off in financial markets as a result of your views on those slower adoption of enterprises, is that your fundamental, is that your most likely scenario? I am probably, let's put it this way, remember when I said, we don't pinpoint forecast. Yeah. You're right.

I don't want to get pinned down to something. I put a higher probability on that perhaps than the market does right now. Now, some of that is a function. Well, the market doesn't put any probability on it, Daryl. That's not saying anything. Well, there are. Well, there are. That's true.

But, I mean, I guess I worry about it. a lot more than maybe writ large the investment universe does right now. And part of that's based on, you know, this is, you're going to laugh at this. But, you know, one of the few benefits of getting older is experience, I suppose. And I had a younger associate who was doing some work on market concentration and they just wanted my input for something as if they thought I knew something. And they brought a chart in. And it was, I can't remember if it was like, you know, the top 10 largest stocks is the share of the index or it might have been the biggest sector is. the share. So you go back far enough. You had like energy got to be 30 percent and then, you know,

internet got to be 25 and then financials during the housing bubble. I forgot exactly what the nature of the chart was. But, you know, it showed these concentrations go up and come down. I kind of realized if you took all of the titles off of that chart and showed it to somebody and said, this is a real data series. This is something that actually happened. Where do you think the line goes now? Any human would be like, down. Of course. It goes down. Why do you say it? Because it always goes down. You need put the, oh, wait, that's AI. Okay, no, I changed my answer.

It's like, you know, if you can get out of your own head about, because at the, at the peaks of these things, there's always a good explanation for why it makes sense and why it's going to go on forever. And it could. I'm not saying that it couldn't, but there's also that, you know, history too. And I just remember looking at that going, any sane person would say that line's going down. But as soon as you tell them it's AI, oh, no, it's fine. It's totally fine.

So is that analytical rigor? Absolutely not. But maybe that's a little bit of experience talking to. And so, again, I think the technology can continue to be very impactful whether or not we don't have in a market adjustment. If I had to say yes or no, I would say yes. I would be a little bit more concerned about that. Okay. So the kind of the way I would place my views on this, just and I want to get your take, is that there is that there is.

is a middle ground. And the middle ground is that AI is it's not going to be dystopic. It's not going to come on so fast. There's just too many frictions. Legacy companies can't adopt. Yes, capabilities are improving, but there's limits on that, too, given concerns around cyber, national security, so forth and so on. You know, there's a lot of impediments for this thing to become dystopic to the labor. Yeah, there's going to be a lot of turn over here, but there's going to be a lot of growth over there. The net of all that, it all kind of works out. And also on the equity market and the bond market, yes, you're right. I mean, if I look at any measure of valuation, you know, it feels like it's got to come back in. But there's different ways that can happen. You know, one way is kind of an ugly way where we had in the Internet bubble and everything blew. We lost 50% of the value of the equity market, you know, very quickly. Or the market kind of goes sideways for a long while and lets the reality of these productivity gains kind of kick in and catch. up and if that's this case, that's our baseline. That's our baseline view. And that's the consensus,

right? I mean, most, if you look at CBO, if you look at, you know, anyone who does this for a living, you know, actually does the numbers. It's kind of down the middle. That's kind of the baseline scenario. Yeah. You don't think that's of those, of, now going back to your, your way you've talked about probabilities, you don't think that scenario has a higher probability in the scenario where you see a major correction in financial markets? Sounds like no. That's okay. Yeah. Sounds like you think it's, yeah. I don't think markets, that tends to be the way markets reach new equilibrium.

I tend to think they move a little bit more sharply. And some of that is informed by years of doing acid allocation around here, separate discussion. I've been part of an acid allocation group for a long period of time. And one of the things I learned was that valuation is a very, very poor predictor to equity market returns. You can sit around and talk about the market being expensive for a long period of time. But in and of itself, in the near term, it doesn't really tell you much about market movements. Rather, what tends to have to happen is you get an economic deterioration of some sort,

and then high valuations just make the ultimate decline worse. And so the world that you're describing to me in my simple framework and taking that kind of experience on board is one where, you know, the economy kind of chugs along. Everything's all right. The earnings eventually catch up to the valuations and everything's fairly stable. I don't know if the economy can avoid some type of adjustment over the time frame that it would require for that to happen. to not have a more significant adjustment. Does that make sense, kind of? Totally, totally makes sense.

And I'm uncomfortable with it too. So if you look at my distributional possible outcomes in this frame we've constructed, it's a very flat distribution, right? I mean, because, you know, it's hard to, given the uncertainties here, it's hard to feel very confident about any scenario. So totally, I totally get what you're saying.

It's just interesting to me that you're putting more weight on the near-term downside scenario of a correction in financial markets, then the consensus, for sure, and even for me, which I find, you know, fascinating. Yeah, but I've also learned, too, as living through a lot of these things, is if these things can go on for a long period of time, usually to be a catalyst. And I think actually one of the things we've been talking about, I should say, I've been talking about more recently, is the notion of what if the Fed does something that kind of nobody who's been around for a while thinks that they might do.

Because one of the things among many, but one of the things that cracked the internet bubble was eventually the Fed took the Fed funds rate to six and a half percent. And inflation was too, you know? And everyone was like, why it's, it's, and one, so here's a little cocktail party statistic for you. You probably know this. And if you quoted it at cocktail parties, you will guarantee yourself. Who goes to cocktail parties? What are you talking about cocktail parties?

I don't anymore. after telling people this little statistic. But if you take the daily Fed funds rate for the past 20 years, yeah, daily Fed funds rate, a little trivia here, and you look at it every day for the past 20 years, and this is still true because I checked it before I came up here. What percentage of the time do you think it was set at zero? Set at zero over the past 20 years?

20 years. I'd say quickly 15% of the time. 15% of the time. 1-5? 15. $45% of the time. Really? It's been at zero?

45% of the daily, go do the daily Fed Fund rate. The daily. Yeah, and just it was 45% of the time. Chris, what would you have said? What would you have said, Chris? I was going to say 25, but. Marissa, what would you were saying? I was thinking 25, yeah.

Okay. Oh, interesting. You can back check me on it, but I've looked at it so many times over the past two years. Really? reason I bring that up is because if you, and I use the word only in quotes here, but if you've only been doing this for 10 or 15 years, you get the impression, you could have the impression that the interest rate environment and the monetary policy environment that we've been in is normal. But a longer look at history suggests that it's actually not. And so I'm not saying the Fed funds rate is going to six and a half percent, but I do think we have a bit of an inflation issue. I I think the war is likely to drag on for a long period of time, and it may not escalate, but it probably won't de-escalate in a way that it creates disinflationary pressures. The AI CAP-X, I think, is in the moment, inflationary because the productivity benefits won't come for a while.

Right. I don't think inflation is going to spiral out of control, but I do think it's going to be sticky above the Fed's target. And so I'm trying to get people to maybe imagine a world where the Fed funds rate does go to four or to Will that crack this? If there's any cracking to be done, I don't know. But I think it's underappreciated, I think, in my interpretation of history and having to live through it, what role, you know, a Fed tightening cycle played in ultimately, along with other things again, but ultimately bringing the Internet bubble to an end.

Well, I suppose you don't even need the Fed to do anything if the long-term interest rates keep going up, right? I mean, that may even be due more damage, right? I mean, we're at four and three-quarters on a 10-year. What's the 30-year? It's like five-thirty. 30, 535 and headed north. Yeah. And that's, you know, again, longer two, if you have two and a half percent in this call,

just for sake of argument, two and a half percent inflation, to throw two percent on top of that for a 10-year yield historically isn't very high. Then you add on fiscal situation, wars, and all the other stuff that we're kind of grappling with right now. And, you know, maybe four and a half to four and three quarters on the 10 years, the low end. Yeah. Just throwing things out here to try to get people to think differently and not anchor on the past decade because I really think that was more the anomaly than, you know, where we are now. So just I want to connect the dots back to the economy.

It sounds like given your poo-pooing of the wealth effect that even if the stock market goes down, we can kind of sort of get away without a recession. We almost did in 2001 in the wake of the Internet bubble bursting. We had not, the only reason why we probably had a recession back then was 9-11 came along at about the same time. But without that, we might have navigated through. Are you saying we can still navigate through without an outright recession? If it's just, if it's just an equity market decline, my belief would be yes. I mean, obviously, if it feeds through into job markets, and I'm stating the obvious here, then you've got a much bigger problem. But just an equity market decline, yes, I do think.

Like the internet, like the internet period. Yeah, okay, got it. Yeah, because that wasn't a consumer. a recession. If I remember correctly, spending never even went negative. It was a CAPX recession, but the CAPX never reached the frenzy that CAPX is right now in terms of its impact on growth. Okay, I'm going to stop there. We cover a lot of ground, and I'll turn it back to Chris and Marissa. Chris, where do you want to push back on Darrell? What is the one thing he said that you disagree with the most? The wealth effect. I do believe in wealth effects, but my question is really about

credit markets versus equity markets. In your thesis here, if things actually were to deteriorate here, are you more worried about a correction in credit or equity? There has been certainly an increase in borrowing for the AI buildout. Is that a factor? Does that differ from what you see from the internet experience? Yeah, I think I would worry more about in terms of losses, probably the equity market, but maybe that's kind of stating the obvious because that's always where the biggest ones are going to be. And I'm not a fixed income investors, so I listen and incorporate a lot of what I hear our investors saying. I think the concentration in the fixed income markets makes me feel a little bit better about credit markets, that this wouldn't turn into some systemic event like the housing collapse did.

So there would be certainly pockets, and you're starting to see it, I guess, in some of these, you know, default swap spreads and things like that. But I think outside of this exposed sector, the credit fundamentals remain pretty good in the U.S. And profitability, obviously a lot of it's concentrated in AI-related things right now, but profitability writ large is still pretty good, even outside of the AI sector. Debt to GDP or whatever, you know, leverage ratio you want to use in the corporate sector, they're up a little bit, but they're nowhere near where they were, again, during the GFC. And so I think credit markets look pretty well supported, but that doesn't mean that there wouldn't be pockets if we get a very sector-specific, obviously, collapse in activity. But I feel like I'm saying stuff that everyone already knows.

You know, there's a – I don't know if you want to go here, but there is a flip side to the corporate sector remaining fairly healthy. And that is – and I'm sure you've seen the data that the labor share of the economy is hitting rock bottom here. Record low. Yeah. Yeah. And it's kind of, it's very interesting because the data, as you know, it goes back to what, the late 40s when the NEPA account art. And it was, it had its cyclicalities. It went up and down and up and down as the economy went up and down. And then after the GSC, it just collapsed and it has not come back. And I think you're starting to see that reflected in some of the political movements that are gaining steam right now. But I think it's actually, it could be an issue for AI as well. One of the things that we, or one of the data sets that we saw when we had our offsite was that it was like, it wasn't a huge survey, but it was a survey, I believe, of 19 developed countries and about how people felt about AI. And the U.S. was either last or tied for last in terms of our populace's enthusiasm for

artificial intelligence. And we're the country that's leading the charge here in terms of the technology and the investment, et cetera. So it's interesting to see it being kind of unpopular in the place where it's being born. But I think that's a lot of it. think that has ramifications for timeframe mark that you talked about how fast can this be implemented you know there's the adoption curve there's a profitability firms have to be able to afford it etc but there could just be a growing backlash for lack of a better term to AI and that would be partially related to this this income share so what i say is good for for fixed income markets etc but profitability financial indicators still being pretty good there's a there's a flip side of that and that's a labor

share, which again, I think I just opened a whole other can of worms or topic of discussion. You got me. You tested my acumen. Let me test yours. What is the labor share? As a share of non, well, of non-form corporate business output? The actual number? Yeah, just what do you think it is?

I did a post over the weekend. This is why I know. I know the chart and I know it goes like this. Yeah. Is it 65 or 60? That's what it used to be back, Okay. Okay.

It was, it was two-thirds. It's now... Okay. Marissa, what do you think? Is it 40%? No, no, not that low. It's not less than the time. Well, sure.

Pretty close. 52.9% at record low in the second quarter. Yeah. Now, there's measurement issues, all kinds of measurement problems. Probably didn't decline, hasn't declined that much, as much as that would suggest. But it directionally is, you know, very clear. So you, but you make a really good point. Chris, anything else before he asked,

Marissa, the same question. I'll seat to Marissa. Okay. Marissa, the baton is over to you. Anything you want to push back on, see, I'm trying to create a food fight here, as you can tell. He's trying to sow. Descontent.

Discontent. Yeah. No, I'm actually, I'm glad you brought up the labor thing as a labor economist. That's kind of what I'm keenly looking at these days. Do you think talking about the labor share, talking about the potential disruption in the labor market, do you think that there's any policy or any government intervention that we should be thinking about if we are, if there is this possibility that we see a large disruption in the labor market, you know, should we be thinking about retraining or income programs or anything like that? Yeah, well, I guess it depends on your philosophies about what the right government responses are. So I kept it very broad. I was very deliberate when I said effective and efficient redistribution mechanism because

if it happens very quickly, you know, that's that's the answer in a very, very short period of time. That's a Band-Aid, really, when you think about it. But it is a policy. The ones that might change the incentive structure is you could tax capital, obviously, make capital more expensive, rather relative to labor. I don't know if that's a compute tax or an AI tax. or something like that, that would slow probably the disruption,

but maybe not completely change it. Given the concentration that we see within the AI space and the profitability within a handful of companies, you could do antitrust regulations and try to break them up. Again, does that change the ultimate dynamic of this technology impacting the economy and ultimately the labor market? No, but it probably slows it. So, you know, some people would look at that and say,

that's the wrong policy for any particular thing. We should let the market figure this out. And that's why time matters a lot, too. Again, we've had technologies before, and they've been disruptive and to the labor market, but generally over a time frame that new jobs are created, et cetera, and beyond the cyclical effects that you see, it's really hard to kind of determine like a major labor market disruption. But given what the markets, the message they're seeming to be sending is like, this is going to happen tomorrow, you know, whether it does or not. But if it did, I think, you'd be dealing with a completely different technological transformation than the ones that we look at historically as example. So it's kind of a long-winded way of answering your question without actually answering your question.

Maybe you think, you know, that's what an economist does here. But it's the redistribution thing is the quick band-aid, the quick fix because the disruption is so sharp and so significant. If you want to kind of change incentive structures or other things you can do on the regulatory or the tax front. But again, I don't know if those are just slowing things down or if they would ultimately, you know, stop if that is the scenario that we're headed towards. Hey, Daryl, we're running out of time, but I want to end with going back to investments.

So how can you not invest in AI stocks? I mean, they're driving the train. There's such a large part of the indices. I mean, if I do what, you know, I know you're active investors, but a lot of people put their money in index funds, and they're just riding the index, and that's just being driven by Nvidia stock and SpaceX,

and soon-to-be-anthropic and open AI. I mean, what do you do? I mean, it's not obvious to me, you know, how you navigate that. Yeah, the answer isn't that you don't invest in them, but you may size it differently. And you invest in the ones that you believe in. Again, this isn't, sorry, I don't mean to throw an internet company under the bus, but it's not Pets.com or some of the other.

Yeah, yeah, yeah. These are legitimate companies, building legitimate things, et cetera. So it's not that you don't invest in them. You just are a little more conscious about the price that you're willing to pay and maybe sizing in a portfolio. I mean, Capital owns these companies, too. I think the challenge and what makes it hard when you're active investors is you need to actually decide whether you think the benchmark is the appropriate prudent portfolio for a lot of our investors. And I think you may have seen the communication that came out for our chief investment officer, Martin Romo, about this, is our job is really twofold. It's to develop, it's to, you know, help our investors improve their lives through successful investing and beat the benchmarks.

But it's also to be conscious and prudent investors. And if we think that the benchmarks are starting to reflect a concentration of risk that we're just not comfortable with, we have to be comfortable deviating from the benchmarks. and maybe for a period of time seeing that come through in our relative results, because we think it's the right thing to do for our investors. And again, I'm not trying to turn this into a commercial for Capital Group, but I distinctly remember headlines in the newspapers about how Capital Group had lost their way or lost their touch as the Internet bubble was going up. And I happened to take, I'm in our Los Angeles office,

but I spent two years in our San Francisco office just because I was young, and I had friends up there. Those two years happened to be from 1997 to the end of 1999 as the internet bubble was exploding upward. But it was so formative for me because I would be in the office and I would listen to these portfolio managers

and they would say, I can't buy this company. It's garbage. But I'm getting killed on a relative basis because the S&P keeps going up and up and up. And they wouldn't do it. And it was very difficult. And we were getting bad press because we were lagging.

And then all of a sudden, it went the other direction, and Capital came out looking great. Again, I'm not saying that's going to happen again, but I think it is really, really hard to have be benchmark aware, which obviously everybody is, particularly in this environment, but had the conviction to not let it drive your decision-making in these type of buyers if you don't think it's the right thing to do. And so it's never a black and white decision. It's not buy these stocks or don't buy these stocks. We're buying the stocks. We own these companies are good companies. But we're probably, doing it in less of a concentration within our portfolios than you would see in the index

because we don't feel that that level of concentration serves our investors well because there are no facts about the future. And it's not to say the AI will not work out. But there is a possibility that it doesn't. And we don't want to have that concentration if for some reason that ends up being the way the future unravels. You know, the big, the MAG 7, they're making, as you said, a boatload of money. I mean, they're highly profitable. Maybe a little less at the moment because they're investing gobs of capital to expand out the Internet infrastructure. But we've got these companies coming, anthropic, open AI, others, you know, in that ecosystem that aren't profitable.

But the thinking is that they, I mean, obviously revenues are rising very rapidly, but so are CAP-X. How do you think about, is that making queasy as well? I mean, they're real companies, obviously. They're doing real things. They're massive companies, very successful, but it's requiring a lot of CAP-X for them to kind of get to this place. Is that part of your nervousness around what's happening here or not? I think it is a little bit because that was,

the historical corollary would be fiber build-out during the Internet. And it's not that that that, that fiber didn't ultimately end up getting used, it did. And then some. It's just that it was built out at a much faster pace and ultimately was needed. It was a long way of saying there was overcapacity. I don't know if there will be similar overcapacity in this particular CAPEX cycle. A lot of the analysts, Jared included, who are AI bulls, will say, you know, all of that capacity is already spoken for.

There's not enough compute as it is. And they keep building and building and building. That may be the case, but I know there was similar stories back when the fiber optic cable was being put into. And so sometimes, again, Mark, I hope this isn't considered a dodge, but it just comes down to time. All of that capacity, if it is over capacity, may eventually get used, but will it get used in the time frame to meet all these lease obligations and other financial commitments, etc, that these companies are making? And that's the trillion dollar question. It really is. So it makes me a little uneasy, not that I think it's bad investment.

It makes me a little uneasy that whether or not it will be utilized quickly enough to generate an ROI that makes all of the other math in some of these companies and stock prices work. Well, that's an individual, one person, Daryl's opinion. I should caveat with that. That is not a capital group opinion. The one thing I'm going to try to do in the future is have both you and Jared on at the same time. That would be a lot of fun. But I want to think, is there anything we missed, Daryl, that you'd like to bring up? I just didn't open any question.

I mean, we covered a lot of ground. And I'm very sympathetic to what you're saying. I mean, it makes a lot of sense to me. But anything, just before we call it a podcast, anything? I don't think so. Okay. I do want to say, though, that Jared and I are very good friends. Oh, how can you not be friends with Jared?

like the nicest guy on the planet. Yeah, I know, exactly. The one thing I do get to hold over him is I think it was two or three years ago. We bet he had a bet that the Fed would cut rates and I had a bet that they wouldn't. And I won. So that was the one time I think I've done better. And he invited me over my family. He cooked an excellent steak, I have to admit.

So, but no, Jared is a very good friend and a very good colleague. And again, I don't think we're totally on the different sides of the thing. He sees the benefit of the technology probably more clearly than I do. And in that respect, I trust his views a lot. A lot of it is whether or not it helps out in the nearer term and happens quickly enough. Well, we'll give him our best. You know, he is a great guy. And it was so good to have you on.

I want to thank you again. Marissa, Chris, anything before we call it a podcast? No. No hearing nothing? Thank you. Okay. All right. With that, we're going to call this a podcast, dear listener.

We'll talk to you soon. Take care now.