Better Vantage: AI, the future of work, and what it means for investors

24 minutes 23 seconds

Christine Kashkari: Will AI truly transform the economy and the job market, or will it be just another automation tool? The answer has profound consequences shaping the direction of the economy and your portfolio. Welcome to season three of Better Vantage by Vanguard, a podcast series hosted by Custom Content from WSJ and Vanguard. I’m Christine Kashkari, editorial director at WSJ Custom Programming. With me, as always, is my cohost and our resident expert, Joe Davis, global chief economist at Vanguard. Back in season one, Joe pulled double duty as both cohost and guest to explore what AI might mean for investors. But this debate moves fast, so today he’s back in the guest chair. Joe is, of course, the author of Coming into View, How AI and Other Megatrends Will Shape Your Investments. Joe, welcome to the hot seat.

Joe Davis: It’s great to be back. I tell you, AI is moving fast. I mean, it’s in every conversation I have with clients, with friends, family, my children. If it was relevant last year, it’s even more relevant today. And it’s got some exciting, interesting investment implications too.

Christine Kashkari: So much has changed, so much has remained the same. Where are we with AI and where are you with your projections?

Joe Davis: I’ve been working as an economist, investment strategist for over 20 years. And you know, we started looking at AI and the future of work, what it could do to the workforce, our jobs included, Christine, believe it or not, ten years ago. So it’s been increasing part of my focus, my team’s focus. But four years ago, we took it even more seriously, and we collected the most data in the world to my knowledge on technology, the labor market, what AI could do, look at past technologies and throw them into the soup with everything else that’s going on in the world. We got high debt levels, we got aging of society, to really get a handle. Can we get a handle on where the risk, the opportunities, what the implications are of how AI could evolve and what it could do to growth, interest rates, and our personal lives too? And I can tell you that the diagnosis, the data-driven framework we created when I first started seeing the results three or four years ago, Christine, they were eye opening to me and so much so that we felt compelled to bring that to a broad audience. But the diagnosis, the heart of the diagnosis stands. And that is, one, with increasing odds AI was going to prove to be the most disruptive and transformative technology in over two generations. So, at least as important impacts as the personal computer, but they will look different and we’ll get into that. Secondly is that it won’t be a mass automation loss for all jobs, but there will be pockets clearly of disruption. And I think that’s on a lot of people’s minds. I have children of college ages, on their mind. And so I think we’ll talk about that.

And you know, thirdly is that just as eye openly what we diagnosed in our rich, you know, sort of analysis, Christine, and it’s something I didn’t see coming, is that there is risk and opportunities that emerge in the investment landscape that evolve over the course of the technology cycle. We’re still what I would say is in phase one. So, you see some winners in the markets, but that will gravitate or pivot to phase two. So, I think there’s risks and opportunities for listeners and we’ll get into that. That’s a diagnosis I didn’t see coming. But when we look at in the past, it actually makes a lot of sense once you understand what’s going on. So, I hopefully we’ll talk about them today.

Christine Kashkari: Now, last year, I mean, economists were still debating abstract theoretical protections. Now we have the data, corporate earnings, capex spending, labor metrics. What are they telling you?

Joe Davis: We will see both continued acceleration in AI investment that ultimately over the next three or four years will start to translate into productivity. Some listeners will hear efficiency, which sounds like job loss on automation—we’re going to go there today. But ultimately, it’ll have as important impact on the economy from a macro perspective as a personal computer, but it’ll happen twice as fast. And so that’s generally the good news. But I want to say it’s disruption and disruption sounds negative, but there will be winners and losers during this transition. Over the course of the next two or three years, if we’re right, AI will start to move from just potential way to invest in data centers and chips. This is the, you know, the picks and shovels. And some automation, save time, sounds like job loss. You will also see augmentation. So, I get better at my job or I can do more with the same amount of time. That’s what’s being missed in the dialogue. And we will see three times as many occupations see that effect as the job loss. And finally, what we have not yet seen and there’s still risk to never seeing, and that is what we saw with the internet and with electricity, and that is it creates new industries. It’s a platform. The great example would be with the rise of the internet, there’s companies that started to sell, I don’t know, books online. They started to transform more other sectors. And so things such as online shopping we take for granted today, or the movie industry would not have occurred without electricity. These are things I talk about in the book. That has still yet to be seen, which is why the odds of us entering that higher growth trend world for the next three or four years, it’s become our baseline. But why the odds are certainly not 100%. And where we could be wrong is that we don’t see that augmentative “Hey, Joe, I want this as a copilot for my work, not just as a substitute” or “What are the new industries?” But there’s ways that we can navigate this as workers as well as investors.

Your job in five years

Christine Kashkari: Joe, I’ve heard you talk about this before that if AI does transform and lead to just enormous productivity that would be equivalent to, is it 17 million baby boomers not retiring?

Joe Davis: Not retiring. Well, that’s the unlock, yes, 17 million. That’s both through a combination of saving some time as well as where it’s more efficient and productive in the work that we do. So it’s a combination, but again, it’s overcoming effectively the demographic headwind that many societies have through the disruption that you’re able to get those effects. I would say AI would come at, is coming at the right time given some of the demographic challenges. But I do not want to sugarcoat it. Eighty percent of the occupations in the world will see a meaningful shift in how they do our work, mine included. And I believe yours as well, Christine, over the next three to five years. And when you hear disruption, when I hear that, the human being in me says that’s bad, I’m telling you there’s change. And there are roughly, by our best estimates, one in five occupations, because they are exposed to what AI can do really well, there’s not much leftover and you can start to see a stalling in the job market. And so our assessment was, we’re going to have this profound automation anxiety, I feel it with my friends. I have two college-age children who are entering the work world. I know the concerns and I’m not dismissing them out of hand. They are there and they’re legitimate. But what I’m trying to balance is for those that say mass job loss, by our assessment, it is quite honestly, it is incorrect. They are not looking at the job descriptions because AI can also lead to more throughput. And there’s a lot of tasks in the jobs that I have to do, which is why net-net I’m changing what I do. But I’m going to see at the end of the day, more the need for my job, not less. So, I’d say for the average listener, the way to think about this is think about the percentage of the time at work that you use the computer. So for me, it’s roughly 70%–80%. So, Christine, is it the same? So, that’s the likelihood of AI materially changing how you and I do our jobs in the next three to five years. That’s high. The personal computer, by the way, was 30% to 40%. Now divide that number, 70%–80%, by two. That is how you’re going to spend your time differently on average three to four years from now. So if it’s 80%, I just divide it by two. That’s 40%. That’s two out of five working days I’m going to be doing things differently than what I am doing four years ago. So, what I tell individuals, use AI as much as humanly possible in your job. So, I’ll be bold enough to say this, Christine, let’s try to automate our jobs away. Odds are you actually won’t be able to do it, but you will learn two things. One, you will say, you know what, I’ll become much more proficient in what AI can do. I’m learning. That’s good because I am now more productive, and I will either help my coworkers or I will get ahead of them, I’m going to be honest, in a competitive world. Secondly, I may find I am in one of the 20% of occupations that three or four years from now, we’ll start to see a little bit of, oh, can do a lot of it. All I will say is we have a better indication now rather than avoiding AI. I will have a better sense and better assessment to start thinking about what I could do three or four years from now, because it’s not there, but there is some writing on the wall. With the benefit of hindsight, you could have told in 1990, you’re going to have some job pressures in telecom. And so I’m not sugarcoating this, but it would give you some time because we have identified and most of those affected occupations, there’s a 70% or 80% chance that you can migrate with similar skills to a different field. We’ve taken this really seriously. It’s not 100%, but you can start getting a little bit ahead of that. So it’s in a worker’s best interest to use it as much as possible both from a career accelerant as well as preservation should you be in one of those fields that could be negatively impacted as you look farther out.

Portfolios and the great rotation

Christine Kashkari: What does this mean for investors?

Joe Davis: With all the research we did, the thing that I found pretty striking, which I did not know, was that on average when great technologies emerge over a 7 or 10 year period, there actually tends to be what I would call two phases to the investment landscape in the investing cycle. And that the winners and losers—the asset classes, stocks, bonds, sectors of the market, countries within the equity market—the winners and losers actually changes over the course of that cycle in pretty profound ways. And loosely speaking, I think we’re still in what’s called phase one. It’s the emergent. It’s not, it’s the production of the technology. Everyone starts using it, right? It’s what economists call diffusion. That’s a fancy way for saying it’s just, it’s deploying through the economy, right? It’s the picks and shovels. So today, it’s the chips, it’s the data centers, the computer and internet. It was bandwidth, it’s fiber optics. It’s software and personal computer hard drives. Who do you think is having the highest earnings growth in stock returns? Well, it’s the producers of the technology. We saw this without exception in every past investment cycle, and that is what has carried the S&P 500, quite frankly. It’s been a lot of the large technology companies and they’re, you know, they’re being rewarded for a lot of the innovation that they’re doing. What I was shocked to find is that at some point in the second half of the technology cycle, the investment opportunity shifts and pretty profoundly. But there’s a rationale for it. The leadership changes significantly between the technology producers and the users of the technology. Because remember, if this is a transformational technology, it’s changing the profitability, the revenue, the efficiency of the companies that are now using it. Which means it’s outside the technology field. And it also is because the technology disrupts its own industry the most.

Christine Kaskari: What’s the most important action an investor must take and a mistake that they should absolutely avoid?

Joe Davis: What is clear is investors the highest conviction sort of portfolio exercise is to start deploying new capital or rebalancing your existing portfolio actually away from the technology field. Not because we’re pessimistic on AI, but quite the contrary, we’re actually optimistic on some of its transformative ability because we are likely entering or starting to enter an extended period where AI will not just be picks and shovels but start to benefit the profitability and the new revenue for companies outside the technology space. And so that’s the highest conviction. We could talk about, you know, small cap versus large or other countries, but that is by far the biggest one. It’s the simplest thing. The technology and the Mag 7 have dominated returns for the S&P for some time. There’s a rationale technology base that starts to deploy capital to other parts of the market. It’s what I would call “the great rotation,” and it has happened in the six previous tech cycles. It’s happened in every single one of them.

Christine Kashkari: What about for the mistakes?

Joe Davis: Well, the mistakes I think is being too…is being overly confident in that the AI momentum will continue to solely accrue to the technology producers. That’s similar to what my previous comment is. I think the other thing is I’m trying to keep a dose of humility. I’m so close to this exercise. And so you know, hedging the conviction, the data says don’t make drastic changes in the portfolio. We have a frothy equity market, but it’s not in bubble territory. We have compelling fixed income yields. So I think it’s around deploying new capital. Let’s add some resilience here. It gets a little bit away from some of the concentration in the S&P and it’s done in a thoughtful way that doesn’t lead to radical changes in the portfolio.

Christine Kashkari: You talked a lot about equities. What about fixed income?

Joe Davis: Well, I think we’re talking about resilience of an entire multiasset portfolio. So, I’d say, you know, if AI is transformational like we think in the baseline, value is the first deploy strategy if you’re actively minded in the equity side. The second one is actually just harnessing fixed income and we have real yields or risk premiums, Christine, that are higher even from last year, but it’s been a theme of ours. It provides some resilience if we have an equity market correction and you’re not giving up as nearly as much. So that’s a default. I think if AI is sort of overrated, it’ll push you even more in the fixed income. There you may have concerns of fiscal pressures and long-term Treasuries. You can make a case for a little bit shorter duration. But I think the headline though is you would actually deploy even more fixed income. So, you can see where I’m going here is either world pushes you within equity to deploy capital more into value. And certainly, it’s a rich fixed income allocation in either scenario. And you would go up at actually if AI disappoints. And so that’s what our analytics continue to push. That’s where I would say deploying capital. So, pretty exciting.

What if AI disappoints?

Christine Kashkari: So, Joe, you lay out a compelling argument for AI as a transformative technology that we’ll see sometime in our lifetime. But I’d love to do this exercise with you. The premortems. Not the postmortems, but the premortems. So basically, we’ll lay out a scenario in the future where you are wrong.

Joe Davis: Yeah, sure. It’s a healthy exercise. Yeah.

Christine Kashkari: Let’s do this premortem. It’s 2030. And AI has not delivered the economic promise that we thought and also at the same time delivered a lot of social implications.

Joe Davis: Yeah.

Christine Kashkari: Why do you think it failed?

Joe Davis: Why? I’d say it didn’t evolve as we most likely thought. That’s a world 2030. AI has not become a general purpose technology. The most likely thing that happened though is that AI only automated. It’s what our second most likely scenario is. It’s a 20%. So I got some of the negative effects. Okay, so companies are more efficient, but we have job loss and you get the wealth effects of new industries. Christine or me being better. I’m getting a bigger paycheck because I’m more productive. And so, you know, you got the assembly line for the office. It would be very disappointing. There would be some social impacts because we have higher unemployment. We don’t have the higher growth to go with it. And the stock market is definitely not prepared for that. Society would, that would be disruptive for sure. And we’re keeping a close eye on that because 20% is certainly not zero odds.

Christine Kashkari: What signal should people be watching out for?

Joe Davis: Well, I’ll tell you once we are monitoring right now, Christine, if we’re going down that path in the markets themselves and we’re already starting to see it and that is the Mag 7 start trailing the 493. That’s actually where the concept came from, is this dynamic, that’s one. Secondly, in the labor market, we see less of the automation concerns. We actually continue to see positive even though lower job growth. But here’s the thing, you start to see the hiring rate of college-age graduates start to strengthen again. Why? AI is now a complement of more jobs than a substitute. And then the third one, you would start to see investment rates and new company formation outside of AI companies. And so you start to see new health care companies, you start to see new financial service companies. That’s the creative destruction and the green shoots, knowing that you have a technological transformation coming. And so those would be the three, I think in those order. We’re seeing the first one, second one I think we could see, but time remains. And then the third one could be 2027 and beyond.

Christine Kashkari: Thank you so much, Joe.

Joe Davis: Oh, you’re welcome.

Christine Kashkari: Well, that’s it for today’s episode. Next time, we’re tackling one of the biggest challenges in personal finance, turning a lifetime of savings into income. Join us for that conversation and all upcoming episodes. If you enjoyed this episode and found it helpful, subscribe and share.

Notes:

All investing is subject to risk, including the possible loss of the money you invest. Funds that concentrate on a relatively narrow market sector face the risk of relatively higher share-price volatility. Bond funds are subject to the risk that an issuer will fail to make payments on time, and that bond prices will decline because of rising interest rates or negative perceptions of an issuer’s ability to make payments.

Diversification does not ensure a profit or protect against a loss.

Past performance is not a guarantee of future returns. The performance of an index is not an exact representation of any particular investment, as you cannot directly invest in an index.

This content was created by Custom Content from WSJ, a unit of The Wall Street Journal Advertising Department.