LIZ ANN SONDERS: I'm Liz Ann Sonders.

COLLIN MARTIN: And I'm Collin Martin.

LIZ ANN: And this is On Investing, an original podcast from Charles Schwab. Every week we analyze what's happening in the markets and discuss how it might affect your investments.

Well, hi, Collin. Another week, and bonds and yields are still in the news. So let me tee up one question for you, somewhat big picture. How are Treasury yields connected to things like affordability? I mean, how does it impact the average person out there? And maybe talk a little bit about why this is clearly not just a U.S. problem or phenomenon, but it is broadly across the world, too. So big picture, I just teed up in a very broad way, go.

COLLIN: Well, first, I'd be lying if I said I wasn't relishing the opportunity to really talk about bonds so much. This is kind of a first for me, you know, with yields so high, people care about bonds again. So I'm loving this right now.

LIZ ANN: Yeah. You are popular.

COLLIN: Yeah, exactly. The requests we're getting to comment on things and provide insights, it's really been great. Yeah, I kind of like the loaded question you just passed there about, you know, why do yields matter, right? What are they telling us? How do they impact us as investors or consumers or borrowers, and is it a U.S. thing? It's not just a U.S. thing. Let's just start with that. This is a global issue. You know, year-to-date we've seen most global bond yields from developed market economies rise. And when you look at what's happening in the U.S., and I think for most other economies as well, it's not necessarily for problems, specifically in the U.S. I don't think it is. I think it's due to the shifting outlook for Fed policy. You know, the idea that the Fed was likely going to cut rates, and now they have hiked once and expected to hike a little bit more, and the idea that our growth is strong, resilient. You should see yields where they are right now. Nominal growth is very strong.

That doesn't mean everyone's doing OK. I mean, if you look overseas, France is having some issues right now. We've seen issues over the years with the United Kingdom, where their yields have risen for, I could say, nefarious reasons, if investors are worried about the just sustainability of their debt load, things like that. But here in the U.S., I don't think that's the case. So how does this impact us as investors? I think the most important way is probably through mortgage rates. If, you know, if you're a consumer or a borrower, a potential homeowner, mortgage rates are generally tied to long-term Treasury yields, like the 10-year Treasury yield, maybe the 30-year Treasury yield.

So when long-term yields rise, that tends to pull up mortgage rates as well. And the 10-year Treasury yield just this year alone is up around a full percentage point, 100 basis points. This is where I want to do my standard public service announcement that the Fed does not set mortgage rates. The Fed only sets short-term interest rates. Mortgage rates, like I said, are based on long-term Treasury yields, which are based more on expectations for Fed policy over time, you know, maybe over the next 10 years, if you're looking at a 10-year Treasury, based on growth expectations, inflation expectations. So when the Fed raises or lowers rates at a given meeting, mortgage rates won't necessarily follow that path. In fact, sometimes they can move in opposite directions based on …

LIZ ANN: And by the way, it surprises me that people don't know that. I think I told you this story a while ago. I have a friend that's a real estate agent, and she didn't know that the Fed didn't have control of mortgage rates. So even somebody, you know, in the business wasn't in tune with that relationship.

COLLIN: Well, I think that goes to the idea that, especially when you think about the bond market, there's a lot of maybe some myths out there and maybe some confusion amongst investors. And that's why I really value, you know, our role, Liz Ann, where we get to have conversations with our clients and kind of discuss these sort of things, to empower them with these details that then helps them make better decisions with their investments or get a better understanding maybe about their own finances. But yeah, I'm always surprised by that as well. When the Fed raises or lowers rates, it doesn't necessarily mean mortgage rates will rise or fall.

Now in this instance, the Fed hiked rates and is expected to hike more, and mortgage rates have risen. But again, I think the rise in Treasury yields is driven by that shifting kind of expectation. But there's other parts, other maturities, and other borrowings that that we as investors can focus on. So mortgage rates are generally based on long-term interest rates, but something like a home equity line of credit, a HELOC, that's usually based on very short-term interest rates that the Fed does tend to influence. Auto loans tend to be more short-term in nature. So they tend to be based more on short- and intermediate-term Treasury yields versus long-term Treasury yields. And all this flows into affordability because it makes the cost of borrowing more expensive. It's more money we need to pay back as the borrower. So it can be good thing if you're investing.

So if you don't have many borrowings, you're looking to go into the bond market right now. We do think the opportunity is relatively attractive, even though we're worried that yields might continue to rise a little bit higher. But on the borrowing side, it does make it a little bit more difficult. And right now, especially in the housing market, we kind of have the double whammy of high prices and high borrowing costs. So it makes affordability really difficult.

Let me pass it back to you because, you know, borrowing costs don't just impact us as consumers, but they impact businesses. And, you know, so far, stocks are kind of shrugging off the recent increase in borrowing costs. But how do rising borrowing costs impact the stock market?

LIZ ANN: Well, at its broadest level, the yield that is most relevant to the stock market in terms of things like correlations is the 10-year yield. And we are still running both on a long-term, meaning on a rolling-one-year basis, or shorter-term, rolling-one-month-basis in negative correlation territory between bond yields and stock prices. Tend to get a little bit more of a hit on the downside when yields are rising, relative to when yields are falling. But that is the tightest relationship in terms of your world and my world.

The good news is that we're starting to see a little bit of improvement in breadth. Market breadth as measured by things like the percentage of an index like the S&P outperforming over trailing periods of times, the percentage of stocks within any of the indexes, but let's focus on the S&P that are trading above their 50-day moving average or their 200-day moving average. So there's lots of ways you can measure breadth. And breadth has kind of stunk. There's been a decent amount of halitosis in the market. Slight, slight improvement that we've seen literally over the last day or two. And we've got a few days before this episode drops. So who knows what could happen by the time this episode drops? But good news is we're seeing a little bit of improvement.

Where you do see the impact of higher yields is less at the index level because we've been back in this environment of concentrated performance, a lot of money moving back into the tech space specifically, the sort of AI theme a little bit more broadly. But where you are seeing ongoing weakness is in the more interest-sensitive segments of the market, like utilities, like the real estate sector, which you mentioned as it relates to housing. So that's where you are seeing an impact. If you go down the cap spectrum and into the small-cap space—unlike last year, when non-profitable Russell 2000 stocks outperformed profitable Russell 2000 stocks two to one—it was the nonprofitable stocks were up 20%, the profitable stocks were up 10%. That was last year, calendar year 2025. This year it's close to the opposite.

We have the non-profitable stocks significantly underperforming. And I think that's also a function of higher yields. You also see underperformance and more acute weakness in areas like the zombie stocks, defined as companies that don't even have enough cash flow to pay the interest on their debt. And of course, as it relates to the interest on their debt, that has been going up. So I do see connectivity in terms of the yield environment. It's just not necessarily showing up in these cap-weighted indexes because you have such a bias in performance again to the tech space, especially as we lead into third-quarter reporting season.

And the last thing I'll say is earnings seasons are always important. You know, earnings are the mother's milk of stock prices. But even more so is because I think that the messages coming from companies as they report, these conference calls, arguably, maybe one of the more important macro signals. We think of macro signals as coming from economic data as it comes in or what's going on with geopolitics. I think what companies are going to say, not necessarily just about their earnings that they just reported, but the outlook going forward, for profits, for profit margins, productivity, AI capex, how that feeds into the economy, those bigger-picture messages are what I'm going to be listening for as we head into earnings season.

COLLIN: You know, a lot of the points you made before, it shows how, you know, your world and my world are kind of related when you think about quality. You know, whether it's high-quality stocks and high-quality bonds versus what you would call low-quality stocks. For us, it's kind of simple where we just look at the credit-rating spectrum. And if you're a high-yield company, also called junk or sub-investment grade, those are issuers with credit ratings of BB or below. Those are the ones that tend to be most at risk when rates rise. So yes, on the surface it might seem like higher borrowing cost could be a threat to businesses, but it seems like it has much more of a direct impact in certain areas.

So if you're a junk-rated borrower, specifically the lowest-rated companies, you know, like CCC or even the low end of the single-B-area spectrum, they tend to have pretty volatile cash flows, maybe thin profit margins. So the idea of just a percentage point or two and additional borrowing costs can have a big impact on their ability to pay. And then there's also companies, a lot of times they're smaller companies, they issue floating-rate debt. And their coupon rates are based on short-term interest rates that are correlated to the fed funds rate. So when the Fed hikes, those companies can see that increase in borrowing costs, you know, almost in real time, the next time they have to make that payment. And that's a risk as well.

So we don't see it a risk with the broad credit markets. We still have a favorable view on investment-grade corporate and even high-yield corporates, but we're cognizant of the risks at the lowest parts of the spectrum. So if you are an investor focusing on individual issues, there are clearly risks with low-quality bonds and low-quality issuers.

COLLIN: So Liz Ann, we have a guest this week. So can you tell us a little bit about her?

LIZ ANN: I sure can. It is Dr. Nela Richardson. Nela's been on the show before, and she's the chief economist for ADP Research. I'm sure many of our listeners recognize the name ADP as the big payroll company. Nela is also a regular contributor to Bloomberg and Marketplace from American Public Media. She frequently appears on CNBC, Fox Business, CNN, Yahoo Finance, and in The Wall Street Journal, Fortune magazine, and The New York Times.

Nela has a weekly column, "Main Street Macro," and it examines economic conditions and their effect on small and large businesses, workers, and households. Prior to joining ADP, Neela was at Edward Jones and also served as chief economist at Redfin, while also working as an economist for Bloomberg. She has a PhD in economics from the University of Maryland and has held research positions at the Commodity Futures Trading Commission, Harvard University's Joint Center for Housing Studies, and Freddie Mac.

Nela is a member of the Stanford Digital Economy Lab Advisory Group, the National Academies Committee on National Statistics, the World Economic Forum, Global Futures Council, and the U.S. Monetary Policy Forum.

Well, Nela, I'm so happy to have you on the pod. You and I ran into each other at the Bloomberg TV studio last week. It's always fun to see you in person, and we're really happy to have you on the pod today.

NELA: I'm excited for the conversation. Thanks for having me on again, Liz Ann.

LIZ ANN: Yeah, and as our producer told you, Nela, the last time you were on, it was a very popular episode. And I loved the comment you made as soon as he said that, which is, "I guess people care about their jobs." Well, we care about your job. And let me just start with a really basic question for our listeners that may not be aware of, you know, who ADP is and how the data is gathered for both the weekly and the monthly sort of incredible data that you provide out into the world.

NELA: Yeah, and not to take credit, just to give it focal point to the company, they've been doing this for years. It's really their public mission service to the world of work. But ADP pays one in six workers in the United States. That translates to over 26 million. And over half of all the W-2s that get formed come from ADP-generated activities. So it's a huge, granular, real-time data set. And the unit of observation is something we can all relate to: It's a paycheck.

LIZ ANN: Yeah. And how does it differ from the Bureau of Labor Statistics data that puts out the monthly jobs report that we're all also very familiar with in terms of non-farm payrolls?

NELA: Yeah, there's several differences to highlight. So this is administrative data. It comes from ADP's activities of paying people. And so you ask what our data set is. It's everything on your paycheck, from you know, the age of the employee; how long they've worked for the firm; how many hours, actual hours they worked (we actually know that because we have to pay people by the hour); benefits, whether it's your 401(k) or into your health care. So all those things on your paycheck, we see in an anonymized way, but it allows us to do two things. One, we can count the number of jobs. And two, we can also have this really fine-grained look into pay and how pay is evolving over time.

It also gives us the ability to uniquely match individuals through time. So when we put out pay data, it's the individual's pay this year versus last year, same individual, anonymized of course, but that person-to-person match means we can abstract away from cohort effects, which is a really powerful 15-million-strong tool that we see every single month. That differs from the BLS data. We think of this private sector data as a complement to official government statistics. Official government statistics from the BLS are built on a survey. So employers are asked every month, "How many people did you pay this week?" What we actually measure is how many people are on payroll.

So not only is the data administrative and not a survey, we ask a slightly different question. So whether or not you got paid in the week, but you're still on the company's payroll, we still count you as employed. So if you were off without paid leave, and you didn't get a paycheck that week, you're employed in the ADP count. In the BLS count, you may not show up as employed if you didn't get a paycheck. So there are some nuances. Over time, both of these indices are benchmarked to the same, here we go, big acronym, QCEW, Quarterly Census of Employment and Wages, which is upwards of high 90s census of all the workers in the United States. It's gathered at the state level. The only drag with this data is that it's only available with a six-month lag. So in essence, both the ADP number and the BLS number are trying to capture their basically six-month forward projections of what that QCEW will be like over time.

LIZ ANN: That's great. And I want to talk a little bit more about pay in a few minutes, particularly your new pay insights report. But before we get to that, just share the high-level details of the most recent monthly reading, which, if I recall, in September was a bit stronger than what had been the case in August. And why did that look a bit different from what we saw out of BLS, and are you seeing an improving trend in your data that may be at odds with the perception anyway of a deteriorating trend per the BLS data?

NELA: We are seeing a stronger trend in September, but I think it harkens back to what I was saying about the methodological differences. Over time, these series built in completely different ways look remarkably similar. But month-to-month there might be changes. The previous month in August, BLS was showing more than 160,000 jobs. So they had a very strong month in August. We had a weaker month. I believe ours was revised to 36,000 jobs. And then in September, those numbers were reversed. So they may not capture the same trend at the same time. Over time, they tend to capture the same trend. But what we bolster that data with and that look with is real-time weekly data. And when I say real time, it's as real time as my workday allows. We get the data on a Saturday, and we produce a weekly report on a Tuesday. That was a fast turnaround—I'm not going to lie—for the team to get in the practice of, but we have it down now. It's automated, thankfully.

So what that tells us is that this momentum, this trend, had been going on for several weeks coming out of that August report where we saw a lower number of jobs. So it's not a one-off, it wasn't just for that week. We saw the trend, and anyone who is following NERP, the … what we call the NER Pulse, they saw that trend, and the bigger number that we just released was not a surprise because it had been building for the past four weeks. Liz Ann, I will note, though, it's different numbers this month, but the refrain is the same. It's still health care driving those jobs. We're still seeing additions in leisure and hospitality despite consumer sentiment being downtrodden. We're still seeing weakness in finance and professional business jobs, and we're still seeing modest gains in manufacturing, but slightly stronger gains in construction. So yeah, the number differs, but the story's the same.

LIZ ANN: Do you parse out the information sector or the tech sector, however it's termed, and what are the trends you're seeing there?

NELA: Yeah, we do. Yeah, they've been modest as well. In fact, we've seen outright losses in the information sector. This month it was pretty good, but still modest. About 3,000 jobs created in September.

LIZ ANN: OK. I want to touch on this new Pay Insights report, which I've been doing some reading on, and it's kind of fascinating. And I think it was you maybe that that was quoted as saying, "Once predictable wage growth has been overtaken by complexities of demographic change, persistent inflation, and AI's effect on jobs," and that this new Pay Insights report can more fully reveal the dynamics of today's labor market. So talk a little bit about that report. I was particularly interested in how you parse out both base pay and gross pay, but also job stayers versus job leavers.

NELA: I'm so glad that you like the attractions in this report. We're really proud of it. It starts with about a 14.7-million dataset every single month that we can observe in this individual matched way. We do now break it out into gross pay and base pay. Here's why that's important. Firms are very reluctant to raise base. Generally they pay, and we've done research with the University of Chicago in a paper that was just out over the summer showing this, firms typically like whole numbers. They typically give a bunch of their employees the same whole number. It's usually 3%. If it's not 3%, then it's going to be 2.5%, but it's rarely going to be 2.27% or something in between. They tend to gravitate to those whole numbers. But what we also know is that they are more likely, in a hot market, to give off-cycle bonuses than they are to raise the base.

And so base pay, the research shows, is more aligned with a signal of labor tightness or inflation. Whereas gross pay is more activity-based, and we have actually, I don't know when you'll air this, but we have a blog in the works about overtime, which is showing some really interesting trends in manufacturing, but that gives economists a sense of how labor market activity is changing, which leads to total earnings. So one is for labor tightness as a signal. The other is for consumer spending and purchasing power. Is it going up? Is it going down?

Before we were just offering gross pay. Now for all clients with more than 50 employees, we can look at base pay and look at this really sharp signal of labor tightness. If a firm is raising base, they're really trying to compete for talent.

LIZ ANN: All right, let me pull on two threads there. On the gross pay inclusive of bonuses as a for instance, so bonuses often track corporate profits, too. So does this work tie in anything with regard to profitability in the aggregate or at the company level and the, say, likelihood of bonuses being higher for those companies that do tie some of that extra pay into their own profitability stream.

NELA: Yeah, that's absolutely correct. So we pick that increase in bonuses and the increase of gross earnings, but the timing is also important, too. Most firms tend to give bonuses around the same time. So since they were public about it, for the University of Chicago, it's going to be around June, according to our co-authors. So that means that when firms are giving off-cycle bonuses, it's going to be for Chicago sometime not in June. This is to keep employees in place. So it's an effort to retain employees. Bonuses that are predictable tend to have less power to retain the current workforce than bonuses that are unexpected.

If you expect a bonus every March, and you get a bonus every March, even if it's a bigger bonus, you expected it. And maybe you expected a bigger bonus because the firm was more profitable, and you were obviously the reason why they were more profitable.

LIZ ANN: Obviously.

NELA: But if you get a bonus in November, that's a different signal. It's a bonus you weren't expecting. And so we're tracking those off-cycle bonuses as a measure also of labor-market tightness and strength.

LIZ ANN: You mentioned some interesting trends with regard to overtime. So talk a little bit more about that. I think you may have mentioned it specific to manufacturing. Correct me if I'm wrong, but I just wanted to hear your interesting perspective on overtime and what trends you're seeing.

NELA: Yeah, so manufacturing I think would be very interesting in terms of your stock thesis. Because a lot of what's driving the stock market now is those frontier tech firms and the AI build-out, and that's what you're seeing in this industry. But you have to look beyond headcount.

Looking at jobs numbers alone will not tell you the whole story because this is an industry that is really whipsawed by the demographic shifts, lots of retirees. It's also whipsawed by consumer preferences. We have a client in the manufacturing space that operates 24 hours a day. Can't raise prices high enough to meet all the demands coming from the AI build-out, but also can't find workers to work those second and third shifts. And so they're not alone. There are many manufacturers that are trying to figure out all these things at once: a retiring workforce. How do you get young people in? How do you get people to work second and third shifts? How do you find or upskill those trades? It's a lot to put on talent. So what are our companies doing?

Well, they're trying to make the most of the workers they have with overtime pay. And so we're seeing an increase in overtime pay, but not across the board, not for those manufacturers in the consumer space, like apparel or beverages. We're seeing it in upstream industries. Those industries, whether it's chemical or electricity or steel, that power a boom and power the start of an economic boom cycle. This is a small industry, manufacturing relative to the huge service sector, but it's signaling something very important that this economy is kickstarting in a really interesting way.

LIZ ANN: And manufacturing tends to be a leading indicator for services. You broke a little bit of those connections during the pandemic, at least the first several years of the pandemic, because manufacturing boomed, but services were still shut down. Do you think we might be reconnecting a little bit that traditional relationship between manufacturing as a leading indicator, or do you think they're going to continue to operate on somewhat separate timelines?

NELA: I do think that manufacturing will be a leading indicator, but I think the demographics of the industry have shifted so much that headcount alone won't tell you the signal. You have to look at pay. You have to look at pay premiums. And when we look inside this industry, again, the job changers, this is where the job changers versus job switchers or, excuse me …

LIZ ANN: I was just about to ask you that. Job leavers and job stayers, yes.

NELA: … job stayers. This is where it becomes really actionable because we're seeing those premiums grow. We define that premium as the difference in pay growth between a job leaver, a job changer, and a job stayer. And they have increased overtime in the manufacturing space, meaning people are rewarded for getting a higher-level job. They're rewarded in base a little bit, and a growing bit, but they're really rewarded in gross. That's higher salaries, yes, but also bigger bonuses and, yes, more opportunity for overtime, which is really important in this sector.

LIZ ANN: So are you then seeing a shift up then in the willingness to leave the job because that premium is afforded to those people who leave and go land somewhere else?

NELA: Well, certainly the incentives are there, but you have to match that with the demographics. The demographics are a drag. Older workers tend to leave jobs less frequently, but they're also the most skilled. So it's going to take more to lure an older worker, an experienced worker, into another job than a younger worker. That skills mismatch hits the demographic trend. So we're not seeing a bunch of churn. In fact, when we look at new hires in this industry, it's under 5%, generally, percentage of new hires in this sector. You compare that to like education and health care, which is … could be 15% to 20% in a given month, lot more churn in professional businesses services or leisure and hospitality than in these highly skilled, good sector jobs.

LIZ ANN: So let's stay on wages but tie it into the outlook for inflation and maybe related to that monetary policy. So you've said recently that you don't see wage overheating to be anything to be alarmed about, but that it's something you're keeping a close eye on. So what would move you from sort of watchful to worried? And how does that come into play when you think about the reaction function on the part of the Fed?

NELA: It would take a tectonic shift to move me to worried. Because even when inflation was above 7% in the United States, the labor market was at no risk of tipping over into a wage price spiral. It just didn't happen. And we're nowhere near 7%. In fact, we … the recent data on inflation has gotten better. But here's where I am worried, Liz Ann, and I'm super worried about this. Our research with Chicago showed that during that high-inflation period, 43% of workers saw real wage declines. And the average decline was 8%.

So from 2021 to 2024, we tracked the same 16 million workers, and we saw that decline. Now the way that workers deal with inflation is they switch jobs. They try to outrun it. You can only do that once well. And older workers tend to do it less. So even with job changers in the mix, we saw more than 30% had real wage declines. And that's the starting point for a lot of people. They don't have the purchasing power that they used to. And this, in our view, is the reason why consumer sentiment has been in the dumps, rock-bottom levels, because people, workers, don't care about the inflation rate. They don't. They care about their purchasing power.

LIZ ANN: I couldn't agree with you more. That's why I always chuckle when I … and this happened a lot during the initiation of tariffs where the pro-tariffs folks would say, "Don't worry about this. It's just a one-time level step up in prices." And you and I might live in the weeds of month-over-month versus year-over-year and headline-versus-core and core services ex-housing. But the real world lives in a "stuff is more expensive now than it was before." So that's how they think about inflation, regardless of how we as market watchers and economists think about it. So I think that's a very good point.

NELA: It's absolutely true. It's "Does my paycheck cover it?" And even as an economist, I'll share this personal tidbit. I drive an electric car, so I haven't pumped gas. And I live in New Jersey, so we don't pump gas in New Jersey ourselves.

LIZ ANN: That's right.

NELA: So the double whammy. I haven't pumped gas in a very long time. I did do a nice thing, and I filled my mom's tank. I didn't realize how nice I was going to be until I saw the price. I was really blown away. And this is someone who hasn't been as involved in pumping gas for a while, and I thought to myself, "Wow, this is meaningful." This is a big step up in filling somebody's tank right now.

LIZ ANN: And it's visceral. I think gasoline is probably one of the most visceral prices. And it often defines, when you look at survey data, it often defines how they think about the economy. How they think about inflation is what that number reads at the pump.

NELA: That's right. And sentiment has a big power agenda when it comes to the labor market. It shapes your view of your paycheck. It shapes your view of your job. And in a low-hire, low-fire labor market, you can't outrun inflation. And if you look at where sentiment has been dastardly, it's been for older workers who can't … who don't change jobs as much.

LIZ ANN: Not only that, but I think, and you probably see this in your data, too, that generally on the lower end of the income spectrum, a higher share of disposable income goes toward non-discretionary items. And we've seen more upward price pressure there than some of the discretionary items. So it's almost a double real whammy and maybe exacerbates the K a little bit.

NELA: You know, we track, as part of our monthly Pay Insights report, wage growth, pay growth, by percentile, quartile. So the bottom 25% all the way up to the top 25%. And to me this is very interesting because a lot of people, a lot of researchers, will look at gross pay and say, "Look, the bottom pay growth is growing faster." And that's true. It's because they're working more hours. But if you look at the base, they're basically indistinguishable from each other. So one cohort has to work more just to keep pace with inflation than the other cohort. And I think that's the reality of life now. If, in order to maintain a standard of living, if you don't make that much money, you have to work more hours.

LIZ ANN: All right, now I want to switch to AI. Obviously, part of the recent zeitgeist is the possibility of AI wiping out the human race. We'll leave that to the scientists to debate, but maybe I'd ask you, how is AI impacting the human resources department and how you're thinking about AI and the labor market?

NELA: So we have a canaries dashboard that we have partnered with Stanford's Digital Economy Lab, led by Erik Brynjolfsson. And he and his team is our partner in the National Employment Report, as well. What we're tracking is AI's effect on employment. And what we're seeing, similar to some other research out there, is that AI's impact seems to be showing up in early career for AI-exposed fields, like customer service agents or software developers. We track this data every month. And at this point it looks like for those younger workers between 22 to 25, you're seeing a pretty steady contraction since the rollout of ChatGPT at about 3% a year.

You're not seeing the same level of contraction in non-AI fields. In fact you're seeing a little growth, but taken as a whole, that early cohort seems to be not keeping pace in terms of employment. Where you are seeing growth is for later-stage workers. People longer in their careers were actually seeing a pickup in employment. So it's not a one-note impact. AI that can augment is actually advantaging older workers. AI that is automative is disadvantaging younger workers. But you asked a great question. I love the way you phrased it because you didn't ask me about AI's impact on the economy. You asked me about AI's impact in HR, human resources. And this is our second strand of work because I think this is really exciting. We're taking job postings from ADP, and we're unbundling them into tasks and activities that you can match with official government data known as O*NET, on activities and occupations. So we're unbundling, we're taking a microscope to our data instead of a telescope because AI's impact is at the task level.

And we do this all the time in housing. We can price the fourth bedroom. We can price the powder room. I know that the price of a garage in New Jersey runs about $75,000 in the community I live. So you can actually unbundle using fancy econ stuff.

LIZ ANN: Try having a dedicated parking space if you buy an apartment in Manhattan.

NELA: Oh man!

LIZ ANN: That'll ramp up the price. I don't know, I don't own in Manhattan, but you know, I have fun perusing that kind of stuff.

NELA: Maybe we can solicit from your audience, like what is that, right? I'm sure somebody out there knows. But you could do it for a job. What's for an IT job? We actually unbundled the task and activities tied to standard IT jobs, including those that were developing strategy, educating senior management and clients on technology. Those we saw were high-value tasks, and then we compared them to low-value tasks for the industry, which included diagnostics, testing, documentation, clear differentiation for high-level judgment work when IT jobs versus more maintenance work or diagnostic work. But here's the cool thing: We can track it over time. So there are some high-value tasks that are actually losing value with the inception of ChatGPT in October 2022.

One of those things is explaining technical details. You could use AI for that in a way that you couldn't maybe five years ago. So what we're going to do with this work is not stay in IT. We're going to expand it to hopefully every one of those 700+ occupations in the ADP payroll data even more and track how the value of tasks is changing within industries, within companies. And the hope is that we can help employers see which tasks are of greater value and how to upskill their talent to those tasks.

LIZ ANN: So when is that work coming out?

NELA: In dribs and drops, we've been showcasing it. We've done preliminary work, but this is a big project for us. So we're going to … you'll be seeing it in stages. We're not going to wait till the end of the road to share it. We're going to share it along the way. But it is a journey, it's a big data project, and we're so glad to be kicking off this work.

LIZ ANN: All right, well, make sure I'm one of the recipients of the sharing, even if it's in drips.

NELA: Absolutely.

LIZ ANN: All right, one last question before we close. What are you most optimistic about in the work that you do?

NELA: Can I give you a historical example of why I'm optimistic?

LIZ ANN: Of course.

NELA: So I want to point to the time when productivity was the lowest in the U.S. That was during the oil embargo of the 1970s, early '80s. People just say it's because of the cost of oil, but in reality, that was the time when this huge Boomer workforce entered the labor market. So it's that, and you had this huge unskilled labor force. And then the highest productivity time period, the late '90s, where people credit the internet as the productivity enhancer. And everyone seems to sidestep the fact that those young people, now, 30 years later, was the biggest, most-skilled workforce the U.S. had ever seen and knew exactly what to do with that internet technology. If the internet had come in the 1970s, would we have had the same boom?

LIZ ANN: Very good point.

NELA: I don't think so. I think you need technology. I think you need skilled workers, and yet we expect young people to show up on day one and tell us how to use AI.

LIZ ANN: Well, I do expect my 26-year-old daughter to know the answer to every question I ask that has anything to do with technology, and she usually does.

NELA: That's awesome. That's awesome. We all need those 26-year-olds. My hope is that these young people can be skilled with such knowledge and experience that they can solve all of our woes and worries about AI and truly understand how to harness technology for the greater good for economies, employers, and workers.

LIZ ANN: Boy, that's such an important message, in the midst of this concern about the human race being wiped out. So thank you for letting us end on a very positive note. And I always enjoy our conversations, whether it's in a green room for a quick period of time or on our pod. So very popular guest, and thank you for doing the same again. I'm sure everybody's going to love it. So thanks, Nela.

NELA: Always a pleasure. I'm happy to come on any time.

COLLIN: So Liz Ann, as we always do, let's look ahead to next week. So for you, what matters most? What's on your radar, and what do you think investors should be watching?

LIZ ANN: Well, it's a big week coming up next week. To start it off, we get the NFIB, which is National Federation of Independent Business, Small Business Optimism. And there's a lot of subcomponent parts. So I think that's an important tell in this environment. We already touched on smaller companies. We get some housing data, existing home sales. Probably the big ones would be the Consumer Price Index, the CPI, and all of its component parts. We get retail sales.

And then two days, or a day after we get CPI, we get the Producer Price Index, though, PPI. And one of the things I'm going to look at is we're starting to think about the ripple effects of high diesel prices and how those work their way into the economy. And there's some thought that where we're more likely to see it show up first could be in the Producer Price Index. So looking for some of those ripple effects. We'll also get price data on both imports and exports. And that's important because so much of the AI spend and build-out is driven by imports. And that's where prices have been running a bit high. So those are some of the measures that are on my radar. How about you?

COLLIN: Yeah, the big ones for me will be the CPI, of course. You know, you gave a nice overview there. It matters a lot for the path of Fed policy. So we're going to be looking at, you know, the monthly readings, specifically core CPI. We're looking for, you know, monthly increases of 0.2% or less to give us confidence that the disinflationary trend is back on track. We'll also look at the breadth. This is something that Fed Chair Kevin Warsh has been talking about. So the number of components in the various inflation indexes that are rising at, you know, high, high rate.

So for example, you know, nearly 80% of the components of the Consumer Price Index were rising by 2% or more on a year-over-year basis in August. It's obviously too high when the target for the Fed is 2%. So hopefully that will start coming down a little bit. And then two other things that I'm going to be focusing on: The Fed Beige Book comes out on Wednesday. That's a qualitative look at all the Federal Reserve districts.

And I think that's important because each district president and district in itself, you know, tries to pay attention to what's going on in their region. They talk to the businesses, they talk to the leaders, to get an idea of how are things looking? You know, is growth improving, stagnating, are prices rising, are they having trouble finding labor? Are they looking to cut jobs? Things like that. So it's a very qualitative look.

But I think it helps shape each Fed District Bank president's view about what they think is necessary in terms of monetary policy. So we'll be looking at that. And then finally we get TIC flows, the Treasury International Capital, which comes from the U.S. Treasury Department. It gives us a look at basically who's buying and selling Treasuries, but what we like to focus on are international holders, global holders, because given … I was going to say that given that fiscal concerns are top of mind, I mean they've been top of mind for a long time, Liz Ann, but they seem to be top of mind even more these days. We're going to be looking to see how that foreign demand is looking and to see if some of the large holders like Japan, for example, if they're starting to pare down their holdings. But so far over the past handful of months and years, really, foreign official demand, government demand, central bank demand has held steady. So that's been pretty good.

Well, I think that's it for this episode. As always, thank you for listening. We truly appreciate it. We truly appreciate all the feedback we get. If you don't want to wait a week until the next episode, you can always keep up with us in real time on social media. I'm @CollinMartinCS on both X and LinkedIn. That's Collin with two L's, and the CS is for Charles Schwab.

LIZ ANN: And I'm @LizAnnSonders on X and LinkedIn. Still have lots of imposters, so please make sure you are following the real me. And you can find all of our written reports, as well as videos, and they include lots of charts and graphs and tables, at schwab.com/learn. And if you want to support our show, please consider leaving us a review on Apple Podcasts, a rating on Spotify, or feedback wherever you listen. And we will be back with a new episode next week.

For important disclosures, see the show notes, or visit schwab.com/OnInvesting, where you can also find the transcript.

1. ^ A K-shaped economy is an economic condition where different segments of society or industries move in opposite directions at the same time—forming the two diverging arms of the letter K.
2. ^ https://www.dol.gov/agencies/eta/onet