AI: Fear of Missing Out or Just Fear?

MIKE TOWNSEND: The explosion in artificial intelligence has been the key driver in three consecutive years of double-digit returns in the market, pushing the S&P 500^® to hit record highs 38 times in 2025.

But we are seeing a significant sector rotation so far in 2026, with energy, materials, industrials and consumer staples all enjoying double-digit returns year to date, while AI-heavy sectors like information technology and communication services are among the laggards.   

So what's going on here? The so-called "AI trade" seems to be changing before our eyes. There has been a pivot to focus on things ancillary to the development of artificial intelligence itself—things like the building of data centers and their impact on energy needs and construction. But there's also been a retreat to the perceived safety and stability of stocks that are less prone to disruption from AI—companies that some are referring to as "real economy companies," amidst larger questions about whether AI can deliver on some of the promised impacts on society, on productivity, on just improving our everyday lives. 

Recent polling has showed skepticism about whether AI is having an impact on daily life. Adoption in the workplace has been slowing, not speeding up, as most AI champions expected. Support for regulatory rules of the road for AI, even at the cost of slowing the technology's development, hit 80% in a Gallup poll last year. 

So, is the AI boom over, or just evolving?

Welcome to WashingtonWise, a podcast for investors from Charles Schwab. I'm your host, Mike Townsend, and on this show, our goal is to cut through the noise and confusion of the nation's capital and help investors figure out what's really worth paying attention to.  

On today's episode, I want to take a closer look at what's been going on with artificial intelligence in the markets and, more broadly, what's going on with AI as it continues its staggering pace of change. Joining me in just a few minutes to share his perspectives on the AI landscape will be my colleague Nathan Peterson, director of derivatives research and strategy at the Schwab Center for Financial Research. 

But before we get to that conversation, I wanted to address the big recent news out of Washington: the Supreme Court's decision to strike down the bulk of President Trump's tariffs. In the 6-3 decision announced on February 20, the court ruled that the president had improperly used the 1977 International Emergency Economic Powers Act, known as IEEPA, to impose reciprocal tariffs on imports from about 100 countries, as well as some of the tariffs on imports from Canada, China, and Mexico. It's a huge blow to one of the cornerstones of the president's economic policy, but the decision has left investors, companies, foreign leaders, even Members of Congress with more questions than answers about what's next. 

Immediately after the decision was announced everyone from the president on down began putting out the message that tariffs are not going away. But ending tariffs was never the point of the Supreme Court decision—it was to clarify the rules for imposing tariffs.

And what's clear is that this ruling has severely curtailed the president's ability to impose tariffs or increase or decrease an existing tariff at a moment's notice.

And all the options that the president has announced he will use to impose tariffs going forward have serious restrictions like defined time limits or time-consuming administrative hurdles.

Here's a quick breakdown of those options and the three major sections of the 1974 Trade Act that will come into play. 

The first is Section 122, which allows an emergency declaration of tariffs but sets a maximum of 15% for a maximum of 150 days. The president announced on February 20 that he would impose a temporary 10% global tariff beginning on February 24. He said the next day that he would up that to the 15% level, but as of February 24 he had not filed the order to do that, so the global tariff kicked in at 10%. There are exceptions for a variety of products, including critical minerals, automobiles that are already subject to their own tariff regime, and agricultural products like beef and tomatoes. At the end of the 150 days, which will come on July 22, Congress would need to vote to extend the tariff past that date, which it is unlikely to do given the deep divisions on Capitol Hill over tariffs generally. It also would not be surprising to see the Section 122 tariffs challenged in court, as, similar to IEEPA, the section has never before been used to impose tariffs.

There is also Section 232. This provision allows tariffs on products or sectors that threaten national security. The law requires the Commerce Department to conduct an investigation, issue a report, and provide time for public feedback. I've often referred to this as the regular way that tariffs happen. It's the most common way tariffs have been imposed for decades and is the mechanism that the administration used last year to impose tariffs on steel, aluminum, copper, furniture, and other products. Those tariffs currently in place were not impacted by the Supreme Court decision. 

Finally, there is Section 301. This provision allows tariffs on countries after an investigation by the U.S. Trade Representative's office determines that another country is using unfair trade practices. It's a similar process to the Section 232 tariffs, with both typically taking several months, and often a year or more. 

There are nearly a dozen investigations under Sections 232 and 301 already underway, and the White House has indicated that it will quickly launch more. But it's a stretch to think the administration can complete those investigations and impose tariffs before this summer. 

Another big question after the ruling was about tariff refunds. The Court was silent on the question of whether the estimated $175 billion in tariffs paid by companies to date will have to be refunded, but it seems likely they will have to be. 

Litigation is certain. Companies from Costco and Staples to J. Crew and Bumble Bee tuna are among dozens that pre-emptively filed lawsuits seeking refunds. FedEx was the first major company to file a suit after the court ruling, and lots more companies are likely to follow. Lower courts will have to clarify the how and the when, and companies may be required to file burdensome paperwork to claim refunds. The process could take months, even years, and the administration is likely to make it as difficult as possible. Treasury Secretary Scott Bessent seemed to preview the administration's stance on February 20 when he said about refunds, "I've got a feeling the American people won't see it."

In his dissenting opinion, Justice Brett Kavanaugh wrote that the refund process "is likely to be a mess," which certainly seems to hit the nail on the head. But it doesn't negate the fact that a refund process will almost certainly unfold in the weeks and months ahead.

Another big takeaway is that there's no question that the global trade environment is dramatically reshaped. There will be some clear, short-term winners. Brazil, for example, will see the 50% tariff Trump imposed fall to 15%. Imports from Vietnam, which have spiked in recent months, will see the tariff fall from 20% to 15%. U.S. companies may accelerate their imports from these and other countries in the near term to take advantage of the lower rates. 

Meanwhile, there's a question about whether trading partners like Japan, South Korea, and the European Union, that struck deals with the United States in recent months—all of which set the tariff level at 15%—will seek to revisit those deals. Europe's trade chief has already called for a freeze in the ratification process of the U.S.-EU trade deal. The frustration comes from the fact that the new global tariff the president launched on February 24 gets added on top of existing tariffs—meaning the EU will end up subject to a higher tariff rate than it agreed to in the trade deal. 

All of these changes bring a new set of uncertainties for U.S. companies. The tariffs landscape they had settled into last fall has been upended once again and now companies have a new set of questions without clear answers. When, and whether, will they get refunds? How long will it take for longer term tariff rates to be settled? Should they revamp supply chains? How will customers react—will they expect to see lower prices right away? Will some customers demand refunds from businesses for the higher costs they paid? 

At the end of the day, the Supreme Court decision was not that surprising—a majority of the justices had seemed quite skeptical about the administration's position during oral arguments back in November. That may be why the market reaction to the decision was fairly muted, as investors seem to have priced in the outcome. 

But the scope of the decision will have profound ramifications for companies, investors, and the economy that have only begun to be sorted out. Tariffs are likely to dominate the economic discussion for months to come.

On my Deeper Dive today, I want to take a closer look at artificial intelligence and how investors are thinking about AI right now, both as an investment opportunity and as a broader driver of economic activity, from jobs to inflation to productivity. To do that, I'm pleased to welcome back to the podcast Nathan Peterson, director of derivatives research and strategy at the Schwab Center for Financial Research. Nate has been with Schwab for more than 20 years in a variety of roles, primarily focusing on research and education about options. He has been closely following the rise of artificial intelligence over the past few years, particularly AI's rapid growth as one of the primary shapers of the day-to-day equity markets. 

Nate, welcome back to the podcast. So great to have you on. Thanks for taking the time to talk to me today.

NATE PETTERSON: Thanks Mike. It's great to be here.

MIKE: Well, Nate, I think every investor, whether you're a trader who watches minute market movements every day or just someone who checks their portfolio now and then, is aware of what a powerful force artificial intelligence has become, not just in the markets, but in our everyday lives. As investors, we all want in on the AI trade. And in our daily lives, we're all trying to figure out just how AI can help make things easier, as well as what we should be wary of.

But I think there is also a lot of confusion out there about what AI can and can't do and just how transformative it will be. And as investors, we've seen, especially recently, how volatile that AI trade can be. So I'm really looking forward to this conversation as we try to sort through how investors should be thinking about AI right now. But before we get to that, let's start with something really basic. Exactly what do we mean when we say artificial intelligence?

It's really exploded into the mainstream in the last couple of years, mostly because of the rise of large language models like ChatGPT, and its many competitors. But AI has been part of our lives for decades now, even if we may not have recognized it. So Nate, what is AI?

NATE: Yeah, Mike. Artificial intelligence or AI represents a broad field of systems that aim at mimicking human cognitive functions like reasoning, learning, problem solving or decision making. And AI isn't a new branch of computer science. Our listeners are likely familiar with virtual assistants like Apple's Siri or Amazon's Alexa, and they've been around for over a decade. But AI really shifted into a higher gear with the launch of OpenAI's ChatGPT back in November of 2022. AI is primarily specialized for specific tasks rather than artificial general intelligence, AGI. That would be your Terminator 2 type of scenario in terms of a fully autonomous AI system. But at this stage, you've probably heard a lot about agentic AI, which has been on the rise. Agentic AI describes systems designed to act autonomously as agents on behalf of humans and other AI systems—learning, making decisions, adaptability, goal orientations. They have clear objectives to achieve. So you may have of the terms like AI chatbots and AI assistants. So for context, chatbots, they allow humans to have conversations with AI and are designed to mimic human conversation. AI assistants are also chatbots, but they utilize other tools to expand on those capabilities. For example, if you use OpenAI's free version of ChatGPT, that's a chatbot. And if you pay for their premium version of ChatGPT, that's an AI assistant. So it's more powerful. AI assistants are still reactive though, meaning they provide responses based on your input, your inquiry. But the next step up there is the AI agents. And these AI software programs are more autonomous. They have a wide range of functions that extend beyond just natural language processing. They solve complex tasks including software design, IT automation, coding, and conversational assistance. On IBM's website, they use the analogy of a football player who has both an assistant and an agent. That assistant is there to serve based on whatever the football player requests, while agents are going to be more proactive in going out there and finding the best opportunities on behalf of that football player. AI agents are not completely autonomous. They do operate with human-defined parameters, like objective setting, permissions, and access controls. And most AI agents do have human intervention. This is so they can provide oversight, they can do strategic alignment, and then a big topic, they can ensure ethical compliance.

MIKE: Well, that's a great foundation to level set. How do you think about where AI started and where it is now?

NATE: Very good question. You've probably heard a lot of talk around which inning are we in, even regarding to the infrastructure build-out or AI in general. There's been a lot of estimates out there. Let me just point to an article from our colleagues at the Schwab Center for Financial Research, Liz Ann Sonders and Kevin Gordon, who have been writing about the evolution of artificial intelligence. They initially described it through a three Cs framework, Create, Catalyze, and Cultivate. They've since updated that third C from Cultivate, which was intended to capture solely the benefits of AI, to now Cascade, which is intended to account for the potential AI disruption.

And for our audience, just know that we will provide a link to the AI Cascade article. So Mike, the first phase deemed create represents the foundational breakthrough period involving the development of large language models, the hyper scaler computing build outs, and the launch of generative AI as a commercial product like ChatGPT.

The next phase is catalyze, and this refers to the second order effects that stem from those computational demands. So if create was about demand for silicon or investing in semiconductors to ramp up computing power, catalyze is about the increase in demand for the necessary physical infrastructure to accommodate that computational demand.

So think about things like data centers and what goes into them: steel copper electricity and water for cooling. As their article points out, since these types of capital expenditures, cycles, historically overshoot, investors are likely wondering where we stand in that capex arc. The last phase, which Liz Ann and Kevin suggest we may currently be in is Cascade. And this involves the diffusion of AI across industries, which will potentially enhance innovation and productivity, but also provide disruption in the form of competition. So as the article states, "companies that integrate AI to enhance margins and pricing power may outperform, while those whose offerings are easily replicated by AI models may face structural headwinds." So Mike, while all three of these Cs are overlapping phases, we appear to be in this sorting out phase as the cascading effects of AI proliferate throughout the global economy.

MIKE: Yeah, I think about the fact that we're in that sorting out phase, that's such a great way of thinking about it. And I think a lot of investors are thinking about that too. You know, there's both a lot of excitement around the potential benefits of AI and, almost the same level of concerns. We're still early on in our understanding of what AI can do, but when you have your trader hat on, you're looking out on the horizon for where AI is headed. So share with us some of the top tier things you're seeing, and also maybe give us your thinking on making decisions about when to invest in a company in this space.

NATE: Of course there's a lot of excitement around this technological renaissance, as I've heard it described. And this is a massively transformative technology for the global economy. If you look at what Jensen Huang the president and CEO of NVIDIA, says about AI, he said it's becoming the foundation for the largest infrastructure build out in human history.

If you look at the amount of spending from the likes of Alphabet, Amazon, Meta, Microsoft, these are what we call the hyperscalers, they're expected to spend $600 billion just between those four in 2026 on data centers and AI infrastructure. And some are estimating $3 to $5 trillion is going to be spent on AI by 2030.

Mike, these types of mega cap tech stocks have been doing so well over the past decade. They've historically been cash flow rich. They've had fortress balance sheets. They have wide moats in their business models. They have sticky ecosystems. They're an asset light business model. And yet over the past six months, there's been a little bit of disparity around, or questions from investors, around this massive spending because it's quite a reversal from what they're used to seeing out of the Mag-7.

Let's go back to some of the potential benefits of AI. The most obvious benefit is productivity gains. 

NVIDIA CEO Jensen Huang calls AI a massive computing platform shift that will help transition labor force from task to purpose. So when you think about tasks, we're talking automation and efficiency, streamlining operations. AI can take over repetitive, time-consuming tasks. And this will free up human capital to focus on purpose. AI can also analyze vast data sets that can help predict patterns or trends to assist with optimization of workflows and can also help avoid, say, mechanical failures. In terms of the impact on industries, think about healthcare in terms of disease detection, faster drug discovery. Within manufacturing, there'll be supply chain optimization, inventory management. Within marketing, they already do targeted advertising, but this can even become more efficient, potentially more profitable. Within mining and energy, you can optimize extraction, you can improve safety and you can have these predictive models that can come in to assist with maintenance. So I could go on and on in terms of the number of industries that it's going to impact, but we're still kind of figuring that out as well. 

On the consumer level, many who are listening to this call have likely already interacted with the likes of a ChatGPT or a Claude or a Grok or a Gemini. And if you can imagine in the future, eventually, we're likely to have a 24-7 personal assistant that's going to be available to answer any questions that you have. It's going to help organize your life by handling all of your tasks, whether it's bill pay or scheduling appointments, booking tickets, even purchasing a home, managing financial investments. You get this computational power, this assistant that's really going to provide a lot of leverage in your daily life activities.

On the corporate level utilizing AI, there's going to be enhanced product service offerings, productivity boost through automation, employees are going to have more power and resources to function with. So some, perhaps not all technology companies, and this includes software companies, will likely be able to implement AI to help with operational efficiencies and reduce OPEX or operational expenditures. And as they use the power of AI, they'll be able to enhance their product offerings, which ultimately could boost customer demand and adoption. 

And Mike, with all the concerns about AI causing company disintermediation, and this just means AI replacing the middlemen or traditional service providers with automation, there'll be companies across multiple industries that will or are already using AI to their benefit, even within the software industry.

Salesforce recently said it is building plugins for Anthropic's latest AI offerings. DocuSign announced a partnership with Anthropic, and Intuit announced plans to develop custom AI agents with them as well. So as an investor, you likely want to ask if the companies that you're invested in, are they embracing this change? Are they enhancing their offering through AI, and of course, is this ultimately showing up on the bottom line?

MIKE: When I've been talking to investors over the last year or so, this issue of jobs really comes up over and over again—this notion that AI is a job killer. But there's a lot of diverse opinion out there about just what that means. I mean, AI obviously can't replace lawyers arguing in court or a doctor's personal interaction with you about your health. Is AI just going to be replacing those basic jobs, those kind of repetitive task oriented jobs that you mentioned?

I've heard it said that those who learn to work with AI instead of railing against it are the ones most likely to succeed. So how much and what kind of job loss is already happening?

NATE: Yeah, Mike, the jury, it's still out as to how disruptive AI is going to be on the labor force. But it is certainly top of mind for investors, if not for your everyday employee. It's one of the biggest concerns out there. And it seems likely that there's going to be some industries that will be negatively impacted while others stand to benefit. So for example, if you have entry level tech jobs, programming or coding jobs or customer service jobs that can be replaced by AI assistants, they're likely going to be negatively impacted. And if you look at the recent headlines around corporate layoffs, they've been concentrated in the technology space. Maybe you've heard that Meta, for example, is having 50% of its coding or more being done by AI now. Well, that's labor replacement there.

It's not just Meta platforms, it's Amazon, Alphabet, Salesforce, they've all announced plans for headcount reductions over the past six months. But I would also say on the other side, keep in mind these hyperscalers are doing massive spending on the infrastructure buildout. So some of these layoffs could just be related to streamlining operations or making adjustments in operational expenses. 

It's probably worth noting, since it grabbed a lot of headlines, but Citrini Research put out an article entitled, "The 2028 Global Intelligence Crisis", which was a hypothetical report from the future that painted a pretty grim picture of how AI could erode the labor market, specifically in white collar jobs. And the stock market reacted pretty negatively to this report. Now, this report wasn't based on any facts, but the market response still kind of speaks to investor sensitivity around the potential for this disruption. And it still just speaks to the elevated concerns on the topic.

Let's speak to some of the positives on the job creation side. If you think about the resources that are gonna be needed to accommodate the AI infrastructure build-out, jobs tied to data center construction, like construction workers, plumbers, electricians, high skilled technicians, network technicians, maintenance—so you also need those skilled technical workers to take care of maintenance and maintain these data centers—you can see how this would be a boon for the labor market. 

Jensen Huang, who has a positive outlook on AI job creation, recently described AI as a five-layer cake spanning across energy, chips, data centers, models, and applications. And he believes each will create new demand for labor. So he expects the largest economic benefit to emerge from that application layer, the top of the cake, after all the foundational parts have been put in place. At the application level, AI models are integrated into our real world systems and are transforming industries and delivering economic value and productivity.

MIKE: You know, Nate, it really seems like headlines are driving the market. You mentioned this Citrini Research paper, which despite being a hypothetical scenario, seemed to drive a market reaction virtually all by itself. So are these headlines what's behind the sell-off in the software sector and other industries like transportation, logistics, insurance, wealth management companies? Have they just been reacting to these kind of one-off press releases and papers that come out and seem to produce a big reaction in the market?

NATE: Absolutely, Mike. It's the frequency or velocity of these headlines that have been coming out, most of them from Anthropic, I would add, but there are a lot more AI companies out there than just OpenAI or Claude. We know that the potential for having an AI company disrupt some of these industries just seems to be easier and easier, in terms of the headlines that have been coming out.

Mike, perhaps the industry that's been impacted the most by AI disruption concerns, and it's seen the most wealth destruction this year, and that's software. So many of the well-known SaaS or Software as a Service, software names like Salesforce or ServiceNow or Workday, they're down roughly 50% from their peaks and they're frequent guests on the 52-week low list. And this is driven in my view by two primary concerns. Number one is obviously competition from AI that is going to replace those business models. And if you don't have Anthropic that can have a comparable system that can replace these incumbent business models, one would think that they can offer at a lower price, which would put pricing pressure on that offer from those companies. So that could impact margins negatively. At least that's the concern. 

The other concern is that because AI will enhance productivity for corporations, corporations can then swap employees for AI—that will reduce the number of seats or the number of employees at the companies. And because these software companies are collecting revenue from licenses, the number of seats that they're able to issue through these subscription revenue services, this potentially can translate to less revenue for those software companies.

But it's not just software. We've also seen several AI-related headlines that have fueled disruption concerns in industries like real estate, like you mentioned, trucking, logistics, wealth management. For example, Online platform Insurify released an AI-powered comparison tool on ChatGPT that allows users to compare car insurance rates. AI firm Algorithm Holdings developed a tool called Simicab, which the firm deemed as "the world's most well-orchestrated transportation platform." Anthropic announced an AI-powered vulnerability scanning tool, this is Cybersecurity Impact, called Claude Code Security which is used to analyze and detect complex code vulnerabilities that conventional methods often miss. And then on February 24, Anthropic conducted a virtual presentation where they announced 10 new AI tools for Claude Cowork that improves things like design, human resources, wealth management.

So Mike, these types of headlines have really impacted investor sentiment, and they've hit the sell button before they're asking questions. It's the sensitivity of the market around these types of headlines and the potential impact that are impacting investors. As we discussed earlier though, these types of industries potentially stand to benefit from AI, both from internal operational efficiencies and potentially providing a more robust offer to clients.

So while investors appear to be managing risk by selling exposure in those areas, the net impact of AI on these industries could be a benefit, and we'll just say it's unknown at this point in time.

MIKE: Investors seem to be asking whether AI can disrupt the business model of the company I'm invested in. And that seems to be kind of driving a lot of that sentiment right now. But Nate, what about some of the other risks that you're watching that could be a pain point for the AI companies themselves?

NATE: Top of the list, and this is not something new. I'd say that if you look at the performance of the Mag-7 or the Fab-5 over the past six months, it's underperformed the market. There's been this broadening of the market and, at least in part, that's due to concerns around overspending by the hyperscalers. This kind of points back to Liz Ann and Kevin's article about that capex arc and are they spending too much money? Again, these used to be these cash flow rich fortress balance sheets that are now using that money to go out there and invest in AI, which is a question mark in terms of how much profitability they're going to achieve or whether it's going to be commoditized and it's going to be difficult to make a lot of money. So that also points to ROI, return on investment, potentially for these hyperscalers. So that's led to some investor pause around that space.

The AI disruption, that's still an unknown, but it seems apparent that there's going to be winners and losers, and if you don't know how to select from the winners and the losers, and I'm not sure if anybody knows at this point in time, that at least kind of increases the tail risk on many of these types of companies.

And then there's other ancillary concerns, environmental issues or regulatory issues. And, Mike, I wanted to ask you, how do you see regulation playing out in Washington when it comes to AI?

MIKE: Yeah, you know, honestly, I think we're still quite a ways from Washington getting its act together when it comes to regulation. I often make the comparison to how Washington was wrestling with cryptocurrency about three, four years ago. Members of Congress at that time really just were not educated enough about the crypto space to know how to go about regulating it. It really took another couple of years for members to get their arms around digital assets. And the result ultimately was the first major legislation regulating crypto passing Congress and being signed into law last summer. Right now, the level of understanding in Washington about AI, kind of where the understanding was about crypto three or four years ago. Everyone kind of generally agrees that we need some guardrails, but no one seems to know how to get going on that or what those guardrails should look like. The EU has comprehensive regulation that was thought maybe that could be a model for the U.S., but that just hasn't happened yet.

We have some states trying to take the lead on regulation, but then President Trump signed an executive order in December to try to ensure that there are federal standards for AI regulation and that those standards don't stifle innovation so that the U.S. can be a global leader in AI. I really think the jury is still out on whether that will happen.

NATE: Yeah, Mike, you know better than most that it can take a while for Washington to move on anything. I do think this is an area, as I said before, where everything is moving so fast that it can be hard for any government to react. And even when Washington thinks it has regulations in place, other issues can crop up. I've seen also headlines around the "not in my backyard" concern, for example, which is really a grassroots issue that has its own set of implications for elected officials.

MIKE: Yeah, you're absolutely right. I've said this before on the podcast, but the issue of the environmental and community impact of data centers is becoming a huge political issue where I live in Northern Virginia, which has the highest concentration of data centers on the planet and 70% of all internet traffic runs through them. Electric bills are up; water bills are up. People complain that the ceaseless noise of the fans keeps their kids up at night. But at the same time, there are 31 data center projects currently approved in one Northern Virginia County on top of the 33 data centers that are already there. There is of course some short-term job boost because of construction, but a lot of long-term concerns, I think, for the neighbors.

But Nate, this raises an important question for investors and that's about the opportunities in what I refer to as companies that are ancillary to the AI boom. Construction, for example, utilities. In fact, as you know, the energy sector is outperforming everything so far in 2026, up about 23% year to date. The utility sector is second, up about 16%.

While of course the two sectors that house the big spending AI companies, information technology and communication services, are both offering negative returns year to date. So is that just sector rotation or is there something bigger going on here? Should investors be looking at these maybe less exciting sectors for opportunities right now?

NATE: Well, Mike, I think it's both AI disruption concerns and rotation. I think you've noticed the broadening of this market rally that we've seen as evidenced by the Russell 2000 really kicked into gear last October, the S&P equal weight outperforming the S&P market cap weighted. And there's been a valuation reset in many areas within the tech space, which are driven by those AI disruption concerns.

But what it's also done is it's driven investors towards relative safety areas of the market—sectors you reference like industrials, utilities, real estate and energy. These are the sectors that are viewed as obvious economic beneficiaries of the AI investment cycle, the picks and shovels of this secular growth story and are essentially immune to the headline risk around the AI disruption, the disintermediation that we've seen recently. But these aren't under the radar plays and many stocks have been strong performers over the past year. So as an investor, if you're considering investing in these sectors, I think you have to consider the elevated valuations versus historical average. And you'll likely need to monitor for inflections in that capex arc that we mentioned from the article.

It's that point at which the investment binge hits some kind of a saturation point. And Mike, that can be very difficult, if not impossible, to predict for investors.

MIKE: Yeah, that's absolutely right. An important point.

Nate, we've talked through some of the benefits and risks of AI. We've talked about how there has been some targeted disruption to certain kinds of companies and their stocks in recent months. Given all that, let's wrap up with how should we be thinking about the investing landscape right now when it comes to AI?

NATE: The winners and losers is going to be an ongoing research investigation by investors. 

Again, it can be a little bit challenging just because AI is so fast moving. But I'm going to point back to the AI cascade article from Liz Ann and Kevin once again, and I'll share some of their suggestions. So number one, they say monitor earnings revisions outside of mega cap tech. So be on the watch for trends and those revisions in both the sectors and industries. 

Number two, watch for capex to revenue ratios for signs of peak infrastructure spending. And again, this refers back to the capex arc, you're looking for signs of that investment saturation. As for individual companies, focus on pricing power and margin durability as AI diffuses across the economy. 

And lastly, Mike, you know that there's been a lot of discussion over the past five to seven years about this top-heavy market cap-weighted S&P 500. A lot of the large to mega cap techs are in so many ETFs and mutual funds that are out there. And in part this has to do with the digitization of the global economy, but the weighting in these ETFs and mutual funds can be skewed to a lot of technology stocks because of that. So if you're an investor and you have some of these funds, you probably want to take the time to review their holdings, look at their individual components of those ETFs or mutual funds, just to understand how much AI exposure that you have. And if you feel there's too much concentration, then consider rebalancing to some of the other sectors that will not be as reliant on this AI boom.

MIKE: Nate, great thoughts as always. Really enjoyed this conversation. We'll definitely have you on later this year because probably the entire sector and industry will have changed given how fast AI is moving. But I want to thank you for joining me today. This was a really interesting conversation.

NATE: Great, thanks so much, Mike. Really appreciate you having me on.

MIKE: That's Nathan Peterson, director of derivatives research and strategy at the Schwab Center for Financial Research. You can find Nate's commentary at schwab.com/learn.

That's all for this week's episode of WashingtonWise. We'll be back in two weeks with a new episode. Take a moment now to follow the show in your listening app so you get an alert when that episode drops and you don't miss any future episodes. And don't forget to leave us a rating or a review—those really help new listeners discover the show. For important disclosures, see the show notes or schwab.com/washingtonwise, where you can also find a transcript.   

I'm Mike Townsend, and this has been WashingtonWise, a podcast for investors. Wherever you are, stay safe, stay healthy, and keep investing wisely.