Are AI Stocks in a Bubble?

Is the AI boom built on hype, or is it backed by real fundamentals?

In this video, Morningstar's Michael Field unpacks the complex AI economy, breaking down the five key layers that power the ecosystem, from hardware, energy, and data centers to AI models, software applications, and integration services. Field explores why infrastructure remains the dominant driver of spending today and why understanding where value is being created is critical for investors evaluating AI opportunities. 

The discussion also examines one of the market’s biggest questions: are AI stocks in a bubble? Drawing on valuation analysis, earnings growth, and fair value estimates across the AI landscape, we assess whether current market enthusiasm is justified by underlying business performance. The result is a balanced, research-driven perspective on the scale of AI investment, the demand fueling it, and what investors should consider when navigating this market.

The full research, Opportunities in Artificial Intelligence, dives into:

• A clear investment case for AI, including bull and bear scenarios, valuation signals, and bubble risk analysis.
• A combined equity and manager research framework ranking US AI-themed funds by true AI exposure.
• Identification of AI “hidden gems” and “false friends”, where manager positioning and fundamentals diverge.

Robby Greengold: Could you build us a simple map of the AI economy to get us to understand from electricity to data centers, chips, cloud computing models...help us understand what this map looks like.

Michael Field: Sure, Robby. I'm not sure if I can lay out a simple map, it's not that simple in essence, but I'll do my very best. Ultimately, you've got five kind of nodes on the chain in AI. You've got hardware and computing, which is the one I'm going to come back to which is probably the most important one that leads on to data management. Think of companies like Western Digital, things like this. You've got the third step in the chain which is models, and that's companies like OpenAI and Anthropic which people are going to be familiar with. Fourth step is apps. So that's software firms and how they integrate, and we're going to talk about those in a while too because you alluded to their poor share price performances. And then five, you've got the integrators right at the end of that food chain.

But as I mentioned at the beginning, that first step, the hardware and computing, basically that's the computing power, what's driving the whole system, energy, you know, companies having to build their own energy networks to actually supply this. And then the factories and things like this themselves, where the chips themselves are coming from. So hardware and computing comprises about 75% of spending in AI. So these huge sums that we're reading about at the moment from AWS and Nvidia etc., that's almost entirely generated in that first step in the food chain.

Greengold: Can you talk about the the huge sums that have been invested? What is the scale that we're talking about here? 

Field: Certainly. So you're talking, around $750 billion has been earmarked this year for infrastructure spending, which is an insane amount of money when you think about the scheme of the S&P 500, what that represents, how big a step up it is from the usual capital expenditure that we're seeing for firms across all industries. So the capital expenditure solely from AI now is dwarfing anything from the rest of the S&P put together.

It's crazy and I can understand the hesitancy or the concern behind that poll numbers as well that we saw coming through, right? More than half the people saying we're partly in an AI bubble. It's kind of neither here nor there really saying we're partly, right? That doesn't preclude you from investing still. But it doesn't exactly mean you're hanging off it either, right? A lot of these people are invested in some shape or form through indexes in this AI universe we have.

Field: Ultimately, yeah, the scale of the investment is huge. Is it concerning? It would be if there was no revenue behind it. It would be very concerning. But if I look at Nvidia's numbers here, well, they've got numbers coming out next month, but if I look back 2 months ago, revenues were up 85% year-over-year. For a five trillion market cap company, that seems absolutely insane. The scale of that is huge. And they're generating gross margins of something like 75% on that revenue number. So, you can see why people are willing to buy this stock. You can see why it's gone up 1000% over the last three years or so.

Greengold: Well, what does the paper say? What do you say about whether AI is currently in a bubble?

Field: So this slide here I just put up gives you some of the answer, right? It gives you some of the answer. It doesn't give you all of the answer. Ultimately, valuations is a huge step when we look at whether we're going to win a bubble or anything, right? It's very easy to look at hindsight, when we look at the tech bubble. With these kind of things it's easy to look back now and say, okay, these things were trading at 30 times, you know, eyeballs and clicks and all these random valuation metrics that people basically fabricated at the time to justify such wild valuations.

But if we look at this, a bottom-up price to fair value estimate for this index, the Global AI and Big Data Index, and that's basically looking at the aggregate of what the share prices are for the index constituents and then what we think each of these individual companies are worth and then coming to some metric that describes the two. You can see roughly we're at the right hand side, very end of that chart now, and we're trading at about an 8% discount to our fair value estimate. So that would suggest immediately that we're not in a bubble but it would also suggest that the discount is moderate.

It's not the same as a few months ago, when Kenneth [Lamont] and I luckily wrote the first paper on this, and it was trading at a 20% discount at the time and we couldn't believe our luck. Our conclusion was no, we're not in a bubble and actually there's a huge discount. You should probably be buying these stocks immediately. Obviously that was a short-lived period, but certainly we don't think they're in a bubble on a valuation perspective, and if you look at the inherent demand for these product,s it doesn't seem to be just a flash in the pan. It seems to be that there's actually real demand behind there and something like longer-lasting demand as well i we look the numbers that some of these companies are actually reporting.