Introduction and the Rise of the "Magnificent Ten Stocks"
In my opinion, I believe these “Magnificent Ten Stocks” as they’ve come to be known, are overly valued, and there are a myriad of ways to quantify that on a spreadsheet. I do believe these AI thematic stocks have formed a bubble, but you know, not all bubbles are made from the same material.
Sometimes bubbles are made of soap suds and they pop almost as quickly as they form.
Sometimes, a bubble can be made from mylar. This one seems to be pretty resilient, but just recently, I believe we saw evidence of how fragile this bubble might actually be at the end of January.
The Factors Behind the AI Bubble
I think the key to understanding the magnitude of this event is to understand how it formed in the first place. Between 2021 and 2022, inflation began to increase significantly and reached levels that had not been seen in over 40 years. That had a chilling impact on patterns of consumption. You quit buying as many things as you once did.
We can see that by looking at the ISM manufacturing index from its high in April of 2021, through its low in early 2023. Any reading above 50 is indicative of expansion.
Here, we can see that the index dropped from a reading of 64.9, to 44, in almost two years. Manufacturers didn’t just decide to stop making things. They responded to your patterns of consumption. Manufacturers broadly employed technology in an effort to make fewer things, with the same profit margin, with fewer workers, and as remotely as possible. This was an inflation crisis driven rapid adoption of technology.
Â
The problem is that this graphic suggests that the crisis that drove the bubble in the first place might be over. The ISM manufacturing index is once again above 50, indicating that expansion is underway. This is being driven by your patterns of consumption. Now the emphasis is seemingly on being bigger, not leaner. We see this in other metrics as well like capacity utilization, durable goods orders, factory orders, and new orders. So, the million-dollar question remains, how much longer will this bubble remain, if the crisis that drove it in the first place, is seemingly over?
Â
For me, this is the 20,000 pound gorilla in the room. For those of us who remember Y2K, in my opinion you can see very stark parallels between these two periods of time. The one key difference is that I could circle January the first of the year 2000 on a calendar. At 12:01 AM on that day, that crisis ended. We don’t have that luxury today.
Can certain narrowly defined sleeves of stocks melt up without a rational spreadsheet justification? Yes, absolutely. It happened 25 years ago.
Back then, Cisco, among other technology stocks did that very thing. In the lead up to Y2K, the Federal Reserve once again made an enormous amount of money, extremely available, and very cheap. Companies largely borrowed, and bought their way through the crisis. As a result, we saw the basket of technology skyrocket to untenably high levels.
After 1 January, all of the machines worked, the crisis was over, orders dried up, and the values of the stocks plummeted.
These companies did not become 80% less valuable overnight. In my opinion, this was a collapsing cascade of algorithms. Understand that the high-frequency trading that we had 25 years ago was driven by horribly outdated technology by today’s standards. Today, through high-frequency, algorithmic driven trading, billions of dollars-worth of securities can be exchanged in a matter of seconds. I don’t know if we will see a similar phenomenon this year, but it will not surprise me if there is a retracing of gain among these stocks.
From time to time, when I draw a correlation between Cisco in 2000 and Nvidia in 2025, it’s not uncommon to be met with “it’s different this time”.
Actually not. There is a nearly perfect correlation between the patterns of behavior that we observed 25 years ago with Cisco, and today with Nvidia. Now, I can’t guarantee that Nvidia won’t rocket through the top of the spreadsheet, but you can’t ignore, what seems to me anyway, to be mounting evidence that maybe a correction in valuations may be closer at hand than not.
At the end of last year, we observed that 10 stocks comprised 40% of the weight of the S&P 500 Index. There are 503 stocks in the S&P 500 Index. Ten of them, carried 40% of the weight. We have never seen anything remotely close to this in the history of the index, and frankly, it dwarfs what we observed 25 years ago during Y2K, the last time we observed a rapid adoption of technology driven by a crisis.. Let that sink in.
Additional Reforms and Impacts on Student Loans: A Holistic Financial Spectrum
In my opinion, there is much Pollyanna in the media regarding how this might impact the market. There are 503 stocks in the S&P 500 index after all, correct? Yes, but remember they are weighted by capitalization. This is not an evenly weighted index. Looking at a graphic such as this tells me that indeed the fortunes of the index could largely be driven by only a handful of stocks.
I am seeing bombastic headlines suggesting that the index itself might be poised for a correction, and detail how the stock market might experience a long, sharp, and deep pull back this year. Well, I suppose that’s possible at any time, and at a high level I agree, in my opinion the headline index itself might appear to be struggling should 10, out of 503 stocks, retrace gain in a significant manner.
However, it is also my opinion that much of the remaining market is undervalued, and as gains are harvested, those proceeds may find new homes in other, undervalued areas of the market. It could be that while the index appears to be losing ground, investors notice that their accounts are appreciating in value.
Market Dynamics and Risks of Overvaluation
So once again, we cannot be caught up with sensationalized headlines. You must look more deeply than the surface to appreciate the odd period of time within which we find ourselves, and understand how we got here. Those who forget history are doomed to repeat it.
All we have to do is look back in time at other parabolic moves to see how events in the past have led to factions of equities becoming overly valued. You’ll notice the magnificent seven in red on this graphic, and it would appear that we are nearing the inflection point where we have seen many other parabolic moves rollover and retrace gain. Clearly, I can’t predict when that will happen, but I believe it remains in front of us, so investors in my opinion should be aware of the premium they are paying as this AI thematic parabolic move continues to age.
At a time when economic data suggests that manufacturing capacity should increase, to meet increasingly greater demands for goods and services, supported by increasingly lower inflation, and capacity to consume widening, I have to ask, when does the market recognize that a shift away from being leaner, to being bigger, may be unfolding? Should that realization occur, then how does that translate into what seems to be an overly valued, narrow sleeve of stocks that were driven to these levels by an inflation crisis that seems to be resolving? Again, I can’t give you a date on a calendar, but it hasn’t happened yet, so it remains in front of us.
In the early morning hours on Monday, the 27th of January, a rumor launched from China while we slept. As we awoke and consumed our news that morning, we heard about some new technology called Deep Seek, an open source AI model that was purported to be more efficient and cost effective than Chat GPT.Â
Well, I suppose we’ll find out in time if that claim is true or not, but that’s really not the point. The point is that it only took a rumor, for the AI thematic segment of the market to capitulate violently
Nvidia set the record on that day for the largest one-day loss in the history of the US stock market. Indeed, that wasn’t the first time Nvidia had claimed that title. Nvidia actually owned eight of the top ten, worst one day losses in the history of the stock market.
So, this shows you how nervous this overly valued market segment might be, and with only a rumor, valuations can change very quickly. I don’t know if this will be a persistent theme throughout the year, but again, that’s really not the point.
I’ll never tell you anything that I can’t put on a spreadsheet. When I say that this sleeve of ten stocks is overly valued, I think this chart demonstrates what I mean pretty well. Pick your geometric mean regression. We have never seen a basket of stocks trade at more than three standard deviations away from an historical market mean that dates back to 1900. Again, for context, you can clearly discern the Y2K period of time.
From the perspective of valuation, there’s a lot of industry chatter about how the market is overly valued in terms of trailing and forward earnings multiples.Â
Well, again, you have to keep in mind that 40% of the market they describe is comprised of only ten stocks, and again in my opinion, they seem to be overly valued.
In this chart, if you look at the cap weighted market itself, it seems to be trading at 24.7 times earnings. However, if you looked at the market from an evenly weighted perspective, where Nvidia carries the same weight as Kellanova for example, you get a far different picture. An equally weighted market is seemingly trading at 18.3 times earnings. That’s a 26% difference.Â
Now, I don’t have a chart to demonstrate this metric, but when I carve these ten stocks out of the index, by my calculations, it seems to be trading at around 14.8 times earnings, and historically, that’s healthy. In other words, in my opinion, the vast majority of what we call the market isn’t overly valued at all as a function of earnings multiples. These ten stocks are heavily skewing that metric.
Strategic Approaches for Investors
How might an investor incorporate AI thematic areas into their portfolios then? Well, again let me offer this disclaimer by saying this is not to be construed as investment advice to anyone reading this article. As always, investing strategies should be carefully considered across many factors that are unique to you, including tolerance for risk.
I am intrigued by the rapid emergence of data centers across the world, and in the United States in particular.
Ok, well it occurs to me that these data centers must be kept at a constant temperature, so HVAC solutions are important. They must have rigid security protocols in place. Building automation is a central element as well.Â
You can’t brown out a data center, so having a source of uninterruptable power is critical. Power distribution, server cabinets, and power distribution are all critical elements of data center construction.
They also eat power. In October, it was announced that Constellation Energy would restart Unit One at Three Mile Island, and sell the power to Microsoft. We neither have the capacity to generate the amount of power that data centers will require, nor the grid to deliver it, so, in the future, as data centers emerge, I believe that power generators will be seen as growth stocks, and not the value stocks as we’ve historically known them.
The point is, that there are ways to incorporate this AI thematic trend into portfolios without direct ownership in stocks of companies that may have melted up to frothy levels. You just have to think outside of the box and do your research.
In closing, let me again remind everyone that while a topical discussion like this is useful, it should be just another facet of a product agnostic approach to holistic financial planning. Where a portfolio is allocated at any given time should entirely be a function of long-term goals that you’re trying to accomplish, and reflect your investor DNA in terms of risk tolerance for example. Don’t just jump into a market area because you think there’s a quick dollar to be made. Every decision should be made in the context of, how does this fit into my long-term planning goals.
(*) David R. Guttery, RFC, RFS, CAM, is a financial advisor, and has been in practice for 33 years, and is the Chief Executive Officer of Keystone Financial Group in Trussville, Alabama, and inSOURCE Financial Advisors in Lincoln, Nebraska. David offers products and services using the following business names: Keystone Financial Group and inSOURCE Financial Advisors – insurance and financial services | Ameritas Investment Company, LLC (AIC), Member FINRA / SIPC – securities and investments | Ameritas Advisory Services – investment advisory services. AIC and AAS are not affiliated with Keystone Financial Group or inSOURCE Financial Advisors. Information provided is gathered from sources believed to be reliable; however, we cannot guarantee their accuracy. This information should not be interpreted as a recommendation to buy or sell any security. Past performance is not an indicator of future results. Examples are for illustrative purposes only and should not be considered representative of any investment. Investments involve risks, including loss of principal.
Recent Articles
