The 2025 S&P 500 rally has a secret the headline numbers are hiding
The S&P 500 rally in 2025 looks impressive on paper, but there is an important detail the headline numbers do not make so obvious.
Since the end of February, the index has accumulated a gain of about 10%. Sounds solid, right? But when Goldman Sachs strips out stocks tied to Artificial Intelligence from the calculation, the picture changes dramatically — the same index turns slightly negative over the same period. Meanwhile, stocks considered AI winners have soared more than 45% during that same stretch. In other words, what is propping up this rally is not a broad market recovery. It is basically AI carrying the entire index on its back 🤖.
And that raises a question investors are starting to ask more frequently: what happens to the S&P 500 if Artificial Intelligence stocks lose steam? The answer may be less comfortable than the current numbers suggest.
The concentration nobody wants to see
This phenomenon has a technical name: market concentration. And it is at a level few analysts can remember seeing with this much intensity. When a handful of companies — in this case, those directly tied to the rise of Artificial Intelligence — end up accounting for a disproportionate share of the performance of an index as broad as the S&P 500, what looks like diversification on the surface hides a very specific dependency underneath. This is not just a cosmetic or narrative issue: it is a structural problem that affects how risk is distributed within the index.
Historically, the S&P 500 was built as a barometer of the American economy as a whole. The idea was always that by bringing together 500 large companies across a variety of sectors, the index would reflect a more accurate picture of overall market health. But what we are seeing now is different. Tech stocks with direct exposure to Artificial Intelligence — especially the so-called Magnificent Seven — carry such an enormous market weight that, on their own, they can move the needle for the entire index. When they rise, the S&P 500 rises. When they stumble, the index feels it immediately.
The Magnificent Seven group, for those who are not familiar, includes Apple, Alphabet, Microsoft, Amazon, Meta, Tesla, and Nvidia. The label gained traction in 2023, and since then these seven tech giants have been responsible for an ever-growing slice of the total index performance. What changed in 2025 is that the story is no longer just about these seven companies in isolation — it now encompasses an entire ecosystem: AI infrastructure, AI software, energy utilities that power data centers, industrial suppliers, and every company investors see as a direct beneficiary of the AI investment boom.
Receive the best innovation content in your email.
All the news, tips, trends, and resources you're looking for, delivered to your inbox.
By subscribing to the newsletter, you agree to receive communications from Método Viral. We are committed to always protecting and respecting your privacy.
This does not mean the market is wrong to price these companies with so much enthusiasm. Real revenue growth, expanding margins, and surging corporate demand for AI solutions justify a good portion of the optimism. The problem is not the value itself — it is the dependency. An index that needs a specific group of stocks to keep its rally going is, in practice, betting on a single thesis. And theses, no matter how solid they seem, always carry alternative scenarios the market would rather not price in.
The historical comparison that is turning heads
Jim Bianco, from Bianco Research, took this discussion to another level by stating that the market has not been this concentrated around a single theme in 150 years. The comparison he draws is with the railroad boom of the late 19th century — a period when a single transformative technology dominated capital markets because it was literally reshaping the United States economy.
According to Bianco, railroads transformed the country in a way no other technology had the potential to match — until the arrival of Artificial Intelligence. It is a bold comparison, but one backed by data. Bianco Research used a list of 41 AI-related stocks compiled by JPMorgan and showed that these names already account for nearly half of the total market capitalization of the S&P 500. Think about that number for a second: 41 companies, out of a universe of 500, concentrating nearly 50% of the entire value of the index.
This concentration does not make the risk harmless, but it makes the timing of any investment decision much more difficult. Bianco acknowledges the situation likely constitutes a bubble, but argues that the more important question is not whether we are in a bubble — it is where in the cycle of that bubble investors currently find themselves.
The 1990s lesson that still holds up
To illustrate this point, Bianco looks back to the late 1990s. When Alan Greenspan, then chairman of the Federal Reserve, delivered his famous irrational exuberance speech in December 1996, many investors took it as a signal to exit the market. Those who followed that reading missed a nearly 300% run-up in the Nasdaq before the dot-com bubble burst.
The message here is not that bubbles are good or that investors should ignore risk signals. The message is that bubbles can last far longer than conventional wisdom suggests, and getting out too early can be just as costly as getting out too late. For anyone following the market in 2025, this is the kind of nuance that separates a surface-level take from a genuinely useful read of the landscape.
What the Goldman Sachs data reveals
The Goldman Sachs analysis that brought this debate into the spotlight is not just a theoretical exercise. It shows, in pretty straightforward terms, that the performance of the S&P 500 since the end of February 2025 is fundamentally anchored by a group of stocks with exposure to Artificial Intelligence. When that group is removed from the equation, the index stops being a recovery story and becomes, in reality, a story of stagnation — or even slight decline. This is exactly the type of data portfolio managers use to calibrate the real risk level of a portfolio that, on paper, looks nicely diversified.
Beyond the numbers, this kind of analysis points to something deeper: the narrative that is holding up prices. The market is buying into the AI story with a lot of conviction, and that in itself is not a problem. But when conviction is concentrated so intensely around a single theme, any event that challenges that narrative — whether it is more aggressive regulation, a slowdown in earnings growth at these companies, or even a shift in institutional investor sentiment — can have an amplified effect on the index as a whole. The rally that looked robust can show its cracks quickly.
It is worth noting that Goldman Sachs is not the only bank paying close attention to this phenomenon. JPMorgan, Morgan Stanley, UBS, and other major institutional players have been publishing internal reports flagging the level of concentration in the S&P 500 and the risks tied to a reversal in sentiment around Artificial Intelligence. This is not isolated alarmism — it is a conversation happening on the top floors of global finance, and it is gradually reaching retail investors as well.
AI as a market engine: how far can this go?
The question showing up most frequently on analyst desks right now is simple: has Artificial Intelligence already delivered enough in terms of concrete results to justify the weight its stocks carry in the market? This is a genuine debate with no easy answer. On one hand, companies like Nvidia, Microsoft, and Meta have been posting real revenue growth directly tied to AI adoption — this is not just future expectation, there is actual money coming in. On the other hand, a significant portion of current pricing is still discounting a future that has not arrived yet, and that kind of bet tends to get volatile when results take longer than expected to materialize.
There is also the macroeconomic factor. The interest rate environment, inflation, and global growth still have a lot of open variables for 2025, and any macro turbulence tends to hit assets with elevated valuations the hardest — exactly the profile of the big AI stocks. If the Federal Reserve changes its tone on rate cuts, for example, or if U.S. employment data starts showing weakness, the market could quickly reallocate capital into more defensive sectors, and the S&P 500 would lose the very engine that has been keeping its rally alive.
Even so, completely dismissing the potential of Artificial Intelligence as a transformative force would be a mistake. What the most experienced analysts are signaling is not that AI is going to disappear or that companies in the sector are overvalued without any basis. The point is more subtle: a healthy market over the long term needs a broader foundation of growth. When the rally depends on such a small number of names to sustain itself, it becomes fragile by nature — not because the theme is bad, but because concentration itself is already a risk factor independent of the quality of the assets involved.
The AI ecosystem moving the market
An important distinction that market coverage in 2025 has been making is that AI leadership in the S&P 500 is no longer limited to the traditional big tech names. The ecosystem has grown. It now includes:
- AI Infrastructure: chipmakers, server manufacturers, and networking equipment companies that form the backbone of data centers
- AI Software: companies developing platforms, language models, and automation tools powered by artificial intelligence
- Energy Utilities: power companies supplying the increasingly energy-hungry data centers
- Industrial Suppliers: cooling, construction, and logistics companies supporting the physical expansion of AI infrastructure
This thematic expansion is, on one hand, a sign that AI investment is maturing. It is no longer just Nvidia selling GPUs — there is an entire value chain being priced by the market. On the other hand, it also means the Artificial Intelligence theme permeates even more sectors of the index than it appears at first glance, which reinforces the thematic concentration thesis even when concentration by individual name may look slightly more diluted.
What is on the radar for the coming months
The next rounds of corporate earnings are going to be critical for understanding whether Artificial Intelligence stocks can maintain the pace the market is pricing in. Any sign of a slowdown in AI infrastructure spending — whether from big tech or from the companies consuming these solutions — will be read as a warning signal, and the effect on the S&P 500 could be quite visible, precisely because of the concentration the Goldman Sachs analysis made so evident.
Another point worth watching is the movement of institutional investors. Pension funds, sovereign wealth funds, and large asset managers have concentration limits in their mandates, and some are already approaching those limits when it comes to tech stocks with AI exposure. When that rebalancing starts happening on a broader scale, it can create selling pressure on the very assets that have risen the most — an effect retail investors rarely anticipate with precision.
The message from the S&P 500 right now is relatively straightforward: as long as AI leadership holds firm, the rally can keep looking strong in the headline numbers. But if stocks tied to artificial intelligence start losing momentum, the index could reveal that it has far less support underneath than the overall level suggests.
At the end of the day, what the market is testing in 2025 is whether Artificial Intelligence can be not just a powerful narrative, but a sustainable enough economic engine to justify the central role it plays in the S&P 500 rally. The coming months will help answer that question — and the answer is going to matter well beyond the borders of big tech 📊.
