Six years ago, chip stocks barely registered in most passive portfolios.
Today, the sector is one of the most dominant forces reshaping the S&P 500.
Semiconductor stocks now account for a record 19.7% of the index—almost four times their weighting in 2020.
In other words, chip stocks now own nearly a fifth of the S&P 500.
That kind of concentration changes how investors need to view the entire market.
Let’s break down the specific numbers… the under-the-radar risks that the market is ignoring… and what it means for your portfolio.
AI spending is flowing straight into chip earnings
Hyperscalers like Microsoft, Amazon, and Google have committed roughly $750 billion in capital expenditures for 2026 to build out AI infrastructure.
The companies making the chips that power that infrastructure—like Nvidia (NVDA), Broadcom (AVGO), and Advanced Micro Devices (AMD)—have seen their earnings (and their stock prices) follow that money higher.
Nvidia alone has a market cap of around $4.8 trillion, and Broadcom’s market cap has climbed to roughly $1.8 trillion.
In fact, Broadcom and Micron Technology (MU) have both joined the S&P 500’s top 10 by market cap as of mid-2026, a shift that would have seemed improbable even two years ago.
AMD is up over 160% year-to-date, and Intel (INTC) has surged over 270%.
But the real shift is happening inside the S&P 500 itself.
ETFs attracted more than $1 trillion in inflows year-to-date through late June 2026, about 45% above the record pace from the same period the prior year.
A lot of that money flows into passive index funds, which means every dollar going into an S&P 500 fund is, by definition, buying more chip stocks than it ever has before.
Simply put, the inflows keep coming, and the weighting keeps climbing.
The spending boom comes with an ROI question
One of the biggest questions is whether the spending justifies how high chip stocks have climbed.
Goldman Sachs has flagged that the AI capex boom has lifted semiconductor profit margins sharply, but rising depreciation expenses from hyperscalers will start to reverse some of that profitability over the next few years.
Let’s break down what this means: When Microsoft, Amazon, Google, Meta, and other hyperscalers spend hundreds of billions of dollars on AI infrastructure, that money shows up as revenue for chipmakers.
But for the hyperscalers, those chips and data centers become assets on the balance sheet, and the cost gets recognized gradually over time through depreciation.
Depreciation can become a drag on future profits if the AI infrastructure fails to generate enough revenue, productivity gains, or cost savings to offset it.
In plain English: The chip companies are getting paid now, but the buyers still have to prove the spending earns an attractive return.
And if that return starts to look weaker than expected, the pressure to slow future chip orders grows.
J.P. Morgan’s asset management team noted that 2026 earnings-per-share estimates for U.S. semiconductor names have been revised up 22% since January, which tells you the market still believes the growth story is intact…
But elevated expectations cut both ways: even a solid quarter can disappoint if it doesn’t beat by enough.
The S&P 500 is more concentrated than investors realize
The concentration is another problem.
When nearly 20% of the S&P 500 sits in one sector, the index stops being as diversified as most investors assume. A passive investor who thinks they own “the whole market” actually owns a portfolio in which one out of every five dollars is riding on the semiconductor cycle.
That’s a meaningful exposure that many people don’t realize they have.
In other words, the index funds that hundreds of millions of ordinary investors rely on for broad diversification are now, by design, heavily tilted toward a single technology cycle.
There are two ways to think about what happens from here.
The bull case is straightforward: Data center demand is real. The companies buying chips are the most profitable businesses on the planet, and they’ve explicitly told investors they plan to keep spending.
Under that scenario, the 19.7% weighting is less a sign of excess and more a reflection of where the economy is shifting.
The bear case is more uncomfortable: Sector concentration at this level has historically marked late-cycle behavior.
In other words, when one area of the market becomes this dominant, it usually means the easy money has already been made, and any wobble in the underlying spending story gets amplified by the sheer size of the exposure.
What it means for your portfolio
The smartest approach starts with understanding what you actually own.
If you hold an S&P 500 index fund, you’re already deeply exposed to the semiconductor cycle. That’s not necessarily a problem: The AI buildout is real… But so is the concentration.
Meanwhile, it remains to be seen whether hyperscaler spending ultimately earns back the ROI to justify today’s valuations.
The market is betting it will.
And if you own the S&P 500, you’re making the same bet, whether you realize it or not.
Want more insight into what’s moving the market… and how to position your portfolio?
Check out our new all-in-one investing hub, Curzio Alpha!
Members get access to all our research and stock recommendations across sectors and market trends… in a single portfolio.
