The AI trade used to be simple: buy the enablers, ignore the noise, and let the CapEx tsunami do the rest.
Now? It’s turning into a food fight.
Here’s how the character of the AI market has shifted and why the vibe is a lot less euphoric than it was a year ago.
Scarcity > Scale
Instead of rewarding the companies building AI infrastructure, the market is obsessing over the ones sitting on supply bottlenecks.
Memory shortages have become the new main character. That has been a windfall for DRAM and HBM players like Micron $MU ( ▼ 0.67% ) and SK hynix, while semiconductor equipment names upstream are quietly benefiting too.
The signal is clear: near-term scarcity is getting valued more than long-term AI vision. Investors want pricing power today, not promises about platform dominance five years from now.
CapEx Fatigue Is Real
Last year, hyperscalers could announce eye-watering AI budgets and get standing ovations.
Now the market wants receipts.
Even when companies like Meta $META ( ▲ 1.46% ) show accelerating revenue growth, their stocks struggle to hold gains if CapEx guidance keeps climbing. Investors are no longer assuming that every dollar poured into data centers will automatically earn a great return.
“Spend big” used to be bullish. In 2026, it’s a debate.
The Squeezed Middle
There’s another group feeling the pain: the companies that use AI hardware but don’t control supply.
Chip buyers outside the hyperscaler elite are getting pushed to the back of the line. That has weighed on firms like Qualcomm $QCOM ( ▼ 2.11% ) , where tight component availability has been a factor in softer outlooks.
In other words, AI demand is booming — but not everyone gets equal access to the tools.
Software Is the New Punching Bag
Public markets have decided that “AI disruption” is less an opportunity and more a threat — at least for now.
Much of the software sector is getting hit regardless of earnings quality. Strong quarters and solid guidance are being shrugged off as investors worry that AI-native tools could erode pricing power, margins, or entire business models down the line.
It’s a rare setup where good fundamentals are losing to long-term existential fear.
The Winners Aren’t Public Yet
Adding to the weirdness: many of the companies best positioned to benefit from AI disruption still aren’t listed.
Heavyweights like OpenAI, Anthropic, and Musk-linked ventures tied to SpaceX and xAI are widely expected to tap public markets in the future.
Big funds know that when those IPOs hit, they will need room in portfolios. That expectation alone can make today’s AI trade feel “negative sum” — capital rotates out of current winners in anticipation of future ones.
Even the Kings Are Stalling
Perhaps the biggest mood shift: the two poster children of the AI chip boom have gone sideways.
Nvidia $NVDA ( ▲ 3.08% ) and Broadcom $AVGO ( ▼ 1.74% ) have both lagged for months, even as the broader market has eked out gains. When leaders stop leading, enthusiasm tends to cool fast.
The bottom line: AI is still the defining tech theme of the decade. But in 2026, the market isn’t pricing it like a clean, rising tide anymore.
It’s pricing bottlenecks, trade-offs, and second-order effects — and that makes the ride a lot bumpier.