Fear moves in herds, and it doesn't discriminate between the strong and the weak.
Fear, when it moves through a market, does not pause to read balance sheets. This week's sweeping sell-off in AI chip stocks has pulled down the exceptional alongside the struggling, creating the kind of gap between price and underlying reality that patient investors have long learned to watch for. Nvidia, ON Semiconductor, and Intel each carry different stories and different risks, but all three now trade at valuations shaped more by collective anxiety than by the particulars of their businesses. The question the moment poses is an old one: whether the price the market offers today reflects what these companies are actually worth, or merely what the crowd fears they might become.
- A wave of doubt about whether AI spending can ever justify its costs has sent the entire chip sector into an indiscriminate freefall, punishing strong and weak companies alike.
- ON Semiconductor suffered the sharpest single-session blow — a 23% collapse — after announcing a $7 billion all-stock acquisition of Synaptics that alarmed investors with its dilution and integration risk.
- Nvidia, despite reporting 85% revenue growth and guiding toward $91 billion next quarter, has slid 18% from its peak and now trades at a valuation meaningfully below the premium it held through most of the AI boom.
- Intel's dramatic recovery from near-collapse has stalled at a complicated juncture: revenue is rising, Tesla has signed on as a foundry customer, but the foundry division still lost $2.4 billion in a single quarter.
- Analysts are drawing a clear hierarchy of risk: Nvidia looks like a discounted quality business, ON Semiconductor pairs real recovery with a speculative deal, and Intel remains a leap of faith into a turnaround still bleeding cash.
The chip sector is in freefall, and like most indiscriminate sell-offs, it is punishing the excellent alongside the mediocre. A wave of skepticism about whether AI spending will ever justify its enormous costs has sent the whole industry lower, and when fear moves through a sector this way, it rarely bothers to distinguish between a company with genuine problems and one that simply got caught in the current. That gap — between what the market is pricing and what the business is actually doing — is where opportunities sometimes hide.
Nvidia is the obvious place to start. The stock has fallen roughly 18 percent from its 52-week peak, yet the company behind it continues to perform at an extraordinary level. Revenue jumped 85 percent year over year to $81.6 billion in its most recent quarter, with data center revenue climbing 92 percent. Management guided the next quarter at $91 billion, suggesting the acceleration hasn't stalled. After the decline, the stock trades at roughly 29 times earnings — a significant discount from the premium multiples it commanded through most of the AI boom — with gross margins still hovering near 75 percent. The risk is the same one hanging over the entire sector: if doubts about AI spending prove justified, Nvidia's stock could fall much further. But of the three companies examined here, it most resembles a genuine opportunity.
ON Semiconductor didn't merely get swept up in the broader downdraft — it got demolished, plummeting more than 23 percent in a single session after announcing a $7 billion all-stock acquisition of Synaptics. The deal is meant to deepen its position in edge AI and physical AI, but the all-stock structure dilutes existing shareholders, and the market reacted with visible displeasure. Zoom out from the single-day panic, however, and the underlying business is genuinely turning a corner. After a prolonged slump, first-quarter revenue rose 5 percent — its first growth in several quarters — and its AI data center business more than doubled. CEO Hassane El-Khoury called it evidence the company had moved beyond its cyclical trough. The stock's elevated price-to-earnings ratio reflects depressed profits at the bottom of a cycle rather than overvaluation; if the recovery sustains, today's price could eventually look like a bargain.
Intel has been the sector's most dramatic story. The stock climbed from a 52-week low near $19 to above $141 as new CEO Lip-Bu Tan's restructuring began to show results, and this week's sell-off barely touched it. Revenue rose 7 percent to $13.6 billion, and the company reportedly won Tesla as its first major customer for its most advanced manufacturing process — a meaningful validation for its struggling foundry operation. Yet the picture remains complicated: Intel is profitable only on an adjusted basis, and its foundry division lost $2.4 billion in the quarter even as revenue climbed. After a run this substantial, the stock already prices in a turnaround that is only partially complete, asking investors to make the biggest leap of faith of the three.
The chip sector is in freefall this week, and like most indiscriminate selloffs, it's punishing the excellent alongside the mediocre. A wave of skepticism about whether artificial intelligence spending will ever justify its enormous costs has sent the whole industry lower, with Friday bringing fresh waves of selling. When fear moves through a sector this way, it rarely bothers to distinguish between a company with genuine problems and one that simply got caught in the current. That gap—between what the market is pricing and what the business is actually doing—is where opportunities sometimes hide.
Nvidia is the obvious place to start. The stock has fallen roughly 18 percent from its 52-week peak, sliding back toward $190, yet the company behind it continues to perform at a level most would call extraordinary. In the quarter that ended April 26, 2026, revenue jumped 85 percent year over year to $81.6 billion. More tellingly, data center revenue—the engine of its AI business—climbed 92 percent to $75.2 billion. Management guided for the next quarter at $91 billion in revenue, suggesting the acceleration hasn't stalled. What makes the current pullback worth examining is the valuation math. After this decline, the stock trades at roughly 29 times earnings, a significant discount from the premium multiples it commanded through most of the AI boom. For a company still growing at this pace, with gross margins hovering near 75 percent, that's not an expensive price to pay. The risk, of course, is the same one hanging over the entire sector: if the market's doubts about AI spending prove justified and growth normalizes, Nvidia's stock could fall much further.
ON Semiconductor didn't merely get swept up in the broader downdraft—it got demolished. The stock plummeted more than 23 percent on Friday alone, landing around $91, after the company announced a $7 billion all-stock acquisition of Synaptics, its largest deal ever. The move is meant to deepen its position in edge AI and so-called physical AI, the embedding of intelligence into everyday objects. Because the deal is structured as an all-stock transaction, it dilutes existing shareholders, and the market reacted with visible displeasure at both the price tag and the integration risks involved. But zoom out from the single-day panic, and the underlying business is genuinely turning a corner. After a prolonged slump in demand from automakers and industrial customers, the company's first quarter revenue, for the period ended April 3, 2026, rose 5 percent year over year—its first growth in several quarters. More impressively, its AI data center business more than doubled. CEO Hassane El-Khoury described the quarter as evidence the company had "moved beyond the cyclical trough on a path to recovery." The stock's price-to-earnings ratio sits in the 60s, a figure that reflects how depressed profits have become at the bottom of the cycle rather than a sign of overvaluation. If the recovery sustains, today's price could eventually look like a bargain.
Intel has been the sector's most dramatic turnaround. The stock has climbed from a 52-week low near $19 to a high above $141 over the past year as new CEO Lip-Bu Tan's restructuring has begun to show results. This week's selloff barely touched it, with shares easing only to around $128. The business itself is improving: first-quarter revenue rose 7 percent to $13.6 billion, adjusted earnings per share came in at $0.29, and the company reportedly won Tesla as its first major customer for its most advanced 14A manufacturing process—a significant validation for its struggling foundry operation. Yet the picture remains complicated. Intel remains profitable only on an adjusted basis, and its foundry division lost $2.4 billion in the quarter even as revenue climbed. After a run this substantial, the stock already prices in a turnaround that is only partially complete. Of the three, Intel asks investors to make the biggest leap of faith: betting that Tan can sustain momentum in a business that is still losing serious money where it matters most.
None of these three stocks is a low-risk proposition. The selloff that has dragged them down could deepen if the market's concerns about AI spending intensify. But the risks are not equal. Intel requires faith in an unproven turnaround still bleeding cash. ON Semiconductor pairs a genuine cyclical recovery with a massive, dilutive acquisition whose value remains unproven. Nvidia, by contrast, is a fast-growing, dominant business whose shares have fallen below what the underlying fundamentals suggest they're worth. It carries the same AI cycle risk as the others, but of these three, it's the one that looks most like a genuine opportunity rather than a speculative bet.
Notable Quotes
We exceeded expectations as demand strengthened through the quarter and we have moved beyond the cyclical trough on a path to recovery.— Hassane El-Khoury, CEO of ON Semiconductor
The Hearth Conversation Another angle on the story
Why does a sector-wide selloff create opportunities at all? Shouldn't the market be efficient enough to separate good companies from bad ones?
In theory, yes. In practice, fear moves in herds. When doubt about an entire industry spreads, it doesn't discriminate—it just sells. A company growing at 85 percent with 75 percent margins gets hit the same as one struggling to find customers. The gap between what the market is pricing and what the business is actually doing is where you find opportunities, if you can stomach the volatility.
So you're saying Nvidia's 18 percent drop is just noise, not a real warning signal?
Not exactly. The risk is real—if AI spending actually does cool, Nvidia's growth could collapse. But the company is showing no signs of that happening. Revenue is accelerating, not decelerating. The question is whether you believe the market's doubt or the company's numbers.
ON Semiconductor got hit much harder. Is that because the Synaptics deal is actually a bad idea, or just because the market hates dilution?
Both, probably. All-stock deals dilute shareholders, and integration risk is real. But the underlying business is genuinely recovering for the first time in quarters. The market is punishing the deal announcement without giving the recovery enough credit.
Intel seems like the wildcard here. It's still losing billions in its foundry business.
It is. Intel is asking you to believe in a turnaround that's only halfway done and still bleeding cash where it matters most. That's a bigger bet than the other two. With Nvidia, you're buying a proven business at a discount. With Intel, you're betting the CEO can finish what he started.
If you had to pick one, which would you actually buy?
Nvidia. It's still the highest-risk of the three because it's tied to the same AI cycle everyone's doubting. But it's the only one where the business fundamentals clearly exceed what the stock price reflects.