NextFin News - Qualcomm and Arm both walked into earnings season with AI-heavy growth stories, but their forecasts failed to give investors the clean upside they wanted. Qualcomm guided fiscal third-quarter revenue to $9.2 billion-$10.0 billion, below a consensus that was closer to $10.2 billion, while Arm said it expected fiscal first-quarter revenue of $1.0 billion-$1.1 billion, a range that looked adequate on the surface but not convincing enough to re-rate the stock higher. The reaction was not just about one quarter. It was about whether the semiconductor market is still rewarding narrative strength when the next guide is only modestly ahead of, or even short of, what the market already expected.
Qualcomm’s own numbers showed why the bar is tricky. In fiscal second quarter 2026, the company reported revenue of $10.6 billion, non-GAAP earnings per share of $2.65, and QCT automotive revenue of $1.326 billion, up 38% from a year earlier. Handset revenue fell 13% to $6.024 billion, IoT revenue rose 9% to $1.726 billion, and the company returned $3.7 billion to stockholders through dividends and repurchases in the quarter. The problem for the stock was not that Qualcomm lacked growth points. It was that the market is now asking whether automotive, IoT, data center ambitions, and on-device AI can offset a smartphone business still tied to memory constraints and a cautious handset cycle quickly enough to justify a higher multiple.
Arm’s issue was slightly different but rhymed with Qualcomm’s. The company has been one of the market’s favorite AI infrastructure names because its licensing model gives it leverage across phones, servers, PCs, and edge devices without needing to sell chips itself. That makes each forecast a test of how much of the AI thesis is still prospective and how much is already embedded in the share price. A revenue guide of $1.0 billion-$1.1 billion is not weak in absolute terms, but it can still disappoint when investors are looking for a sharper acceleration, especially after a long run of optimism around Arm’s royalty stream and AI-adjacent exposure. In other words, the question is no longer whether Arm belongs in the AI conversation. The question is whether the conversation is already fully priced.
That is why the market response was so telling. Both companies are now being judged less by whether they can tell a credible growth story and more by whether the next quarter justifies the premium already built into the names. For Qualcomm, the guide implied that handset weakness and memory constraints could keep the near-term mix softer than bulls hoped. For Arm, it suggested that licensing and royalty momentum still need more proof before investors pay for a faster growth regime. The result was a classic expectation gap: strong strategic narratives, but guidance that did not stretch far enough beyond consensus.
The deeper issue is that the semiconductor equity market is no longer rewarding every AI connection equally. Investors have become more selective about transmission. They want to know whether AI improves revenue now, whether it only improves the product roadmap, or whether it simply makes the story easier to sell. Qualcomm’s guide pointed to the first category only in part: automotive and IoT are growing, but smartphones still dominate the earnings bridge. Arm sits in the second and third categories more than the first. It benefits when the industry shifts toward more compute at the edge, but its revenue recognition lags the excitement. That delay matters when rates are still high enough to punish distant cash flows and when every multiple already assumes a cleaner growth path than the company just described.
Why Qualcomm’s Guide Was Not Enough
Qualcomm’s forecast disappointed because it did not solve the market’s core problem: the company’s strongest growth areas are improving, but the legacy business still sets the pace. In fiscal second quarter 2026, automotive revenue rose 38% year over year to $1.326 billion and IoT revenue gained 9% to $1.726 billion, but handsets still generated $6.024 billion and declined 13%. That mix matters more than the headline revenue figure because investors do not buy Qualcomm for a single quarter. They buy it for a mix shift. If the mix shift looks incremental rather than transformative, the multiple can compress even when revenue holds up.
The company tried to frame the cycle as one with a visible bottom. Qualcomm said in its release that it now estimates handset revenue from Chinese customers will reach a bottom in the third quarter and return to sequential growth in the following quarter. That is important, but it is still a cyclical argument. Memory shortages, handset inventory management, and replacement timing can all improve later; they have done so before. The question is whether the market wants to pay today for a bottom that remains a quarter or two away. The answer, at least in the immediate reaction, was no.
That makes this a cyclical issue first and a structural issue only second. The cyclical leg is obvious: handset demand, inventory normalization, and memory pricing can all move through familiar boom-bust patterns. The evidence for that is historical. Qualcomm has lived through multiple handset slowdowns, from post-boom digestion periods to channel corrections and replacement-cycle pauses, each of which eventually gave way to renewed demand once inventories cleared and device refreshes resumed. The company’s current language about a Chinese handset bottom fits that pattern. But the structural leg is newer. Qualcomm’s diversification into automotive, industrial IoT, edge AI, and data center silicon is designed to reduce dependence on handsets over time. That shift is not a temporary rebound; it is an effort to change the earnings base.
The structural problem is that the market will not pay for the destination until it sees repeated proof of the journey. Qualcomm’s $1.326 billion in automotive revenue is real progress, but it is still far smaller than handset revenue. A 38% increase sounds like a breakaway; a $1.326 billion base says otherwise. The same is true for the company’s broader AI pitch. Management can say that AI agents are reshaping its roadmap, and that a hyperscaler custom silicon engagement is on track for initial shipments later this calendar year, but investors still need to see how much revenue falls out of that story, and when. Without that, the market treats AI as optionality, not as a replacement for handset cyclicality.
“We are in a period of profound industry transformation — the rise of AI agents is reshaping our roadmap across every platform we develop.”
Cristiano Amon’s line captures the strategy correctly. It also exposes the problem. A roadmap can change faster than a revenue base. Qualcomm is trying to make the jump from a handset-linked cycle to a broader edge-compute platform, but the earnings bridge still depends on the old engine. That is why the stock can weaken even after a solid quarter: the market is not doubting the story; it is doubting the timing.
The second-order effect is more important than the first-order disappointment. If investors conclude that Qualcomm’s diversification is real but slow, they stop valuing the company like a rerating candidate and start valuing it like a mature cyclical with some growth optionality attached. That changes the discount rate in practice, not the formal one. It means the stock’s multiple responds less to long-dated promises and more to whether each quarter beats the near-term consensus cleanly. In that regime, a guide that lands inside a reasonable range can still feel disappointing if it does not move the market farther than expected.
Why Arm’s Forecast Was Good Enough To Exist, But Not Good Enough To Excite
Arm’s forecast failed for a subtler reason. The company does not need to post the same kind of revenue bridge as Qualcomm, because its asset-light model and IP licensing structure create a different earnings pattern. But that very structure also means the stock trades on duration: how long it can keep compounding royalty streams as Arm-designed architectures spread through smartphones, servers, PCs, automotive systems, and edge devices. When the company’s forecast lands near expectations rather than above them, investors start asking whether the growth curve is bending enough to justify the multiple.
Arm’s current position makes that question unavoidable. The business has become a proxy for the spread of AI-capable devices, but it is still exposed to the timing of design wins and the pace at which those design wins turn into royalties. That is a slower transmission mechanism than Qualcomm’s. Qualcomm can point to units, ASPs, mix, and segment revenue in the near term. Arm must wait for licenses to convert into deployed silicon and for deployments to convert into recurring royalties. That lag is why a forecast that is technically within range can still read as underwhelming.
The market reaction fits a broader pattern in semiconductors: investors are rewarding companies that can show near-term monetization of AI, and becoming less forgiving of names whose AI exposure is mainly strategic. Arm remains central to the AI hardware ecosystem, but centrality is not the same as immediate revenue acceleration. If the forecast does not imply stronger near-term royalty growth, investors can admire the franchise and still rotate away from the shares.
That is where the structural-versus-cyclical call matters most. Arm is not dealing with a normal end-market lull that will quickly mean-revert. Its problem is partly cyclical because customer adoption can pause, but the more important issue is structural: expectations have been repriced around a world in which every additional edge device, server, or custom chip broadens Arm’s royalty opportunity. Once a company is valued on the speed of a structural adoption curve, even a decent guide can disappoint if it does not confirm acceleration. The cycle may explain the quarter. The structure explains the reaction.
The counter-thesis is straightforward and serious: both stocks may have sold off not because the outlooks were bad, but because the bar was absurdly high. On that view, Qualcomm’s guide was conservative but still healthy, its automotive and IoT momentum remains intact, and Chinese handset bottoming points to sequential improvement ahead. Arm’s forecast, meanwhile, keeps the AI licensing story alive and supports the idea that every new generation of compute still leans on Arm architecture. That thesis is not trivial. It is probably the mainstream bull case. It says the market is overreacting to a one-quarter guide rather than repricing the business models.
The answer is that the bulls are right on the longer arc and still wrong on the immediate tape. A premium multiple is not a reward for having the right long-term strategy; it is a claim on the pace at which that strategy becomes measurable. Qualcomm and Arm both failed that test. Neither company told investors that the next quarter would be dramatically better than the market already assumed. That is enough to cap enthusiasm in a market that has become very sensitive to the difference between a good story and a faster earnings conversion rate.
The falsifying signal for this bearish near-term read is concrete. If Qualcomm’s handset revenue from Chinese customers does return to sequential growth in the following quarter and automotive revenue keeps compounding at a rate near or above 30% year over year, the stock should stop being judged as a pure handset-cycle name and begin trading more like a multi-engine platform. For Arm, a sustained upside surprise in revenue above the current guide band, paired with a visible step-up in royalty intensity from AI-centric deployments, would show that the current disappointment was only a timing issue. Until then, the market is likely to keep demanding proof rather than promises.
What The Market Is Really Pricing Now
The key second-order point is that the market is not simply asking whether Qualcomm and Arm can grow. It is asking whether growth is coming fast enough to outrun the compression in patience. That is a different question. It explains why solid numbers can still disappoint. It also explains why a guide that looks acceptable on paper can feel weak in a sector where investors have spent the last two years paying up for every credible AI exposure. The market is narrowing the gap between story and delivery.
That narrowing has cross-asset implications. For Qualcomm, softer near-term guidance can keep pressure on the stock even if the company’s longer-term diversification strategy remains intact, because the immediate debate is about mix and timing. For Arm, the effect is more valuation-oriented: a guide that does not accelerate the royalty curve makes it harder to defend a premium tied to the spread of AI architecture rather than to current earnings power. In both cases, investors are moving from narrative valuation to evidence valuation. That is not a permanent verdict, but it is a meaningful regime shift in how these names are being judged.
Short term, the stocks are hostage to sentiment and the next print. Medium term, the question is whether Qualcomm can turn automotive, IoT, and data center efforts into a larger share of revenue while Arm proves that AI adoption is lifting royalties more quickly than the market modeled. Long term, the structural story still matters: Qualcomm wants to be less dependent on handsets, and Arm wants to be the architecture behind the next generation of compute. But neither company gets the long-term premium for free. They have to show the bridge quarter by quarter.
Base case: the selloff and investor hesitation persist until the next set of guidance updates confirms that sequential improvement is real. Upside case: Qualcomm’s handset trough arrives on schedule, automotive growth stays near double digits, and Arm shows that AI-linked deployment is turning into sharper revenue acceleration. Downside case: handset weakness lasts longer than expected, memory constraints keep pressuring purchasing patterns, and Arm’s royalty curve proves less elastic than bulls assumed. The trigger to watch is simple. If the next quarter does not translate strategy into harder numbers, the market will keep treating both companies as expensive stories with uneven execution.
This is not the market rejecting AI. It is the market refusing to pay endlessly for AI that still lives in the future.
Explore more exclusive insights at nextfin.ai.
