By Maksym Misichenko · CNBC ·
What AI agents think about this news
Despite impressive earnings and AI capex growth, panelists express caution due to potential capex digestion risk, cyclicality, customer concentration, and margin squeeze from competition and component costs.
Risk: AI demand peaking early or a demand-supply mismatch leading to a margin squeeze
Opportunity: Potential shift in mix toward higher-margin enterprise deals less exposed to hyperscaler ASIC displacement
This analysis is generated by the StockScreener pipeline — four leading LLMs (Claude, GPT, Gemini, Grok) receive identical prompts with built-in anti-hallucination guards. Read methodology →
Don't fight the tape. That's the message from strategists at Piper Sandler on Friday morning, and you know what, it's hard to argue with their take. "The market's price action continues to remind us of a simple mantra from the movie Maverick— 'Don't Think, Just Do,'" they wrote. "The primary trends are firmly bullish, rewarding investors who avoid over-analyzing the macro noise and instead follow the tape's leadership. By sticking with momentum leaders and avoiding laggards, investors are capturing the market's grind into new highs." Importantly, in many cases, the winners are winning thanks to earnings, not simply multiple expansion. Take a look at server maker Dell Technologies , the latest company to report a blowout quarter on the back of the artificial intelligence spending boom. If you're only looking at the price action — shares up over 30% on Friday — you could be forgiven for thinking that we're starting to see bubble behavior. However, as was the case with Nvidia from the start of its historic run following the launch of ChatGPT in late 2022 , it's not a bubble if the fundamentals back it. That is certainly the case with Dell. Wall Street has already raised its fiscal 2027 full-year earnings estimate to $16.85 a share, up from $13.12 prior to Thursday night's report, according to FactSet data. Meanwhile, estimates for next fiscal year have jumped to $20.21 a share from $15.18. That represents a 28.4% increase to FY27 numbers and a 33% increase to FY28 numbers. What that means is that, despite Friday's surge in the stock, based on a share price of $410, the price-to-earnings ratio on FY27 numbers has only expanded from 24.2 times to 24.3 times. On FY28 numbers, the stock trades at roughly 20.3 times earnings, which is actually down from the 21 times multiple we thought we were looking at Thursday before learning that the estimates have just been way too low. Put simply, the price action will have you thinking this is unsustainable. That's exactly what folks were saying about Nvidia after that legendary May 2023 earnings report , when Wall Street realized just how offside it was on the earnings tidal wave about to hit. The point is, if you're looking at these moves and thinking you need to bet against them, think again. The AI spending boom is real. We are simply now seeing Wall Street catch up to what the hyperscale CEO's have been saying: the AI opportunity is far bigger than most can imagine, and the spending on data centers and computing infrastructure is required. On this note, Dell's customer count for AI servers has surpassed 5,000, up over 50% in the last six months. Talk about broadening demand, which is something Nvidia preached when it reported last week. Nvidia and Dell are close partners, and Nvidia is a clear winner from this quarter too. We expect that will become even clearer in coming days, when CEO Jense Huang gives his keynote address at Computex in Taipei. At the same time, we're not advocating chasing a move like this in Dell -- no matter how impressed we are by the numbers. The reason to proceed with caution has more to do with market mechanics and less about the fundamentals or valuation. A lot of people just made a lot of money, and when that happens, some profit-taking should be expected. Consider: There may be some fund managers with Dell positions that now exceed an appropriate weighting thanks to these monster gains, forcing them to peel some off. This gets to the idea of why we like to trim stocks after parabolic moves. In this case, we're saying you should wait for a name like Dell to cool off before initiating a position. That can certainly make this a hard market to get money to work in, but what we can do is keep track of winners and estimate revisions, and use that to build a target list — similar to our Bullpen watchlist — that we can revisit when volatility strikes (as it always does). As for the AI stocks we already own, this is simply another proof point that for all the money made in the AI trade — whether that's in semiconductors like Nvidia, power, or network infrastructure, there is still plenty of upside left to be had because demand is off the charts. (Jim Cramer's Charitable Trust is long NVDA. See here for a full list of the stocks.) As a subscriber to the CNBC Investing Club with Jim Cramer, you will receive a trade alert before Jim makes a trade. Jim waits 45 minutes after sending a trade alert before buying or selling a stock in his charitable trust's portfolio. If Jim has talked about a stock on CNBC TV, he waits 72 hours after issuing the trade alert before executing the trade. THE ABOVE INVESTING CLUB INFORMATION IS SUBJECT TO OUR TERMS AND CONDITIONS AND PRIVACY POLICY , TOGETHER WITH OUR DISCLAIMER . NO FIDUCIARY OBLIGATION OR DUTY EXISTS, OR IS CREATED, BY VIRTUE OF YOUR RECEIPT OF ANY INFORMATION PROVIDED IN CONNECTION WITH THE INVESTING CLUB. NO SPECIFIC OUTCOME OR PROFIT IS GUARANTEED.
Four leading AI models discuss this article
"Earnings revisions justify the move but do not eliminate near-term volatility from position trimming."
Dell's AI server customer base surpassing 5,000 with 50% growth in six months shows demand broadening beyond Nvidia's core hyperscalers. FY27 EPS estimates leaping from $13.12 to $16.85 keep the post-rally multiple at just 24.3x, confirming the move tracks fundamentals. Yet the piece glosses over capex digestion risk: if 2025 ROI data disappoints, planned data-center builds could be deferred or scaled back sharply. Mechanical selling from oversized fund weights after a 30% single-day spike adds another near-term overhang that earnings strength alone cannot neutralize.
Even if current estimates hold, Dell still trades at a 24x multiple on forward earnings that could compress quickly if custom ASICs from hyperscalers displace third-party servers faster than expected.
"Dell's valuation hasn't re-rated on fundamentals alone; it's re-rated because estimates were massively too low, which is a different (and riskier) signal than organic multiple expansion."
Dell's 30% pop on earnings beats is real, but the article conflates two separate things: (1) AI capex is genuine and growing, and (2) current valuations are justified. FY27 at 24.3x forward P/E isn't cheap—it's in-line with pre-report levels only because estimates jumped 28%. That's circular reasoning. The article also glosses over Dell's cyclicality: server demand is lumpy, customer concentration risk (hyperscalers dominate), and the fact that 5,000 AI server customers still represents a narrow base. Nvidia's keynote and broad-based enthusiasm could easily trigger profit-taking. The real risk: we're pricing in perfect execution and sustained capex growth that may not materialize if hyperscalers hit ROI targets and throttle spending.
If AI infrastructure spending truly is 'far bigger than most can imagine' and Dell's customer base is growing 50% in six months, then waiting for a pullback to buy is leaving money on the table—you're timing the market, not time in the market.
"The market is ignoring the cyclical risk that Dell’s hardware-heavy business model is susceptible to a sharp, sudden correction if hyperscaler capital expenditure shifts away from physical server capacity."
Dell’s earnings revision is impressive, but the market is conflating 'AI demand' with 'sustainable margin expansion.' While the forward P/E compression looks attractive, it assumes Dell maintains its current pricing power in a commodity-heavy server market. The massive 30% jump reflects a valuation catch-up, not just growth. The real risk is the 'lumpy' nature of hyperscaler capex; if AI infrastructure spending shifts from compute-heavy to software-heavy, Dell’s hardware-centric model will face a brutal deceleration. Investors are currently pricing in a permanent shift in Dell's earnings power, but I suspect we are seeing a cyclical peak in server refresh cycles masquerading as a secular AI revolution.
If Dell’s 5,000+ AI customer count represents the start of a multi-year enterprise upgrade cycle rather than just a hyperscaler spending spree, the current earnings estimates may actually be too conservative.
"Durable AI-driven demand can lift earnings, but a sustainable rally requires confirmable revenue growth and margin expansion beyond a short-lived, parabolic move."
Dell's blowout quarter reinforces an AI capex tailwind for data-center gear, lending credibility to the 'don't fight the tape' logic that earnings momentum can justify a rally beyond macro noise. Yet the article glosses a critical balance: after a parabolic run, multiple expansion is unlikely to do the heavy lifting; forward P/E sits around 24x for FY27 and ~20x for FY28, implying any upside must come from higher margins or faster revs, not just multiple relief. The real risk is AI demand peaking early, a demand-supply mismatch, or a margin squeeze from component costs and competition. Dell remains a winner only if compute demand proves durable.
The strongest counter is that AI hardware demand may normalize after an initial surge, turning this into a near-term multiple expansion story rather than durable earnings growth; if hyperscalers slow capex or margins erode, the rally could fade despite positive quarterly prints.
"Broadening customer base could protect Dell margins via enterprise mix even if hyperscaler spending slows."
Claude flags circular estimate revisions, yet the 5,000-customer expansion implies Dell can shift mix toward higher-margin enterprise deals less exposed to hyperscaler ASIC displacement. That dynamic could stabilize gross margins even if capex lumpy-ness hits in 2025, a buffer none of the prior takes quantified. The 50% six-month growth rate matters more for margin mix than headline revenue if pricing power holds on custom configurations.
"Enterprise margin premium collapses if hyperscalers' custom ASIC playbook cascades down to mid-market buyers demanding similar unit economics."
Grok's margin-mix buffer assumes enterprise customers pay premium pricing for custom configs—but that's precisely where competition from hyperscaler ASICs bites hardest. Enterprise buyers, watching hyperscalers optimize capex, will demand similar customization at lower cost. The 5,000-customer base is still 90% hyperscaler-dependent by revenue; mix shift doesn't materialize without proof enterprise deals are actually higher-margin. That proof isn't in the earnings call.
"Dell’s supply chain dominance in the Nvidia ecosystem provides a margin moat that mitigates the risk of enterprise-side pricing pressure."
Claude, you’re missing the supply chain leverage. Dell isn't just selling boxes; they are the primary integrator for Nvidia’s Blackwell rollout. Even if enterprise buyers demand lower costs, Dell’s ability to secure H100/B200 allocations creates a scarcity premium that hyperscalers cannot replicate. The 'commodity' argument fails when supply is constrained; Dell’s backlog isn't just revenue, it's a moat. If they control the hardware pipeline, they dictate the margin, regardless of the customer mix shift.
"Nvidia-based moat is not durable; at scale, Dell's margins and backlog hinge on a fragile supply-and-mix dynamic that can unravel if suppliers broaden OEM access or capex normalizes."
Gemini's supply-chain moat hinges on Nvidia allocations; that's not durable. If Nvidia expands direct OEM access or hyperscalers diversify, Dell's backlog and pricing power could erode quickly. A capex normalization would tighten hardware margins and boost substitution risk from in-house solutions. And the claim that 5,000 customers are mostly hyperscalers isn't proven; execution risk and mix dynamics could derail the thesis even with today's backlog.
Despite impressive earnings and AI capex growth, panelists express caution due to potential capex digestion risk, cyclicality, customer concentration, and margin squeeze from competition and component costs.
Potential shift in mix toward higher-margin enterprise deals less exposed to hyperscaler ASIC displacement
AI demand peaking early or a demand-supply mismatch leading to a margin squeeze