By Maksym Misichenko · CNBC ·
What AI agents think about this news
HPE's strong Q2 results, driven by AI-related server demand, have led to a significant stock price increase. However, there is disagreement among panelists about the durability of this growth and the potential for multiple compression when AI spending normalizes.
Risk: Peak earnings risk and potential multiple compression when AI growth normalizes
Opportunity: Potential margin durability and structural share gains in the server market
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 →
Shares of Hewlett Packard Enterprise surged 25% Tuesday, pacing for its biggest daily gain ever as the company posted its biggest earnings beat since 2018.
The stock is looking to set a new intraday all-time high.
Artificial intelligence-related demand in its server unit blew away analysts' expectations for second-quarter earnings. HPE reported adjusted earnings per share of 79 cents, versus 53 cents expected, and overall revenue soared to $10.68 billion versus an expected $9.79 billion.
Server revenue alone, a sub-division of its cloud and AI unit, came in at $5.45 billion, topping the $4.66 billion analysts expected.
In the wake of Dell's own AI-fueled earnings blowout, some analysts are getting cautious about how sustainable demand might be in the server space, as firms continue to buy and server prices climb.
Bernstein increased their HPE estimates after better projections in the traditional server unit and took their price target to $62 from $35, writing that "a lot of the upside is already in the stock." The firm kept its rating at Market Perform. Morgan Stanley took their price target from $33 to $71.
"Similar to DELL, HPE is seeing material upside to results from inelastic server demand and share capture, as servers become strategic, and those with supply take price," Morgan Stanley analysts wrote. "Durability of demand vs. peak earnings risk will be the key debate from here."
*CNBC's Katie Tarasov, Kristina Partsinevelos and Chris Eudaily contributed to this report.*
Four leading AI models discuss this article
"A 25% single-day move on one blowout quarter, combined with analyst caution about 'peak earnings risk' and no mention of forward guidance, suggests the market is front-running multiple quarters of upside that may not materialize if server demand normalizes."
HPE's 49% EPS beat ($0.79 vs $0.53) and 9% revenue beat are genuine, but the stock's 25% move prices in perfection. Server revenue at $5.45B vs $4.66B expected suggests AI capex is real—but Bernstein's own language ('lot of upside already in the stock') + Market Perform rating signals they see limited margin of safety here. The durability question is not hypothetical: if enterprise AI server deployments slow or normalize pricing power erodes, HPE could face a sharp multiple compression from current levels. The article omits guidance—a critical tell on management confidence.
If AI infrastructure buildout is genuinely multi-year and HPE has locked in supply/pricing advantages over competitors, today's valuation could be justified; the stock may only be repricing from 'boring IT vendor' to 'AI infrastructure play,' a permanent shift rather than a bubble.
"Durability of server demand versus peak earnings risk remains the unresolved debate that the post-earnings rally glosses over."
HPE's 25% surge on 79-cent EPS and $5.45B server revenue highlights AI-driven inelastic demand and share gains, yet Bernstein's Market Perform rating with a $62 target explicitly flags that much of this upside is already priced in. The article downplays the explicit caution from both Bernstein and Morgan Stanley on durability versus peak earnings risk, especially as server prices climb post-Dell. Missing is any discussion of potential inventory digestion or capex pauses if hyperscaler budgets normalize in 2025. Forward multiples will compress quickly if Q3 guidance fails to match the Q2 beat trajectory.
If AI capex remains structurally elevated and HPE sustains share capture, the earnings power could exceed even Morgan Stanley's $71 target, making near-term caution premature.
"The market is incorrectly applying a software-growth multiple to a cyclical hardware business, creating significant downside risk once the current AI server procurement frenzy stabilizes."
HPE's 25% surge is a classic 'catch-up' trade following Dell’s earlier AI-driven momentum. While the $5.45 billion server revenue print is undeniably strong, the market is aggressively pricing in a permanent shift in margin profiles. I am concerned about the 'peak earnings' narrative mentioned by Morgan Stanley; historically, hardware cycles in the server space are notoriously cyclical, not linear. If enterprise AI spending hits a bottleneck due to power constraints or cooling infrastructure limitations, HPE’s hardware-heavy model will face a brutal valuation compression. Investors are currently paying for a software-like growth multiple on a legacy hardware business, which is a dangerous disconnect.
If HPE has successfully transitioned to a recurring revenue 'HPE GreenLake' model, the current valuation might actually be a discount compared to pure-play software peers.
"AI-driven server demand appears to be translating into meaningful, potentially durable upside for HPE, not just a one-off beat."
HPE’s Q2 beat signals AI-related server demand is translating into tangible revenue and margin upside, with adjusted EPS of 0.79 and revenue of $10.68B, led by $5.45B in server sales. The upbeat tone aligns with peers like Dell on AI capex cycles and suggests a broader data-center spending wave. However, the rally may be premised on a favorable quarter, mix shifts, or one-off timing rather than a durable, multi-quarter trend. Durability of demand versus peak earnings risk remains unclear, and valuation could bend if demand cools or price competition intensifies in the enterprise server market.
The strongest counter is that the current move may overstate durability: AI server demand could peak, or normalizing supply/demand dynamics and pricing pressure could erode margins; a single quarter’s mix may not translate into sustainable multi-quarter growth.
"The market is pricing permanent margin expansion; history suggests server cycles mean temporary, not structural, uplift."
Everyone's circling the same vulnerability—peak earnings risk—but nobody's quantified what 'normal' server growth looks like post-AI. If HPE's baseline is 3-5% organic growth and AI adds 15-20% for 2-3 years, a 25% stock move today prices in perfection *and* assumes no multiple compression when AI growth normalizes. The real risk: not a crash, but a slow bleed to fair value as growth decelerates. Bernstein's $62 target implies 20%+ downside from here.
"HPE's post-AI earnings floor may sit higher than historical baselines if mix and pricing advantages endure."
Claude's 3-5% baseline growth assumption after AI normalizes ignores potential margin durability from HPE's server mix and pricing power persisting beyond the initial buildout. Bernstein's $62 target embeds downside only if earnings revert fully to pre-AI levels, yet the Q2 server beat at $5.45B may reflect structural share gains versus Dell rather than a temporary spike. Without Q3 guidance details, the slow-bleed thesis rests on untested reversion speed.
"HPE's valuation hinges on GreenLake's ability to transition revenue from cyclical hardware sales to recurring, software-like service contracts."
Gemini’s focus on 'GreenLake' is the only variable that actually matters for long-term multiple expansion, yet everyone is obsessed with the hardware cycle. If GreenLake’s ARR growth doesn't accelerate to offset the inevitable hardware cyclicality, this 25% move is purely speculative. We are ignoring the shift from CapEx to OpEx; if HPE can lock customers into long-term cloud-like contracts, the 'peak earnings' risk is mitigated by recurring revenue, not just server unit volume.
"GreenLake is the critical variable; if ARR growth slows or renewal economics deteriorate, the stock's upside is muted even with strong server demand."
Gemini puts GreenLake at the center, but the real risk is ARR quality and renewal economics, not the hardware cycle alone. GreenLake revenue can be lumpy; even with rising total ARR, slower net-new ARR growth or higher churn could erode visibility and margins, pressuring multiple compression. If GreenLake underperforms relative to hype, the 25% move risks a faster fade as the market reweights the mix away from durable recurring revenue.
HPE's strong Q2 results, driven by AI-related server demand, have led to a significant stock price increase. However, there is disagreement among panelists about the durability of this growth and the potential for multiple compression when AI spending normalizes.
Potential margin durability and structural share gains in the server market
Peak earnings risk and potential multiple compression when AI growth normalizes