KEY POINTS
- Snowflake surged more than 30% the day after its earnings report, while HPE gained 16% on better-than-expected results and rare 2027 guidance issued in June.
- Both moves signal that enterprise AI spending is accelerating at the platform and infrastructure layer, not slowing — a direct rebuttal to post-June 5 chip-selloff pessimism.
- Traders should watch whether HPE's 2027 guidance holds up under analyst scrutiny when markets reopen Friday June 20, and whether SNOW can sustain its post-earnings level above the 30% gap.
Snowflake stock surged more than 30% the session after its earnings report — the kind of move that resets a thesis, not just a price target. Paired with HPE's 16% single-day gain on strong results and 2027 guidance issued in June, the message from the enterprise AI layer is unambiguous: real money is still being committed to AI infrastructure at scale, and the buyers are not blinking.
The Numbers That Moved Markets
HPE's move is the more strategically significant of the two. CEO Antonio Neri didn't just beat the quarter — he issued full-year 2027 guidance in June of 2026, a level of forward conviction that's rare even in bull markets and almost unheard of when a sector has just watched $923 billion in semiconductor market cap evaporate in a single session. Neri's pushback on dot-com-era comparisons was pointed: "There is real demand," he said, and explicitly stated he sees no AI demand peak on the horizon. That framing matters because it directly addresses the bear case that emerged after Broadcom's cautious AI chip outlook triggered the June 5 semiconductor wipeout. HPE is not a chip company — it is the infrastructure layer between the chips and the enterprise — and its visibility into forward-demand curves is arguably more accurate than any single semiconductor vendor's guidance.
Snowflake's 30%-plus gain is a different kind of signal. Snowflake sits at the data layer — the platform where AI models are trained, fine-tuned, and deployed against enterprise data sets. A 30% post-earnings gap at a company of Snowflake's scale is not a short squeeze or a momentum trade; it reflects a repricing of revenue estimates by institutional money. CEO Sridhar Ramaswamy held his tone at the company's annual partner conference in San Francisco, showing no hedging on demand. When a CEO stays on offense at a partner conference the week after a blockbuster earnings print, that's a signal about pipeline visibility, not just current-quarter execution.
What the June 5 Rout Got Wrong
The June 5 semiconductor selloff — which erased roughly $923 billion in market cap across the top 10 decliners and sent the Philadelphia Semiconductor Index (SOX) down more than 8% — was triggered by a specific set of inputs: Broadcom's cautious AI chip outlook, a deepening memory chip crisis, and a projected collapse in global smartphone demand. Those are real problems, but they are concentrated in specific parts of the semiconductor stack. The market's mistake on June 5 was treating them as a read-through to all AI spending, when the actual dynamic is a structural divergence between booming AI data center demand and softness in automotive, consumer devices, and non-data center communications.
HPE's 2027 guidance and Snowflake's earnings both argue that the divergence is real and that the AI infrastructure layer — servers, networking, data platforms — is insulated from the consumer and automotive chip cycles that spooked the market. Broadcom's earnings tone was the trip wire on June 5, but Broadcom is a different business than HPE or Snowflake. Broadcom's exposure to custom AI silicon and networking ASICs means its guidance swings are amplified by hyperscaler capex timing — a lumpy, project-driven revenue stream. HPE's server and hybrid cloud business reflects a broader, more diversified demand base across thousands of enterprise customers. Those customers are still buying.
What Traders Watch Next at Friday's Open
Markets reopen Friday June 20 after Juneteenth, and the HPE and Snowflake prints are the most constructive data points traders have had since the June 5 wipeout. The immediate question is whether the post-earnings gaps hold — or whether the broader semiconductor overhang drags even the strong performers back. SNOW's 30% gap is a wide one to defend in a sector where sentiment is still fragile. Any negative newsflow on Samsung's labor situation, the Tencent/Blackwell export story, or additional cautious guidance from chip vendors could provide cover for profit-takers to close post-earnings longs in both names.
The more durable trade is the thematic one. Deloitte's 2026 Semiconductor Industry Outlook projects global semiconductor sales reaching a historic peak of $975 billion this year, driven by AI infrastructure. That number doesn't exist without the enterprise software and platform spending that HPE and Snowflake are reflecting. The AI investment cycle is not a single-layer story — it runs from TSMC's 3nm wafers through NVIDIA's accelerators, through HPE's servers, and into Snowflake's data platform. June 5 broke one link in that chain on sentiment, not fundamentals. HPE's 2027 guidance and SNOW's 30% gap are the market's attempt to reprice that distinction.
For positioning going into Friday: HPE's 16% gain sets a new technical floor, and Neri's 2027 guidance gives fundamental bulls a specific anchor to hold against. On SNOW, the critical level is whether it can consolidate above the pre-earnings price plus 20% — a partial giveback of the 30% gap would still represent a net-positive repricing and would confirm institutional accumulation rather than a one-day momentum flush. Watch the opening 30 minutes on June 20 closely. Volume confirmation on both names will tell you whether the institutional money that drove these post-earnings gaps is holding or distributing into the holiday-shortened week's first session.
