HPE — Hewlett Packard Enterprise Company

Sentiment flipped decisively bullish after the 2026-09-02 Q3 FY26 report: revenue of $12.21B (+34% YoY), non-GAAP EPS of $1.11 vs. $0.93 consensus, Cloud & AI operating margins more than doubling YoY to 17.0%, a record $7.6B AI systems backlog, a $3.5B hyperscale AI inferencing contract, and an expanded Oracle/Juniper networking collaboration. Management raised FY26 non-GAAP EPS to $3.75-3.85 and introduced a FY27 framework of 13-17% revenue growth, 16-20% EPS growth and at least $5B free cash flow, plus a pledge to return 75%+ of FCF and accelerate buybacks. Sell-side followed hard: BofA to $82 (subsequently $88), Barclays to $79, Citi to $76, Morgan Stanley upgraded to Overweight in August. The stock returned +19.4% in the week into 2026-09-13 as institutions re-rated HPE from a slow-growth hardware value name into an AI/networking infrastructure winner.

The trend is a genuine inflection, not a blip. Revenue ran $6.73B/$7.18B/$7.67B/$8.48B through FY24, then accelerated every quarter: $9.30B (Q1 FY26, +19.1% YoY), $10.68B (Q2 FY26, +40.7%), and $12.21B (Q3 FY26, +32.7%). Gross margin tells the same story and is the most important number here: 28.4% in Q3 FY25, 30.4% in Q4 FY25, 35.9% in Q1 FY26, 36.5% in Q2 FY26, and 40.1% in Q3 FY26 - roughly 1,170bps of expansion in four quarters, which is a mix shift toward networking (Juniper) and higher-value AI systems, not just volume. Operating margin went 5.4% -> 4.4% -> 9.7% -> 7.0% -> 11.4%; the 11.4% in Q3 FY26 is the highest in the 12-quarter window. GAAP diluted EPS swung from -$0.82 in the Apr-2025 quarter (a -$1.05B GAAP loss tied to Juniper purchase accounting/charges, despite positive operating income of $255M) to +$1.06 in Jul-2026. Two caveats worth stating plainly: the Apr-2025 loss quarter still distorts every trailing-twelve-month metric, and Q3 FY26 net income of $1.54B exceeded operating income of $1.39B, implying a below-the-line boost. But directionally, four straight quarters of accelerating revenue with expanding margins is a structural trend, and the accruals ratio of -0.12% at 2026-07-31 confirms cash generation is running ahead of reported profit rather than behind it.

Split verdict - expensive on trailing numbers, fair on forward numbers. As of 2026-09-13, the trailing GAAP P/E was 30.9x against a 5-year average of 11.7x, P/S was 1.96x against a 5-year average of 0.77x, and P/B was 3.13x; FMP's own valuation context labels this 'Rich vs. its own history.' But that history describes a pre-Juniper, pre-AI HPE that deserved a ~12x multiple. On forward numbers the picture is different: consensus next-FY (FY26) non-GAAP EPS of $3.798 puts the stock at 16.4x, and growing that at the low end of management's FY27 framework (16%) implies ~$4.48 in FY27, or roughly 13.9x. PEG is 0.81. Trailing twelve-month free cash flow of $4.16B ($896M + $827M + $609M + $1.824B) is a 5.1% FCF yield, rising to ~6.1% on the guided >=$5B FY27 figure. The honest caveat: GAAP earnings power is far below non-GAAP (TTM GAAP EPS $2.06 vs. ~$3.80 FY26 non-GAAP guided), so the 'cheap on forward earnings' argument depends on continuing to accept non-GAAP definitions that exclude large amortization and restructuring charges. Net: fair-to-slightly-rich at $62, not a bargain.

The market is paying roughly 16x the FY26 non-GAAP EPS that is 87% already booked and ~14x a FY27 number management only just framed. That is not a stock pricing in fantasy - it is a stock pricing in 'the guidance is real but I'm not paying a software multiple for it yet.' The asymmetry is now about estimate revisions rather than multiple expansion: the FMP consensus target of $67.75 (12 analysts; range $54-$86) implies only ~9% upside from $62.09, and even the broader post-print cluster around $69.18 only gets you ~11%. That means the stock needs further upward revisions - FY27 EPS moving from ~$4.48 to $5.00+ - to justify today's price, and the bull outliers (BofA at $82-88) are the ones underwriting exactly that. Watch the earnings line, not the revenue line: if the $7.6B AI backlog converts at the margins Cloud & AI showed in Q3 (17.0% operating margin), EPS revisions follow; if component costs eat that margin, the multiple compresses fast. Note also that the analyst rating breakdown is 19 buy / 19 hold / 1 sell - a consensus label of 'Buy' hiding a genuinely split house view.

As of 2026-09-13, HPE was extended in every direction. RSI(14) stood at 64.1 - elevated but not yet overbought - while price sat 13.6% above its 20-day SMA (~$54.6), 21.5% above its 50-day (~$51.1), and 82.6% above its 200-day (~$34.0). The one-week return of +19.4% versus a one-month return of just +5.61% tells you the entire move is a single post-earnings gap, and the three-month return of +28.9% says the same about the broader run. This is a momentum breakout, not a base. Buying here means paying up 13.6% above even the very short-term trend line with the 52-week high ($64.25) directly overhead as resistance. Entry-timing implication: momentum traders own it now; anyone with a value disposition should wait for a retest of the $53-55 zone where the post-earnings gap and the 20-day average converge, or at minimum wait for the price to consolidate sideways and let the moving averages catch up. Chasing a stock 82.6% above its 200-day is how you inherit a 20-25% drawdown on the first wobble.

• AI infrastructure capex supercycle: enterprise demand is shifting from pilots to production-scale inference and agentic AI workloads, which is directly fueling HPE's HPC, AI systems and networking orders (networking orders up 36% YoY).
• Component supply tightness: memory, NAND and CPU availability - not demand - is the binding constraint. HPE expanded multi-year purchase commitments from $6B to $30B to secure supply, which protects revenue but converts a shortage into a future fixed-cost obligation if demand cools.
• Interest rates and enterprise IT budgets: rate volatility affects enterprise capex cycles and customer financing through HPE Financial Services, though strong corporate cash flows and AI urgency have so far insulated hardware budgets.
• Networking competitive/regulatory landscape: the Juniper Networks integration is tracking ahead of synergy schedules, but HPE is now competing head-on with Cisco and Arista in AI data-center networking, and large-scale networking M&A draws regulatory scrutiny.
• Sector rotation and multiple risk: HPE is now traded as an AI infrastructure name, so any broad de-rating of AI-linked equities hits the multiple regardless of HPE's own execution; the stock's 1.44 beta amplifies that.
• Capital return regime: management's pledge to return at least 75% of free cash flow via accelerated buybacks makes the share count a second lever on EPS, but buybacks funded alongside $17.3B of long-term debt limit how aggressive that can get.

• Q4 FY26 earnings in early-to-mid December 2026, with guidance of $13.9-14.8B revenue and $1.20-1.30 non-GAAP EPS - the first real test of whether the raised numbers are achievable.
• FY27 formal guidance and the conversion of the record $7.6B AI systems backlog, plus any incremental hyperscale AI awards following the $3.5B inferencing contract disclosed on 2026-09-02.
• Juniper Networks integration milestones and synergy realization, plus progress deploying Juniper networking inside Oracle and other AI data centers.
• Accelerated share repurchase execution in Q4 FY26 (>=75% of FCF) and any resulting EPS accretion or share-count reduction.
• Supply-chain relief: if memory/NAND/CPU availability improves, HPE converts backlog faster and margin pressure from component costs eases - a double positive.
• Further sell-side target revisions; with 19 of 39 analysts still at Hold, upward rating migration is an underappreciated source of demand.

• Valuation risk after a 187.7% six-month move: TTM P/E of 30.9x vs. a 5-year average of 11.7x and P/S of 1.96x vs. a 0.77x average leaves no cushion if growth guidance slips by even one quarter.
• GAAP vs. non-GAAP gap: TTM GAAP EPS is $2.06 while FY26 non-GAAP guidance is $3.75-3.85 - a >45% gap driven by Juniper amortization, restructuring and stock-based comp. If the market stops capitalizing the non-GAAP number, the stock re-rates violently lower.
• Component cost and supply exposure: $30B of purchase commitments, up from $6B, is a large fixed obligation; a demand air pocket or a memory pricing spike would hit gross margin (currently 40.1% in Q3 FY26) directly.
• Customer concentration in AI: a handful of hyperscale buyers drive the backlog, so a single delayed or re-scoped deployment is a material revenue event.
• Balance sheet and liquidity: $17.3B of long-term debt against $26.5B of equity (D/E 0.76), $6.2B of cash, a current ratio of 1.11 and a quick ratio of 0.72 - thin working capital headroom for a company now carrying much larger purchase commitments.
• Technically extended: RSI 64.1 with price 13.6% above the 20-day and 82.6% above the 200-day makes any negative headline disproportionately painful.

This is a real operating turnaround, not a story stock: four straight quarters of accelerating revenue, ~1,170bps of gross-margin expansion in a year, a record $7.6B AI backlog, and a FY27 framework the company has never been in a position to issue before. But as of 2026-09-13 the price already reflects most of the near-term good news - $62.09 vs. a consensus target of $67.75, with my own forward-based fair value at $72, implies only low-to-mid teens upside against genuine execution risk (component supply, $30B of purchase commitments, a GAAP/non-GAAP earnings gap of nearly half). I would want to own HPE, not buy it here. Anyone entering today is buying 13.6% above the 20-day average after a 19.4% one-week rip; the better risk/reward is a pullback into the $53-55 zone or a multi-quarter holding period that lets the FY27 numbers actually print. This is analysis, not personalized financial advice.

Forward P/E on non-GAAP EPS, cross-checked against free cash flow. I applied 16x to my FY27E non-GAAP EPS of ~$4.48 (FY26 guidance midpoint of $3.80 grown at the low end of management's 16-20% FY27 framework) - a multiple sitting between Dell (~13x) and Cisco (~18-20x) that reflects HPE's improved networking/AI mix and high-teens EPS growth - giving $71.7. Cross-check: guided FY27 FCF of >=$5.0B across ~1.324B shares is $3.78/share, and at a 19x FCF multiple (a 5.3% yield, reasonable for a mid-teens grower) that is $71.8. Because both independent methods converge near $72, I set fair value at $72.00; note this rests on non-GAAP definitions, and a GAAP-based multiple on the same growth would produce a materially lower figure.