Dell Technologies Inc.

Dell closed 2025 at $125.88, bottomed at $111 on Jan 20 as memory-cost fears peaked, and has since re-rated in three steps — each a guidance raise rather than a multiple story. The Feb 26 print (+22% the next day) set a $12.90 FY27 EPS guide Morgan Stanley said may be unrealistic; the May 28 print (+33%) lifted it to $17.90; the Sept 1 print (+15.8%, the S&P 500's top gainer) lifted it to $25.50. On Sept 9 the shares touched $562.99 intraday, then reversed to close at $535.25; they trade at $517.45 this morning. Between June 1 ($466) and Aug 31 ($456) the stock went nowhere as estimates climbed — it took a third raise to move it.

Quarterly revenue has doubled in five quarters, from $23.4B in the April-2025 quarter, and the FQ3 guide ($49B, +81%) points higher. Two cautions on reading the bars. First, a growing share is price: Dell's PC shipments fell 4.9% year-over-year in calendar Q2 (Omdia) while Client Solutions revenue rose 20%, and management describes traditional-server growth as partly “uplifted by the increased cost of the underlying material.” Second, AI revenue tracks supply and customer data-center readiness, not bookings: FQ1 and FQ2 AI revenue were nearly identical ($16.1B, $16.4B) while orders swung from $24.4B to $60.9B.

The cleanest way to see what drove FQ2 is to hold AI constant. AI-optimized server revenue barely moved from FQ1 to FQ2 ($16.1B to $16.4B). Traditional servers and networking rose $2.0B to $10.5B and storage $0.5B to $4.9B. Yet ISG operating income jumped $1.7B, from $3.06B to $4.78B, and the segment margin went from 10.5% to 15.0%. Unless AI servers — which management concedes dilute the rate — suddenly became highly profitable, nearly all of that $1.7B came from $2.5B of incremental traditional-server and storage revenue. That is a scarcity signature. Dell had memory and CPUs when peers did not (Morgan Stanley's Taiwan checks credited it with “better access to memory supply (and pricing)”), it gained “more than 10 points” of traditional-server share in two quarters, and it repriced repeatedly.

This is not a criticism — Dell runs the best supply chain in the industry and is being paid for it. The question is duration, and the FQ3 guide shows the premium plateauing rather than compounding: ISG revenue up ~145% to ~$34.6B on $19B of AI servers, traditional servers and storage roughly flat sequentially at ~$15.5B, and ISG operating income flat near $4.7B (a margin guided “up just over 1 point” from FQ3 FY26's 12.4%). TrendForce's July survey expects server DRAM contract prices to rise another 13–18% in 3Q26 and to keep rising quarterly through 2H27 at a moderating pace, with 2027 RDIMM bit supply growing only 15–20%. Two details cut against Dell: long-term agreements now cap increases for the largest U.S. cloud buyers, shifting pricing pressure to buyers without them, enterprise OEMs among them, and Jeff Clarke conceded that “more cost-sensitive customers are extending their upgrade cycles.” Our read: the premium persists through most of FY28, then fades as new DRAM capacity arrives — a plateau year for traditional-server profit, not a growth year.

Dell reports two segments. The Infrastructure Solutions Group (ISG: $31.8B in FQ2, 68% of revenue, 81% of segment operating income) sells AI-optimized servers — rack-scale NVIDIA and AMD systems integrated, deployed, serviced and increasingly financed for neoclouds, sovereigns and 6,500+ AI customers — plus PowerEdge servers, networking and storage, where Dell-IP demand has grown faster than the market for six straight quarters; with most of the installed base still on 14th-generation or older servers, management sees a durable refresh runway. The Client Solutions Group (CSG: $15.0B) is PCs, overwhelmingly commercial. Services ($5.9B) were flat — the growth is hardware.

In AI servers Dell competes with Supermicro, HPE and Lenovo for neocloud, enterprise and sovereign builds; the largest hyperscalers buy mostly direct from Taiwanese ODMs such as Foxconn and Quanta, a channel Dell largely cedes. Its edge is scale, supply access, deployment speed and financing. The whole complex is riding the same price-led upcycle — HPE's FQ3 revenue rose 33% (helped by Juniper) and it guided FY26 EPS to $3.75–$3.85, above consensus, while Supermicro guides FY27 net sales to $65–$72B. Dell trades at a premium to the box-makers and a discount to Everpure.

We anchor to Dell's FY27 guide and build FY28 from the exit rate, not the trend. Guidance implies FQ4 EPS of ~$7.10 (the $25.50 full year less $11.90 in H1 and $6.50 in FQ3). Annualizing the second half ($13.60) gives ~$27.20, and a few points of buyback accretion take it to our $28.00 FY28E. That assumes AI revenue grows to roughly $90B on the backlog while traditional-server and PC pricing and ISG margins plateau at second-half levels. The $30.91 consensus needs another leg of growth: FY28 revenue of ~$236B, 16% above the second-half exit rate.

Implied Price = P/E × FY28E non-GAAP EPS (fiscal year ending Jan-2028)

The grid's most useful cell is not ours. At consensus FY28 EPS ($31) and the same 19x, it returns $589 — the Street's mean target to the dollar. The disagreement between us and the Street is therefore entirely about FY28 earnings, not the multiple: whether Dell grows beyond its second-half run-rate (consensus) or plateaus near it (us).

Michael Dell, Chairman and CEO, and his affiliated holders own the 276.7M high-vote Class A shares — 44% of the 635.8M outstanding — and he has not sold in 2026 (his last sale: 6.25M shares at ~$160 in October 2025). Vice Chairman and COO Jeff Clarke runs operations and the AI business under a one-time 2.5M-share performance option, vesting in 2031, tied to market-cap and adjusted-free-cash-flow hurdles — notable, since adjusted FCF excludes the cash the financing ramp consumes. CFO David Kennedy has held the seat since September 2025. Execution has been superb: a 10% headcount cut (to ~97,000) held cost growth far below revenue growth, taking FY27 operating expenses to ~8% of revenue, the lowest in Dell's 42-year history.

The chart shows what the income statement hides. FQ2 non-GAAP net income nearly tripled year-over-year while GAAP free cash flow halved, absorbed by financing originations, inventory (doubled since January to $21.3B) and receivables, with payables of $49.7B covering much of the inventory. Adjusted free cash flow adds back the financing on the logic that DFS is a funded finance arm — but since January financing receivables rose $6.1B while DFS debt rose only $0.5B, to $9.6B, so the core balance sheet and the bond market funded almost all of the increase. Neocloud customer Boost Run expanded a DFS financing partnership alongside a $1.44B Dell order (per its SPAC partner), and Dell invested in UK neocloud Nscale's $2B Series C.

Capital return looks generous — $6.3B returned in H1, a 20% dividend increase, the share count down 5% — but repurchasing $3.8B of stock in a $1.0B free-cash-flow quarter while issuing bonds and extending $7.5B of customer credit is leverage by another route. At 0.8x core leverage it is affordable; it is also why we want FQ3 originations back toward the AI shipment rate before paying a higher multiple.

Positioning after a four-fold run is hedged, not euphoric. Chain-wide, puts outnumber calls (put/call ratio 1.48), and the protection is long-dated — the January 2027 series alone carries 78K puts against 45K calls, holders insuring gains through the FQ3 print. Near the money the picture flips, with calls leading and the $600 strike the largest call concentration. Thirty-day implied volatility is ~62%, mid-range for Dell's past year (IV rank 43), and early flow today leaned bearish ($22.8M of bearish against $18.7M of bullish premium).

The Street is overwhelmingly constructive and still moving up. Among the 22 brokers we track, 18 rate Buy/Outperform against 4 Hold-equivalents (UBS, TD Cowen, Morgan Stanley, Wolfe); the FactSet analyst consensus is Overweight with a $589.33 mean target as of Sept 9, up from $195 in mid-May, and published targets reach $735. The skeptics cluster at $499–$500: UBS argues the multiple already reflects fears that AI-driven compute growth “may slow sharply” in FY28–FY29 and that steep industry-wide price increases “raise durability concerns,” while Morgan Stanley says the valuation already reflects Dell's strength. Targets rose by less than estimates — UBS lifted its FY27 EPS forecast 34% but its target only 10% — so the $589 mean is a bet on the FY28 number, not on a higher multiple.

We would move to BUY if: (1) FQ3 shows ISG operating margin of 14.5% or better on ~$19B of AI revenue, against ~13.5% implied by the guide — evidence the pricing premium is still building rather than plateauing; (2) DFS originations fall back below ~30% of AI shipments while GAAP free cash flow recovers toward net income; (3) the stock pulls back toward the 50-day average (~$445) with estimates intact, or 16x our FY28E; or (4) February's first FY28 guide points above $30, validating consensus.

We would move to SELL if server DRAM contract prices flatten or fall before mid-2027, ISG margin drops below 12.5% as AI mix rises, DFS past-dues or reserves step up, or AI orders run below shipments for two consecutive quarters, burning backlog. The HOLD rests on FY28 being a plateau; evidence of a decline instead would pull our target toward the $300 bear case.

Dell is having the best year in its history: revenue guided up 69%, EPS up 148%, a $95B AI backlog, share gains in servers and storage, and the leanest cost structure in four decades. The momentum investor's instinct — that estimates rising this fast keep rising — has been right three times since February.

What is in question is what the market is paying for. At $517.45, Dell is priced for a FY28 that holds onto margins a memory shortage created, valued on an adjusted cash flow that excludes the growing credit Dell extends to its own customers. Treating FY28 as a plateau at the second-half exit rate, the stock is worth $530 — not a short, but a stock already at fair value on the best news it is likely to get before February's FY28 guide.