TD SYNNEX Corporation (NYSE: SNX) — The World’s Biggest Box-Mover, Re-Rated on the AI Boxes: A Cost-of-Capital Business Priced for the Boom to Keep Booming

Independent equity research note. The analytical body below carries no recommendation and no price target; the sole exception is the labeled Claude’s Take block.

⚡ Claude’s Take

The author’s own independent opinion and general information — not investment advice. The analytical body below carries no position.

Verdict: HOLD — do not chase at the richest multiple in the company’s history; accumulate-on-weakness only sub-~$190–200; not a short. TD SYNNEX is the scale-and-margin leader of a rational, consolidated broadline-IT-distribution oligopoly — a genuinely well-run business that has, for its entire life, earned roughly its cost of capital and never meaningfully more (ROIC ~8.9% ≈ WACC). The stock has nearly doubled in five months (~$157 in February to a $296.47 all-time high on June 24, 2026, now $244.64) and re-rated to ~10.5× spot EV/EBITDA and ~20× trailing GAAP earnings — versus its own ~7–8× / ~13× decade history and versus distribution peers Arrow (7.4×) and Ingram Micro (6.7×). The entire re-rating rests on one thing: Hyve, its hyperscale custom-server ODM arm, where gross billings grew +117% year-over-year to $5.5B last quarter on the AI-data-center build-out. My fair-value zone is ~$180–220 (≈8–9× EV/EBITDA, ~11–13× forward non-GAAP EPS of ~$18) — which is why $244 is a HOLD, not a buy: you are paying a growth-and-quality multiple for a business whose incremental growth is lower quality than its core.

The framing is a momentum re-rating of a value/cyclical distributor on an AI narrative — “quality-of-earnings-down, multiple-up.” The factor tape confirms it: SNX has run +78% in a year with a 2.5 Sharpe and a trivial 16% drawdown (a one-way street), yet its factor DNA is still mid-cap value + dividend-yield with its single closest statistical cousin being Sanmina, an EMS/ODM — and it carries a negative loading on the Robotics-&-AI factor. In other words, the market is re-rating a value name, not yet pricing a secular-AI compounder, and the realized momentum has outrun the factor identity. The bear case is not that Hyve is fake — it is booming and real — but that it is thin-margin (~3.3% of billings), customer-concentrated (one customer = 11% of total FY25 revenue), working-capital-hungry (the company consumed ~$330M of free cash last quarter and had negative FCF in the 2022 growth year), and cyclical — precisely the attributes a peak multiple should not reward. The June-25-to-July-2 pullback (~17% off the high) was post-blowout profit-taking, not a sponsor exit: contrary to first appearances, the June-29 “3.24M-share secondary” is a de-minimis (~0.27%) Amazon commercial-warrant resale registration, and the legacy private-equity overhang is gone (Apollo fully exited in April 2024; the founding MiTAC family is under 5% and shrinking). So I am not leaning on an overhang to justify HOLD — I’m leaning on price versus quality. It is genuinely cheap on forward non-GAAP (~13×) and the buyback shrinks the count ~4%/yr — which is why I stop at HOLD and explicitly not short.

Conviction: medium. The single piece of evidence that flips me bullish: Hyve holding a >3% operating margin while the two newly-signed hyperscaler programs ramp and the top-customer concentration falls — proof of a durable, diversifying ODM franchise that earns the premium. The single piece that flips me bearish: a hyperscaler capex air-pocket or a top-customer program loss that cracks Hyve’s growth, leaving SNX carrying working-capital drag into a de-rating toward the peer ~7× EV/EBITDA. Tag: “The world’s biggest box-mover, re-rated on the AI boxes.”

📈 Stock Price Action — Five-Year Event Map

TD SYNNEX spent nearly four years as a range-bound, ~$90–135 broadline distributor that the market valued like the low-growth, cost-of-capital business it was — then, from early 2026, it doubled. From a $95.81 low (April 4, 2025) the stock ran to an all-time high of $296.47 (June 24, 2026) before pulling back to $244.64 (July 2, 2026) — now −17.5% off that high, inside a 52-week range of roughly $134–$296. The arc is the whole thesis in one line: the market paid ~7–8× EV/EBITDA for SNX as an IT-plumbing grinder for four years, and is now paying ~10.5× because Hyve turned it into an AI-infrastructure growth story. Price moves below are FACT; the attributed drivers are INTERPRETATION.

Cycle narrative. (1–3) For its first four years as a merged entity, TD SYNNEX behaved exactly as its economics implied: a ~$60B-revenue, ~2%-operating-margin distributor that de-rated in the 2022 rate shock, generated negative free cash flow in that same growth year as it built inventory, and then ground sideways in a 2023–24 IT-demand air-pocket while buying back ~$600M of stock a year. (4–5) A brief 2025 beat gave way to a Q1 FY25 miss and the April-2025 tariff shock that carried it to a $95.81 low. (6–7) From that low the stock compounded relentlessly as the AI-infrastructure narrative found its distribution proxy: Hyve, the hyperscale ODM arm, inflected from a rounding error to ~28% of quarterly revenue, and back-to-back Q1/Q2 FY26 blowouts (Hyve billings +117%, programs with all five U.S. hyperscalers, an HPE global-distribution win) drove the multiple to its richest-ever level and the price to a $296.47 ATH. (8) The first crack came immediately after the June-25 Q2 print: post-blowout profit-taking took the stock ~17% off its high in a week, coinciding with (but not primarily driven by) a small S-3 takedown registering the shares underlying an Amazon commercial warrant — a de-minimis ~0.27% technical, not the legacy-sponsor exit it superficially resembled (Apollo had fully exited in 2024). Each move ties to an earnings 8-K, a guidance change, or a capital-markets event in the public record.

1. Executive Summary

TD SYNNEX Corporation is the world’s largest — or, on a revenue basis, co-largest with Ingram Micro — broadline distributor and solutions aggregator for the information-technology ecosystem. It sits upstream in the two-tier IT channel: it buys hardware, software, and cloud from more than 1,500 vendor lines (Apple, HP, Lenovo, Dell, Cisco, Microsoft, NVIDIA, and hundreds more) and resells, configures, finances, and logistically services them for roughly 150,000 downstream resellers, VARs, MSPs, and integrators — the layer that includes CDW, Insight, and SHI. FY2025 (November-end) revenue was $62.5B on a GAAP net basis (~$80B of gross billings), producing a ~7.0% gross margin, ~2.3% operating margin, and ~1.3% net margin — the classic ultra-thin, high-turns, working-capital-intensive economics of distribution. The company was formed by the September-2021 merger of SYNNEX and Tech Data and had earlier (2020) spun off its business-services arm as Concentrix (CNXC).

The investment tension is unusually clean, and it is a valuation-and-quality tension, not a business-quality collapse. The business is well-run: SNX is the scale and margin leader among pure broadline distributors, it converts (in normal years) ~95% of net income to free cash flow, it has shrunk its share count from 95.5M to ~79.5M since 2021, and it grows the dividend. But it has never escaped the structural ceiling of its industry — ROIC of ~8.9% sits right at its cost of capital, the proof that its real scale/working-capital moat protects share and existence, not excess returns. For four years the market priced this correctly, at ~7–8× EV/EBITDA and low-teens P/E.

What changed is Hyve — and the multiple. Hyve, SNX’s hyperscale custom-server ODM business, is riding the AI-data-center capex super-cycle: gross billings +117% year-over-year to $5.5B in Q2 FY26 (~28% of quarterly revenue), with programs now signed across all five of the largest U.S. hyperscalers. On the back of two consecutive blowout quarters, the stock has re-rated to ~10.5× spot EV/EBITDA and ~20× trailing GAAP earnings (~99th percentile of its own history) — a clear premium to distribution peers Arrow (~7.4×) and Ingram Micro (~6.7×). The core debate: is Hyve a durable secular quality-upgrade that justifies a growth multiple, or a thin-margin (~3.3% of billings), customer-concentrated (one customer = 11% of FY25 revenue), working-capital-hungry, cyclical revenue-optics engine that lowers business quality even as it raises the growth rate? Our reading is that it is real and impressive but lower-quality than the core, and that the richest-ever multiple has already capitalized the optimistic case.

Our framework verdicts: a durable-but-thin competitive position (real scale and two-sided captivity, zero pricing power); a structurally mediocre industry (a rational oligopoly that clears but never beats its cost of capital); high financial quality with important caveats (lumpy, growth-inverse free cash flow and near-meaningless book metrics from acquisition intangibles); and disciplined capital allocation (consistent buybacks, growing dividend, sensible leverage). Valuation embeds continuation of the Hyve boom; the variant-perception question is whether the market has correctly priced a durable AI-infrastructure franchise or extrapolated a concentrated, cyclical growth spurt at the top of the capital cycle. No recommendation or price target appears below; Claude’s Take above carries the only position.

2. Business Overview

What TD SYNNEX does. TD SYNNEX is a technology-products distributor and solutions aggregator — the wholesale intermediary between the manufacturers of IT (OEMs and software publishers) and the ~150,000 resellers/VARs/MSPs/integrators who serve end customers. It does not (mostly) manufacture and does not (mostly) sell to end users. It buys at scale from 1,500-plus vendor lines, holds the inventory, extends the trade credit, configures and kits the product, and ships — taking a thin spread plus vendor rebates on enormous volume. In the company’s own framing it is the “aggregator” that lets a small reseller access the full breadth of the IT industry through a single relationship, and lets a vendor reach a long tail of resellers it could never service directly. The company traces to 1980, is dual-headquartered in Clearwater, Florida and Fremont, California, employs ~23,900 people, and operates in ~100+ countries. (Source: FY2025 10-K, Item 1; ROIC company profile.)

Two product portfolios; two businesses; three reporting segments. The taxonomy matters because management, product mix, and GAAP each slice the company differently:

• Product portfolios (how the industry describes the catalog): Endpoint Solutions — PCs, mobile phones, peripherals, printers, supplies (the “commodity” end); and Advanced Solutions — data-center technologies: servers, storage, networking, security, software, hybrid/hyperscale infrastructure (the higher-value, faster-growing end). In Q2 FY26, Endpoint distribution billings grew +13% (PC ASP-led, mid-single-digit units) and Advanced Solutions +31% (infrastructure and security).
• Businesses (how management now runs and discusses it): Distribution (gross billings $23.4B in Q2 FY26, +22%; ~1.9% operating margin on billings) and Hyve (gross billings $5.5B, +117%; ~3.3% operating margin on billings) — the hyperscale ODM arm, discussed in and .
• GAAP reportable segments (three, geographic): the Americas, Europe, and Asia-Pacific & Japan (APJ). FY2025 revenue split: Americas $36.2B (~58%, +4.0%), Europe $21.7B (~35%, +10.5%), APJ ~$4.6B (~7%). Europe is the faster-growing geography and, per management, a structural share-gain story against pure-play specialists.

How it makes money — and the gross-billings-vs-revenue nuance that dominates the numbers. SNX’s economics are a spread on volume: it earns a thin gross margin plus vendor incentive/rebate income, and its reported profit is what survives after ~4.7% of SG&A. Critically, management now headlines “gross billings” ($28.9B in Q2 FY26, +33% YoY) rather than GAAP revenue ($19.575B) — the two differ by a ~33% “gross-to-net adjustment” because a growing share of the mix (much software, cloud/SaaS, and certain agency arrangements) is recognized on a net (agent) basis, where SNX books only its margin, not the gross sale. This has two consequences the reader must hold onto: (1) reported GAAP revenue growth understates the volume growth of the business, which is why management guides on billings; and (2) the gross-margin percentage is partly a mix artifact — as more revenue is netted down, the denominator shrinks and the percentage optically rises independent of pricing. The honest read is on gross-profit dollars and operating-income dollars, both of which grew strongly (Q2 non-GAAP operating income +49%).

The float mechanics — the business is a balance sheet. The economic core of distribution is often missed: SNX earns its spread by intermediating working capital as much as product. It buys inventory from OEMs on payment terms, holds $9.5B of it, extends trade credit to ~150,000 resellers (carrying $12.7B of receivables), and funds the gap substantially with $17.6B of its own payables to vendors — a ~17–20-day cash-conversion cycle that turns ~18–20× a year. The scale leader can offer the deepest credit lines, the broadest catalog on one invoice, and the fastest fulfillment, all of which are funded advantages a sub-scale entrant cannot match. This is why the barrier to entry is the balance sheet, and why the returns are thin: SNX is compensated for taking credit, inventory, and logistics risk on enormous volume, not for any proprietary product.

PartnerFirst and the omnichannel motion. Management’s operational thesis is a “true omnichannel” model — digital self-serve (the PartnerFirst platform, increasingly layered with machine-learning/agentic-AI personalization, dashboards, and recommendations to lift attach rates and conversion) blended in real time with high-touch specialist selling, segmented by customer tier and reallocated dynamically. The claimed payoff is above-market SMB growth and share gains against pure-play specialists, most visibly in Europe (the fastest-growing geography at +10.5%). Whether this is a durable, quantifiable edge or standard distributor blocking-and-tackling is an open question — but the vendor-side validation (HPE naming SNX one of two global distributors) suggests the scale-plus-capability combination is real to the OEMs who choose their channel.

Recurring vs. non-recurring. SNX is not a subscription business. The overwhelming majority of revenue is transactional product resale tied to refresh cycles, project spend, and — increasingly — hyperscaler build-out programs. The stickier layers are the relationships (150,000 resellers embedded in SNX’s credit lines, portals, and configuration workflows), the growing software/cloud/security streams (double-digit growth, per management), and Hyve’s multi-year hyperscaler programs. Verdict: a well-run, globally-scaled, asset-light-ish intermediary whose reported revenue and gross-margin percentage are heavily shaped by net-revenue accounting; the real franchise is scale, two-sided relationships, and balance-sheet capacity — not recurring revenue.

3. Industry Dynamics

Structure — a consolidated three-player oligopoly. Broadline IT distribution has consolidated over two decades into an effective global oligopoly of three: TD SYNNEX, Ingram Micro, and Arrow’s enterprise-computing arm (ECS). SNX and Ingram Micro are the true broadline twins; together they represent roughly 40% of global broadline distribution revenue, with SNX holding an estimated ~15–18% global share. Below them sits a long tail of specialty and regional distributors (ScanSource in POS/barcode/comms; ALSO Holding and Esprinet in Europe). Total IT spend flowing through the channel is measured in the ~$1.5–1.8 trillion range; SNX’s $62.5B of GAAP revenue (~$80B of gross billings) is a mid-teens slice of the broadline segment. (Sources: SNX FY2025 10-K; Mordor Intelligence IT Hardware Distribution & Channel Market; Omdia/Canalys channel research, 2025.)

Why the two-tier structure persists (the real value-chain rationale). Both ends of the chain genuinely need the distributor, and this is what makes the industry structurally stable even if not lucrative:

• Credit and float. SNX extends trade credit to ~150,000 under-capitalized resellers and carries the paper — $12.7B of accounts receivable and $9.5B of inventory funded substantially by $17.6B of accounts payable. This is a working-capital intermediation machine no sub-scale entrant can fund; the balance sheet is the moat.
• Aggregation and breadth. One relationship gives a reseller access to 1,500+ vendor lines; one relationship gives a vendor access to a long tail of resellers.
• Logistics and configuration. Physical distribution, kitting, integration, and (via Hyve) full rack build.
• Demand aggregation for OEMs, who cannot economically service thousands of small resellers directly.

Critically, this role is more durable than the downstream reseller’s. The disintermediation threat that hangs over resellers like CDW — hyperscaler marketplaces (AWS/Azure/GCP) rerouting the cloud take-rate around the reseller — does not hit the distributor tier the same way, because the distributor’s function is credit, logistics, and physical aggregation, not the cloud agency fee.

Where the profit pool sits. The economics of the two-tier channel are a thin margin stack: the OEM captures the product margin, the distributor takes ~7% gross / ~2% operating on the physical/credit intermediation, and the reseller/VAR takes a further ~15–22% gross / ~5–7% operating on the customer relationship, integration, and services (CDW’s numbers). The distributor sits at the thinnest point of the stack by design — it is the highest-volume, lowest-touch, most-commoditized layer — which is why it is a business measured in turns and basis points rather than margins. The corollary is that a distributor cannot margin-expand its way to value creation the way a services-heavy reseller occasionally can; its levers are volume, share, working-capital efficiency, and mix (nudging toward higher-value Advanced Solutions and services). Hyve is the one move that breaks this pattern — by pushing up the value chain into design/manufacturing, SNX reaches for a fatter (~3.3%) slice — but it does so by trading the safe, wide-moat distribution layer for a contested, concentrated manufacturing one.

Structural profitability — why gross margins are ~7% and operating margins ~2%. Distributors are price-takers with no pricing power. They earn a thin spread plus discretionary vendor rebates on volume and pass OEM price through to resellers — as management put it plainly on the Q2 call, “we are a cost-plus business; if cost increases, we pass it to customers.” The consequence is a business whose returns clear, but never beat, the cost of capital: SNX ROIC ~8.9%, Ingram Micro ~7.3%, Arrow ~8.1%. The moat protects share and existence, not excess returns.

Capital cycle (Marathon lens). The distinction between the two capital cycles SNX straddles is the single most useful industry frame. In IT distribution, capital is disciplined and the supply side is rational: the three survivors all buy back stock and delever (Ingram cut net leverage 2.2×→1.0× post-IPO; SNX reduces its share count ~4%/yr), no one is adding capacity to chase share, and consolidation has removed rather than added competitors over two decades — the textbook late-consolidation, high-barrier, low-return equilibrium that is stable precisely because no one earns enough to attract new entrants. In the AI/hyperscaler end-market, by contrast, capital is flooding in at a historic rate, and this is where the Marathon warning bites: the boom pulls SNX into funding ever more inventory and receivables (its FY22 inventory build produced negative FCF; Q2 FY26 consumed $330M) against thin, concentrated margins, right as its own multiple peaks. High returns in the AI supply chain are attracting enormous capital — ODM capacity, hyperscaler in-sourcing, memory/GPU fabs — which is exactly the configuration that mean-reverts. SNX’s core is on the safe side of the capital cycle; its growth is on the dangerous side. Verdict: a structurally mediocre industry — a stable, rational, high-barrier oligopoly that earns roughly its cost of capital. Better than a fragmented commodity trade (share is stable, returns clear the bar), but with no pricing power and no path to compounding excess returns. Good enough to exist, not good enough to compound.

4. Competitive Position

The moat, named. In Greenwald’s taxonomy, TD SYNNEX’s advantage is scale economies combined with modest two-sided customer captivity — a cost/scale advantage plus network stickiness, not brand, not pricing power, and not a wide moat. The evidence:

• Scale economies. SNX is the largest (or co-largest) broadline distributor globally, with the vendor breadth (1,500+ lines), logistics density, and — decisively — the balance-sheet capacity to float $22B of receivables-plus-inventory that a sub-scale entrant simply cannot fund. Distribution is a fixed-cost-leverage-on-volume business; the scale leader spreads its logistics, IT, and credit infrastructure over the most volume.
• Two-sided captivity. On the reseller side, 150,000 customers are embedded in SNX’s credit lines, ordering portals (PartnerFirst), and configuration/enablement workflows — real switching friction for a small VAR whose financing and fulfillment run through SNX. On the vendor side, top-tier certifications and the recent HPE decision to name SNX one of just two global distributors for its full networking/cloud/AI portfolio (including Juniper) show OEMs increasingly rationalizing to a few global-scale distributors — a dynamic that structurally favors the largest players.

Pressure-testing the moat — where it is thin. Three facts keep this “narrow,” not “wide.” First, zero pricing power: a 7% gross margin and a cost-plus model are the direct proof that the advantage is durability of share and turns, not the ability to price. Second, the excess-returns test fails: ROIC ~8.9% ≈ WACC means the moat protects existence, not value creation — the hallmark of a good operator in a structurally capped industry. Third, vendor and customer concentration SNX does not control: Apple is 11% of revenue and HP 10% on the vendor side (their rebate programs are discretionary and short-term, a gross-margin lever SNX doesn’t own), and one customer is now 11% of revenue on the demand side (Hyve-driven — see below).

Versus peers (numbers). SNX and Ingram Micro are the broadline twins; SNX is both the scale leader and the margin leader among pure broadline distributors:

*Arrow’s 11.2% blended gross margin is a components-mix artifact — ~70% of Arrow is electronic-components distribution (structurally higher margin); its true comparable is its ECS arm (~$9.4B, +18% YoY, AI-driven). WESCO shows that IT distribution is the thinnest-margin distribution niche — industrial/electrical distribution earns ~3× the gross margin.

A different competitive front for Hyve. One nuance the peer table understates: Hyve does not primarily compete with distributors — it competes with contract manufacturers and ODMs (Foxconn/Hon Hai, Wiwynn, Quanta, Celestica, Jabil, Sanmina) for hyperscaler server/rack programs. This is a genuinely different, more contestable arena than broadline distribution: hyperscalers deliberately multi-source their ODMs, design much of the hardware themselves, and can in-source or squeeze on price. SNX’s edge in Hyve is its end-to-end offering (design/co-design through board manufacturing, full rack integration, and supply-chain services) and long-standing relationships (the AWS tie that produced the equity warrant) — real, but not the near-unassailable scale-plus-float barrier that protects core distribution. That SNX’s closest statistical factor cousin is Sanmina (an EMS/ODM) is the market implicitly recognizing that Hyve pulls SNX toward a lower-moat, more-cyclical business model even as it raises the growth rate.

The Greenwald market-share-stability test is passed for core distribution — SNX has held or gained share as a scale consolidator. The ROIC test is the more telling: ~8.9% GAAP (~10.5% adjusted) on a goodwill-heavy base clears the cost of capital and is the hard evidence the advantage is real, but it does not dramatically beat it. Verdict: a durable-but-thin competitive advantage — genuine scale, two-sided captivity, and balance-sheet capacity in core distribution (harder to disintermediate than the downstream reseller’s position), wrapped around a lower-moat, more-contestable ODM business in Hyve. SNX is the strongest player in a structurally weak distribution business, extending into a faster-growing but less-defensible manufacturing adjacency.

5. Growth History and Forward Opportunities

History — a merger, then four flat years, then Hyve. The revenue trajectory is dominated by the Tech Data merger (September 2021), which roughly doubled SYNNEX from a ~$20B (FY2020) to a ~$60B distributor. Post-merger, organic growth was minimal for three years — revenue was $62.3B (FY22), $57.6B (FY23), $58.5B (FY24) — as a post-COVID IT-demand air-pocket and net-revenue mix shift held the top line flat-to-down. FY2025 broke the pattern with +6.9% to $62.5B, and FY2026 has accelerated sharply: Q1 revenue +18%, Q2 +16% GAAP (gross billings +33%). The acceleration is not broad-based demand recovery alone — it is Hyve plus component-cost-driven ASP inflation (rising memory and CPU prices lift hardware dollars even at flat units).

The growth engine — Hyve and the AI build-out. Hyve is SNX’s hyperscale custom-server ODM business: it co-designs and manufactures build-to-order servers, racks, and networking directly for hyperscalers, plus supply-chain services. The numbers are extraordinary and are the reason the stock re-rated:

• Q2 FY26 Hyve gross billings +117% YoY to $5.5B (~28% of quarterly revenue), non-GAAP operating income +89% to $181M, at a ~3.3% operating margin on billings (down 50bps YoY on mix/ramp inefficiency). Manufacturing was ~2/3 of Hyve, supply-chain services ~1/3.
• SNX now has programs with each of the five largest U.S. hyperscalers; three have more than one program; the third is in early ramp, and the fourth and fifth are expected to ramp in late FY26 / early FY27 — explicit optionality that the current Q3 guide excludes (“assumes no material contribution from newly onboarded customers”).
• SNX is expanding manufacturing footprint by >1M square feet in the U.S. and issued an equity warrant to Amazon (a longstanding Hyve/AWS customer) that grows in value as programs expand — a customer-alignment tool and a concentration tell.

The two businesses now look like this (Q2 FY26, on a gross-billings basis):

The table captures the whole investment debate in six rows: the ~81% of the business that is a durable, wide-moat, cost-of-capital grinder is growing a respectable +22% (itself flattered by ASP inflation), while the ~19% that is driving the narrative and the multiple is growing +117% but is structurally lower-quality. The blended growth rate is spectacular; the blended quality is diluting.

The ASP-inflation kicker — real dollars, but a quality question. A material part of FY26’s revenue acceleration is price, not units: memory and CPU costs are rising sharply, and because SNX is a cost-plus pass-through, higher component costs flow straight into higher hardware ASPs and therefore higher revenue dollars at a roughly constant percentage margin. Management confirmed PCs grew on ASPs with only mid-single-digit unit growth, that storage and servers are the most price-inflated categories (memory-driven), and that further price increases were expected in July. This is a double-edged sword: it flatters near-term revenue and gross-profit dollars, but it is (a) not volume growth, (b) potentially demand-destructive if it persists (management is watching consumer-PC elasticity), and © reversible when component costs normalize. A quality-minded reader discounts the ASP-driven portion of the growth as lower-durability than the genuine unit/share/mix gains.

Forward opportunities beyond Hyve. (1) AI-capable device refresh in Endpoint (management believes “the shift to AI-capable devices is just beginning,” plus a Windows-refresh tailwind and B2B PC cycle). (2) Advanced Solutions breadth — storage (called out as the “next” AI wave after compute), networking (Wi-Fi 7 refresh), security and software/cloud (double-digit growth). (3) Vendor go-to-market rationalization — the HPE global-distribution win, which management expects to be “more to come,” structurally favors the largest global distributor. (4) Geographic expansion in APJ (+15.2% in FY25), Latin America, and potentially the Middle East. (5) The two un-ramped hyperscaler programs — Hyve now has programs with all five major U.S. hyperscalers, but Q3 guidance explicitly excludes material contribution from the two newest, which are expected to ramp in late FY26 / early FY27 — genuine, quantifiable optionality that is not in the current run-rate.

Does Hyve improve the economics, or just the optics? This is the question on which the whole thesis turns, and the honest answer is both, partially. On the positive side, Hyve is genuinely accretive at the operating line: its ~3.3%-of-billings operating margin exceeds the corporate ~2.6%, because an ODM carries minimal SG&A relative to its revenue — so Hyve raises consolidated operating margin even as it drags the gross-margin percentage (Q2 blended GM fell to 6.84% as Hyve mix grew). On the negative side, the return on capital is the real test, and it is unflattering: Hyve ties up large inventory and receivables against a handful of hyperscaler counterparties, so a respectable operating margin can still translate into a mediocre ROIC once the working capital is charged against it — and SNX does not disclose Hyve’s stand-alone ROIC (a key open question, potentially resolved if Hyve becomes a fourth reportable segment). The concentration is stark: one customer is 11% of total company revenue, and Hyve’s entire book sits across only five hyperscalers who design their own hardware, multi-source their ODMs, and can in-source at will. The Amazon equity warrant is a two-sided signal — it aligns a key customer, but it also underscores how much leverage the hyperscalers hold over their suppliers. Net: Hyve raises the growth rate and is modestly accretive to reported margin, but it lowers business quality — more concentration, more working-capital intensity, more cyclicality, and a weaker moat — which is precisely why capitalizing it at a richest-ever multiple is the core risk.

Quality of the growth — the crux. The honest read: the core distribution growth is decent-quality (share gains, margin-accretive Advanced Solutions mix, real double-digit software/cloud), but the headline growth is dominated by Hyve, which is lower-quality, and by ASP inflation, which is transitory. Management’s own #1 stated priority — “growing operating profit faster than billings” — is the right frame and an implicit acknowledgment that billings growth alone is low-quality. Verdict: high-rate growth of mixed quality — a genuinely strong core-distribution franchise gaining share, wrapped around a spectacular but thin-margin, concentrated, cyclical AI-ODM spurt (and a slug of reversible price inflation) that the market is capitalizing as though it were durable and high-quality.

6. Financial Quality

Income statement — thin, stable, and improving at the margin. The FY2025 shape is textbook distribution: revenue $62.5B, gross profit $4.37B (GM 6.99%), SG&A $2.95B, operating income $1.42B (OM 2.28%), net income $827.7M (net margin 1.32%), GAAP diluted EPS $10.04. The multi-year record shows remarkable margin stability (GM 6.0–7.0%, OM 2.0–2.3% every year 2020–25) with a modest recent uptick as Advanced Solutions and Hyve mix improved the blend. The economics do improve slightly with scale/mix — operating margin rose from ~2.04% (FY22) to ~2.28% (FY25), and incremental operating margins have turned positive (~3.9% FY25) — but this is a low-margin business by nature, and the “improvement” is measured in tens of basis points.

The six-year record (November fiscal years):

*FY2021 is a stub for the Tech Data revenue (merger closed Sep-1-2021), which is why revenue doubles FY21→FY22. **FY2020 GAAP EPS includes a $194.6M discontinued-operations item (the Concentrix spin); continuing-ops EPS was ~$6.57. The clean read across the table: revenue roughly flat 2022–24 then re-accelerating; margins grinding up ~25bps over five years; FCF swinging with the working-capital cycle (negative in the FY22 build year); ROIC creeping from 6% toward 9%; and the diluted share count falling every year since the merger peak. This is the financial signature of a well-run, low-margin, cyclical compounder-of-the-count — not a high-return franchise.

GAAP vs. non-GAAP. The gap is material and legitimate to understand: SNX’s non-GAAP figures exclude amortization of acquisition intangibles — a large, recurring, non-cash charge from the Tech Data purchase of $296.3M in FY2025 (~$3.56/share); the balance sheet still carries $3.8B of intangibles plus $4.1B of goodwill — plus stock-based compensation and (now small) acquisition/integration costs. The bridge is meaningful: FY2025 GAAP diluted EPS $10.04 versus non-GAAP diluted EPS $13.19 (FY24 $11.68); GAAP operating margin 2.26% versus non-GAAP 2.86%. In Q2 FY26, GAAP EPS was $4.15 versus non-GAAP $4.85. Unlike a software company where non-GAAP flatters by excluding real SBC dilution, here the dominant add-back (intangible amortization) is genuinely non-cash and arguably economic to exclude — but the reader should keep GAAP in view for discipline and note that SBC (~$66M) is small and not an FCF distorter (unlike the SaaS names). Note also that management’s incentive comp is struck on non-GAAP net income, which the intangible add-back flatters.

Returns — read ROIC, not ROE. FY2025 ROE was a headline-grabbing 26.7%, but this is a thin-equity leverage artifact: tangible book value is only ~$7/share (it was negative in 2021–22 post-merger), so a modest net income divided by a thin, goodwill-eroded equity base produces a big ROE. The honest metrics are GAAP ROIC ~8.9% (up from ~8.3% in FY24) and company-reported adjusted ROIC ~10.5% (which adds back the intangible amortization and integration costs). Read together, they say the same thing: SNX earns roughly its cost of capital, with at most a modest positive spread — a good operator that clears the bar but does not dramatically beat it. Return on assets is ~2.6%.

Free cash flow — real, but lumpy and inverse to growth. This is the single most important nuance for a distributor. In normal/declining-growth years SNX gushes cash (FY25 operating cash flow $1.53B, capex just $142M, FCF ~$1.39B; FY23 $1.26B; FY20 $1.64B). But in growth years, working capital consumes cash: FY2022 FCF was negative −$167M on an inventory build, and in Q2 FY26 the company consumed ~$330M of free cash as Hyve’s hypergrowth pulled inventory and receivables higher. Management’s long-term “north star” is ~95% net-income-to-FCF conversion, explicitly caveated that hypergrowth periods consume cash and downturns release it. This countercyclical cash profile is a genuine feature — a downturn would liquidate working capital into a cash flood — but it means FCF-based valuation must be normalized across the cycle, and current-year FCF will look poor precisely because the business is booming.

Balance sheet — moderate leverage, giant working-capital float, meaningless book value. Net debt was $2.18B at FY25-end (~1.2× EBITDA) and net leverage was 1.6× at Q2 FY26 (management’s medium-term framework), below target; cash fell to $1.1B as WC grew. Debt is investment-grade. The working-capital machine is the balance sheet: $12.7B receivables and $9.5B inventory largely funded by $17.6B payables, a ~17–20-day cash-conversion cycle, and — new this quarter — a €650M European receivables-securitization program (BNP Paribas/Santander/Crédit Agricole) that provides cheap off-revolver funding for the WC growth. Book-value metrics are near-useless (tangible equity is a rounding error against $7.9B of goodwill+intangibles), which is why P/B and P/TBV should be ignored in favor of EV/EBITDA and normalized FCF. Verdict: high financial quality for a distributor — stable margins, real (if lumpy) cash generation, sensible leverage, and a well-managed WC engine — with two honest caveats: FCF is growth-inverse and must be normalized, and book/ROE metrics are distorted by acquisition accounting and must be replaced by ROIC.

7. Capital Allocation

The record — disciplined and shareholder-friendly. TD SYNNEX has been a consistent, unshowy allocator. Since the merger it has returned cash steadily through buybacks and a growing dividend. The current authorization is a $2.0B program (approved March 2024, no expiry), up from a $1.0B program in January 2023; the company repurchased 4.4M shares for $596M in FY25 (5.5M/$612M in FY24) and 1.095M shares at an average $175.39 in Q2 FY26, leaving ~$1.0B remaining at May-31-2026. The dividend has grown every year — $0.408 (FY20) → $1.76 (FY25) → raised ~9% to $0.48/quarter (~$1.92 annualized) in January 2026 (payout ~13–18% of earnings, yield ~0.8%). The buyback has done real work: diluted share count has fallen from ~95.5M (FY21) to ~79.5M (Q2 FY26), ~4%/year — meaningful per-share compounding for a business whose absolute earnings grow slowly. Leverage is ~1.6× (below the medium-term framework); total borrowings were ~$4.6B at FY25-end; the April-2024 issuance of $600M 6.100% senior notes due 2034 refinanced $700M of 1.250% notes maturing August 2024 (a refinancing, not incremental leverage). The balance sheet is investment-grade.

Reinvestment and M&A. Capex is trivial (~$142M FY25, ~0.2% of revenue) — distribution is asset-light. The reinvestment story now is Hyve working capital and footprint (~$100M/year of Hyve capex, amortized 5–6 years, ladders leases, and — management stresses — built only against committed long-term programs, not speculative capacity). M&A since the transformational Tech Data deal has been small and tuck-in: Apptium (July 1, 2025, ~$105M, cloud-commerce orchestration for the StreamOne platform) and IPsense (December 2024, Brazil cloud-migration) — capability adds, not scale bets, with no standalone deal 8-Ks. This is appropriate: the industry does not reward roll-up empire-building, and management has resisted it.

Incentive alignment — good, with one watch-item. The FY2025 proxy (DEF 14A, February 2026) shows genuinely sensible metrics for a thin-margin distributor: the CEO’s annual incentive is 70% worldwide non-GAAP net income + 30% adjusted ROIC / return-on-working-capital, with segment leaders additionally on regional return-on-working-capital and contribution margin. There is no gross-billings- or revenue-based bonus metric — capital efficiency is explicitly embedded, exactly right for a business where revenue growth is often low-quality. Long-term incentives are 3-year cliff-vesting PRSUs (zero if the minimum is missed) plus time-based RSUs. CEO Patrick Zammit’s FY2025 total compensation was ~$10.3M (~250× median); say-on-pay passed comfortably. The one watch-item: incentives keyed to non-GAAP net income, which adds back the $296M of intangible amortization — defensible but flattering versus GAAP.

Ownership and insider behavior — the “overhang” is largely resolved. The legacy control blocks are essentially gone: Apollo (via Tiger Parent Holdings), which received ~44% at the 2021 merger, fully exited by April 2024 through a series of registered secondaries each paired with a company concurrent buyback, and reported 0.0% of class in its April-2024 13D/A; the founding MiTAC/Miau family is now under 5% (~4.3%) and trimming. Today’s largest holders are index funds (Vanguard ~10%, BlackRock ~8.5%, Fidelity ~6%); directors and officers own ~0.7%. Insider transactions over the last ~18 months show zero open-market purchases (no code-P buys) against routine option-exercise-and-sell and tax-withholding activity, with the retired CEO and one director the largest net sellers — a mild negative/neutral signal, but consistent with normal comp monetization rather than a conviction tell. Verdict: management has allocated capital intelligently — a disciplined $2.0B buyback that shrinks the count ~4%/year, a sensibly growing dividend, trivial-but-well-targeted reinvestment, restraint on M&A, and ROIC-embedded incentives — all within the constraint that a cost-of-capital business cannot buy its way to excess returns. The private-equity overhang that shadowed the stock for three years is now essentially cleared, which removes a long-standing technical headwind.

8. Changes and Headwinds — Last Two Years

Strategic and leadership changes. The defining shift is the Hyve inflection and the pivot to headlining “gross billings” and the Distribution-vs-Hyve business framing — a deliberate repositioning of SNX from a boring broadline distributor to an AI-infrastructure participant. Alongside it: the CEO succession — Patrick Zammit (ex-COO, ex-Avnet/Tech Data) became President & CEO effective September 1, 2024, with Rich Hume retiring to the board; the CFO change — David Jordan became CFO effective October 2, 2025 (succeeding Marshall Witt); the HPE global-distribution win (a structural vendor-rationalization gain ramping into FY27); tuck-in M&A (Apptium ~$105M July 2025, IPsense Dec 2024); the new €650M EU receivables-securitization program; the near-complete Apollo ownership exit (0% by April 2024); and the Amazon equity warrant tied to Hyve program expansion (whose underlying shares were registered for resale in June 2026 — the “secondary” that superficially resembled, but was not, a sponsor block).

Context — a completed integration and a strategic repositioning. The two years also mark the effective completion of the Tech Data integration (acquisition/integration/restructuring costs collapsed to $7.2M in FY25 from $71.3M, and the merger synergies are now in the base) and, with it, a deliberate repositioning of the equity story. Management has moved the narrative from “the world’s largest broadline distributor” toward “the essential AI-infrastructure channel,” headlining gross billings and the Distribution-vs-Hyve framing, standing up dedicated Hyve capacity and (prospectively) segment disclosure, and courting the AI-infrastructure investor base. This is a legitimate reflection of where the growth is — but it is also a narrative shift that has accompanied, and arguably helped drive, the multiple re-rating, and a skeptical reader should separate the genuine operating inflection from the storytelling that has monetized it. The earlier 2020 spin of Concentrix (CNXC) already streamlined SNX into a pure-play IT-products distributor; Hyve is the second act of that focus, pushing the company up the value chain into manufacturing.

Tailwinds. The AI-data-center capex super-cycle (Hyve +117%); component-cost-driven ASP inflation lifting hardware dollars; a nascent AI-PC/Windows refresh; storage and networking (Wi-Fi 7) refresh cycles; and double-digit software/cloud/security growth. Management reports no demand destruction yet from rising prices and characterizes underlying demand as solid.

Headwinds and watch-items. (1) Component supply constraints — memory and certain CPUs; management explicitly built supply risk into the Q3 guide. (2) Working-capital drain — Hyve hypergrowth is consuming cash now (Q2 FCF −$330M). (3) Customer concentration — one customer at 11% of revenue, plus the hyperscaler-concentration risk inside Hyve. (4) Tariffs and FX — global hardware imports and large Europe/LatAm exposure. (5) The selling-shareholder overhang and post-blowout valuation. (6) A one-time margin flatterer — management flagged a 5–10bps benefit in Q2 distribution margin from strategic inventory purchasing ahead of price increases, explicitly called out as dwindling — a small, honest quality-of-earnings note. Verdict: the last two years have genuinely strengthened the growth profile and the vendor position (Hyve, HPE) while lowering earnings quality (concentration, WC intensity, cyclicality) — a net thesis-changer that cuts both ways and is the source of the entire valuation debate.

9. Risk Analysis (Risk Matrix)

The dominant risks are concentrated on the AI/Hyve growth engine and the valuation that capitalizes it — the same feature that drove the re-rating is the largest source of downside, and the three top-row risks (capex air-pocket, customer concentration, valuation de-rating) are correlated: a hyperscaler capex slowdown would simultaneously stall Hyve growth, expose the customer concentration, and trigger the multiple compression, so they would likely arrive together rather than independently. The historical evidence that a distributor de-rating is fast and correlated with an earnings wobble is on the tape: the stock fell ~25% in March–April 2025 on a single soft quarter plus a tariff shock. The mitigants are real and worth weighting: the balance sheet is investment-grade, the core distribution business is diversified across geographies and ~150,000 customers, and — crucially — the cash generation is countercyclical, so a downturn that cut Hyve volume would release working capital into a cash flood (the mirror image of the current $330M/quarter consumption), cushioning the equity even as earnings fell. A catastrophic or total loss is therefore remote; the realistic downside is a 50%-class de-rating if Hyve growth stalls — the multiple compressing toward the peer ~7× EV/EBITDA while working-capital drag and a lower earnings base compound — not an impairment of the enterprise itself. The asymmetry a buyer at $244 faces is that most of the identifiable upside (the two un-ramped hyperscaler programs, HPE) is already the consensus narrative and partly in the multiple, while the downside (capex normalization) is a when-not-if feature of any capex super-cycle.

10. Valuation Discussion (Embedded Expectations)

Where the multiple sits — and the three honest ways to read it. At $244.64 (~79.5M diluted shares, ~$19.5B market cap; add ~$3.5B net debt for ~$23B enterprise value):

• Trailing GAAP P/E ~20× through Q1 FY26 (TTM GAAP EPS $12.10) — the 99th percentile of SNX’s own decade history (P/S 99th, P/B 92nd, composite ~97th). This is the “richest-ever” headline, and it is true on trailing GAAP.
• Trailing GAAP P/E ~17.4× including the Q2 blowout (TTM GAAP diluted EPS ~$14.07) — still historically rich (own history ~7–16×).
• Spot EV/EBITDA ~10.5× (TTM EBITDA ~$2.2B) versus a five-year average of ~7.2× and an FY25-close 8.0× — a ~45% premium to its own history and a large premium to distribution peers (Arrow 7.4×, Ingram 6.7×). This is the cleanest single read and the one the bear should anchor on.
• Forward non-GAAP P/E ~13× (FY26 non-GAAP EPS ~$18, on Q2 $4.85 + Q3 guide $4.50 + a ~$18–19 full-year run-rate) — not extreme, roughly in line with history and only a modest premium to peers. This is the bull’s anchor: because earnings are inflecting fast, the forward multiple is far below the trailing one.

The tension between the trailing-99th-percentile and the forward-13× reads is the valuation debate: the multiple is only “cheap” if the Hyve-driven earnings surge is durable and repeatable, and only “richest-ever expensive” if it is a peak.

Embedded-expectations analysis — what the price requires. At ~10.5× spot EV/EBITDA for a business that has historically commanded ~7–8×, the market is capitalizing ~3 turns of multiple premium on the thesis that Hyve permanently upgrades SNX’s growth and quality. To justify $244 on a normalized basis, one must underwrite something like: EBITDA growing from ~$2.2B toward ~$2.8–3.0B over 2–3 years (Hyve programs ramping, all five hyperscalers live, HPE contributing), a multiple that holds at ~9–10× rather than reverting to the peer 7×, and the buyback continuing to shrink the count ~4%/year. That is a coherent bull path — but it requires the AI capex cycle to persist, Hyve concentration to not bite, and working capital to normalize into cash. The market is, in effect, underwriting the optimistic case as the base case.

Scenario sketch (illustrative, not a target):

• Bear: Hyve growth stalls / a top-customer program is lost; EBITDA reverts toward ~$1.8–2.0B and the multiple de-rates to the peer ~7× → EV ~$13–14B → equity value well below the current price. The 2025-style −25% air-pockets (Mar–Apr 2025) show the downside is real and fast, and a distributor de-rating tends to be violent because the earnings themselves wobble at the same time.
• Base: core distribution compounds low-single-digits, Hyve grows but decelerates and stays ~3%-margin; normalized EBITDA ~$2.3–2.5B at ~8–8.5× → EV ~$19–21B → roughly the low-$200s, i.e., modest downside-to-flat from $244.
• Bull: all five hyperscaler programs ramp (the guide excludes the two new ones), HPE contributes, Hyve holds >3% margin and diversifies; EBITDA ~$2.8–3.0B at ~9–10× → EV ~$26–30B → a further leg up.

Cross-checking against peers. The peer multiples make the premium concrete: Arrow trades ~7.4× EV/EBITDA and ~12× earnings, Ingram Micro ~6.7× and ~16×, ScanSource ~6.4× and ~11× — a distribution cohort clustered at 6–7× EV/EBITDA and low-teens P/E. SNX at ~10.5× spot EV/EBITDA and ~17–20× trailing GAAP P/E carries a ~50% EV/EBITDA premium to the group. That premium is entirely a bet on Hyve/AI: on the core distribution business alone, SNX would be a peer-multiple stock, i.e., worth something closer to ~$150–180. The market is paying an incremental ~$65–95/share of equity value for the proposition that Hyve is a durable, secular growth-and-quality upgrade rather than a cyclical, concentrated ODM spurt. The dividend (~0.8% yield) and normalized FCF yield (~6–7% through-cycle, but near-zero in the current hypergrowth year) are too modest to underpin the valuation — this is a growth-multiple stock now, not an income or deep-value one, which is itself the tell that the character of the investment has changed at these levels.

What the market is pricing correctly vs. incorrectly. Correctly: that Hyve is real, that SNX is the scale-and-margin leader of a stable oligopoly, that the buyback compounds, and that the distributor tier is more durable than the reseller tier. Potentially incorrectly: that a thin-margin, one-customer-11%, working-capital-hungry, cyclical AI-ODM business deserves a durable-growth multiple, and that the trailing earnings surge (part ASP inflation, part strategic-inventory flatter, part program-timing) is a clean run-rate. No price target; embedded-expectations framing only.

11. Variant Perception

Consensus. The sell side is constructive and chasing: a wave of post-Q2 price-target raises (Morgan Stanley to $374, UBS to $352, Barrington $325, Barclays $278, Goldman/RBC/UBS all raising), framing SNX as the best-positioned distribution proxy for the AI-infrastructure build-out, with Hyve as a durable, under-appreciated growth engine and the core distribution business as a stable, share-gaining, cash-generative base. The news tape skews positive.

The strongest bull case. SNX is a scale-advantaged, cash-generative oligopolist that has permanently changed its growth profile: Hyve makes it the trusted ODM partner to every major U.S. hyperscaler at the dawn of a multi-year, multi-hundred-billion-dollar AI-data-center capex cycle, with two more hyperscaler programs (excluded from guidance) ramping into FY27 and an HPE global-distribution win layering on. On forward non-GAAP the stock is only ~13× — cheap for a business now growing operating profit at 30–50% — and the ~4%/year buyback plus a growing dividend compound the per-share math. The distributor tier is structurally insulated from the disintermediation that threatens resellers.

The strongest bear case. The re-rating to a richest-ever ~10.5× EV/EBITDA / ~20× trailing GAAP multiple capitalizes an AI spurt that is lower-quality than the core business it is being credited to: Hyve is ~3%-margin, has one customer at 11% of total revenue, consumes cash (Q2 FCF −$330M), and rides a hyperscaler capex cycle that will not compound forever and that the hyperscalers themselves can dual-source, in-source, or squeeze. The trailing earnings surge is partly ASP inflation (memory/CPU cost pass-through), partly a self-flagged one-time strategic-inventory margin benefit, and partly program timing — not a clean run-rate. Strip the AI narrative and you have a fine ~8.9%-GAAP-ROIC (~10.5% adjusted) distributor that has always been worth ~7–8× EV/EBITDA — and the stock is ~45% above that. (Note the bear case does not rest on a supply overhang: the legacy Apollo/MiTAC blocks are essentially gone — that long-standing technical headwind has cleared, which is a point for the bull.)

The factor-positioning read (evidence for where consensus may be offsides). The tape shows an extreme one-way momentum run — +78% in a year, Sharpe 2.5, a trivial 16% max drawdown — but SNX’s factor DNA is still mid-cap value + dividend-yield + small-size, its closest statistical cousin is Sanmina (an EMS/ODM), and it carries a negative loading on the Robotics-&-AI factor. The market is re-rating a value name faster than its factor identity has shifted; the realized momentum has outrun the underlying, which is exactly the configuration that mean-reverts when the momentum stalls. The −17.5%-off-ATH pullback plus the secondary is the first crack in the one-way street.

The 3–5 assumptions that matter most, and what would falsify each:

1. Hyve growth is durable, not a peak — falsified by a hyperscaler capex slowdown or a sequential Hyve billings decline.
2. Hyve concentration is manageable — falsified by loss/step-down of the 11%-of-revenue customer or a single Hyve program.
3. The multiple holds above the peer ~7× — falsified by two quarters of decelerating billings + a de-rate toward peers.
4. Earnings quality is clean — falsified by margin give-back as strategic-inventory and ASP-inflation tailwinds fade.
5. The buyback keeps compounding — falsified if WC drain forces capital toward funding growth instead of repurchases.

12. Fact vs. Interpretation Table

13. Open Questions

1. Will Hyve become a fourth reportable segment? The Q2 FY26 10-Q’s XBRL introduces a distinct “Hyve Solutions Segment” member alongside the three geographies — if formalized, it would finally give investors Hyve’s stand-alone economics (and its return on capital).
2. What is Hyve’s actual return on capital? Operating margin is disclosed (~3.3% of billings) but the working-capital intensity means ROIC — not margin — is the real quality test, and it is not separately disclosed.
3. How concentrated is Hyve within the 11%-of-revenue customer? Is the unnamed 11% customer a single hyperscaler, and how much of Hyve depends on it?
4. Durability of the Amazon warrant economics — what commercial cost/dilution does the performance-vesting warrant embed as programs scale?
5. Normalized through-cycle FCF — what does FCF look like once Hyve’s WC build stabilizes and growth normalizes?
6. How much of recent gross-margin strength is ASP-inflation and strategic-inventory (transitory) vs. mix (durable)?
7. The Endpoint vs. Advanced Solutions dollar split is disclosed only in IR decks, not the filings — a transparency gap for modeling mix.

14. What Must Be True

For the bull case to be right (and the stock to work from here):

• Hyve sustains >30% growth and holds a >3% operating margin as the third, fourth, and fifth hyperscaler programs ramp — durable, diversifying AI-ODM franchise.
• Customer concentration falls as programs diversify (the 11% customer becomes a smaller share of a bigger Hyve).
• EBITDA compounds toward ~$2.8–3.0B and the multiple holds at ~9–10× rather than reverting to peers.
• Falsification test: two consecutive quarters of decelerating Hyve gross billings, or a disclosed loss/step-down of a hyperscaler program, or Hyve operating margin falling below ~2.5% — any of these breaks the durable-quality thesis and should re-rate the stock toward the peer ~7× EV/EBITDA.

For the bear case to be right:

• The AI capex cycle air-pockets (as hyperscaler capex guidance softens), Hyve growth stalls, and SNX is left carrying working-capital drag into a de-rating.
• The trailing earnings surge proves partly transitory (ASP inflation and strategic-inventory benefits fade), and margins give back.
• Falsification test: Hyve sustains triple-digit-then-strong-double-digit growth for another 3–4 quarters at a stable ~3%+ margin while the customer base diversifies, and the multiple holds — which would prove the re-rating was an earned quality-upgrade, not a peak-cycle extrapolation, and would invalidate the “richest-ever, cost-of-capital business” bear frame.

15. Source Appendix

The full, dated, categorized source list follows in Appendix B below. Primary sources: TD SYNNEX FY2025 Form 10-K (filed Jan 2026) and prior 10-Ks/10-Qs; Q2 FY26 earnings release, 8-K, and call transcript (June 25, 2026); the June-29-2026 S-3 prospectus supplement and related 8-Ks; and the FY2025 DEF 14A. Quantitative data reconciled via third-party fundamental-data aggregators, own-history valuation-percentile data, and a public quantitative factor model. Third-party industry framing from Mordor Intelligence and Omdia/Canalys channel research. All management commentary is treated as hypothesis and validated against filings and external data.

APPENDIX A — Standard Diligence Questionnaire

TD SYNNEX Corporation (NYSE: SNX) — as of 2026-07-04

Answers are grounded in the public record and labeled Fact / Interpretation / Assumption where it matters. Where a question does not map cleanly to a distributor’s model, the correct sector analog is given.

General

What thoughtful questions have other investors asked about this company? The recurring institutional questions cluster around: (1) Is Hyve a durable secular franchise or a cyclical, low-margin AI pass-through? — the single most-debated point, and the driver of the entire re-rating; (2) What is Hyve’s actual return on capital? (operating margin is disclosed, ROIC is not); (3) How concentrated is the business in a single hyperscaler? (one customer = 11% of FY25 revenue); (4) Why is free cash flow negative in a boom year? (working capital funds growth); (5) Can the buyback continue if WC keeps absorbing cash?; and (6) the now-resolved When will the Apollo/MiTAC overhang clear? (answer: it already has). Sell-side attention has shifted almost entirely from the boring distribution base to Hyve/AI.

Cyclicality & Earnings Nature

Are earnings at a cyclical high or low? Interpretation: a cyclical/secular high. Revenue and operating income are inflecting sharply (Q2 FY26 non-GAAP operating income +49%) on the AI-data-center capex super-cycle plus component-cost-driven ASP inflation — both of which are elevated relative to a normalized environment. The core distribution business is mid-cycle-to-strong; Hyve is at a hypergrowth peak that is unlikely to compound at +117% indefinitely.

Driven by the external environment or internal actions? Both, but predominantly external — the AI capex wave and hardware ASP inflation are exogenous tailwinds. Internal actions (Hyve program wins, the HPE global-distribution award, share-gain execution in Europe) are real and additive but ride the external cycle.

How stable are revenues? Fact: the core distribution revenue is stable-to-slowly-growing (flat for three years post-merger, +6.9% FY25) with countercyclical cash generation. Hyve revenue is lumpy (multi-year program timing) and cyclical (hyperscaler capex). Blended, revenue is now more volatile than the historical distributor profile.

Outlook for products/services? Positive near-term across most categories per management: AI-capable PC refresh “just beginning,” storage/networking (Wi-Fi 7) refresh, double-digit software/cloud/security, and hyperscale infrastructure. The one watch-item is consumer-PC elasticity to ASP increases (SNX is weighted to B2B, mitigating it).

How big will this market be — growing, shrinking, domestic or international? The addressable IT channel (~$1.5–1.8T of spend) is large and low-single-digit growing; the AI-infrastructure sub-segment SNX rides via Hyve is growing far faster but is cyclical. SNX is genuinely global (~58% Americas, ~35% Europe, ~7% APJ; APJ growing fastest at +15%).

Business Quality & Competitive Moat

Is the industry getting more or less competitive? Interpretation: more concentrated, not more competitive. Broadline distribution has consolidated to a three-player oligopoly (SNX, Ingram Micro, Arrow ECS), and vendors are rationalizing to fewer global distributors (the HPE win) — structurally favoring the largest scale players.

How profitable is the business (ROIC, ROE)? Fact: GAAP ROIC ~8.9%, adjusted ROIC ~10.5% — roughly the cost of capital with at most a modest positive spread. ROE 26.7% is a thin-equity leverage artifact and should be ignored in favor of ROIC.

How profitable is the industry — competitors, barriers to entry? Structurally low-margin (7% gross, 2% operating) because distributors are price-takers. Barriers to entry are nonetheless high: scale economies, vendor relationships/certifications, and — decisively — the working-capital capacity to float $22B of receivables-plus-inventory. New entry at scale is effectively impossible; the barrier protects share, not margin.

Can the business be easily understood? Yes at the core (buy wholesale, sell to resellers on a thin spread), with two complications the reader must master: the gross-billings-vs-GAAP-revenue distinction and Hyve’s ODM economics.

Can it be undermined by foreign low-cost labor? Not directly — the moat is logistics, credit, and vendor breadth, not labor cost. Hyve manufacturing competes with Asian ODMs (Foxconn, Wiwynn, Quanta, Celestica) on cost and capability, a genuine competitive front.

Do brands matter? SNX’s own brand matters modestly (reseller trust, vendor certifications); the vendors’ brands (Apple, HP, NVIDIA) matter enormously — SNX is a conduit for them.

Nature of competition? Scale, breadth, credit terms, fulfillment speed, and vendor-program access — not price advertising. In Hyve, it is design capability, quality, and hyperscaler trust.

Customers’ switching costs? Interpretation: moderate. A reseller embedded in SNX’s credit lines, portals, and configuration workflows faces real friction to switch, but can and does multi-source across distributors. Two-sided stickiness, not lock-in.

Financial Condition & Balance Sheet

Assets not fully recognized on the balance sheet? The vendor relationships, reseller network, and Hyve hyperscaler programs are unrecognized intangible value; conversely, $4.1B goodwill + $3.8B intangibles from Tech Data overstate tangible net worth.

Off-balance-sheet liabilities? Operating leases (laddered, modest); a €650M European receivables-securitization program and other A/R financing that funds working capital — economically a financing but structured off the revolver. No pension/major contingent liabilities flagged as material.

How conservative is the accounting? Reasonably conservative on the core; the main analytical adjustments are (1) reading GAAP alongside heavily-adjusted non-GAAP (the $296M intangible add-back), and (2) normalizing the growth-inverse working-capital swings in FCF. Management transparently flagged a small (5–10bps) one-time strategic-inventory margin benefit — a good-faith QoE disclosure.

How CapEx-hungry is the business? Fact: barely — core distribution capex is ~0.2% of revenue. The new demand is Hyve footprint (~$100M/yr) and, far more importantly, working capital, which is the real capital intensity: growth consumes cash (Q2 FY26 FCF −$330M; FY22 −$167M).

Capital Allocation & Management

How much FCF does the business generate, and how is it used? In normal years ~$1.2–1.6B (FY25 ~$1.39B), converting ~95% of net income; in hypergrowth years it turns negative as WC builds. Uses: buybacks ($596M FY25), dividends (~$150M), tuck-in M&A. Philosophy: return excess cash, shrink the count, grow the dividend, keep leverage ~1.5×.

Significant acquisitions recently? No transformative deals since Tech Data (2021). Tuck-ins only: Apptium (~$105M, July 2025), IPsense (Dec 2024).

Buying back shares? Yes — $2.0B authorization (Mar-2024), ~$1.0B remaining; ~4%/year share-count reduction since 2021.

Issuing large amounts of stock to insiders? No — SBC is small (~$66M/yr). The June-2026 “secondary” was an Amazon commercial-warrant resale registration (~0.27%), not an insider/sponsor placement.

Compensation policy of directors/management? CEO pay ~$10.3M FY25; incentives 70% non-GAAP net income + 30% adjusted ROIC (capital-efficiency-aligned; no billings/revenue metric). Apollo-nominated directors are uncompensated and own ~0%. Reasonable; the only watch-item is the non-GAAP (intangible-amortization-flattered) bonus base.

Motivations of management? A professional, non-founder team (Zammit ex-Tech Data/Avnet) running for operating-profit growth and capital efficiency — management’s stated #1 priority is “growing operating profit faster than billings,” an explicit rejection of low-quality top-line optics.

Valuation & Market Data

Is the stock an ADR, MLP, or K-1 issuer? No — U.S. C-corp common stock, NYSE-listed, issues a Form 1099 dividend.

Dividend policy? Regular, growing quarterly cash dividend ($0.48/quarter; ~$1.92 annualized; ~0.8% yield; ~13–18% payout) — raised ~9% in January 2026.

How profitable is the business? Thin but stable: ~7% gross, ~2.3% operating, ~1.3% net margin; ROIC ~8.9% GAAP / ~10.5% adjusted.

Is net income diverging from cash from operations? Yes, and by design — in growth periods OCF/FCF run below net income (WC absorption); in flat/down periods they run above. This is a normal, understood distributor dynamic, not an accounting red flag — but it makes single-year FCF multiples misleading and requires through-cycle normalization.

Risks & Downside

What factors would cause the stock to decline? A hyperscaler capex air-pocket or a Hyve program/customer loss (the dominant risk); a valuation de-rating from the richest-ever multiple toward peer ~7× EV/EBITDA; margin give-back as ASP-inflation and strategic-inventory tailwinds fade; component-supply constraints capping growth; and macro/tariff shocks (as in the −25% April-2025 episode).

Risk of a catastrophic loss? Low. Investment-grade balance sheet, diversified core distribution, countercyclical cash generation (a downturn releases working capital into cash), and no single existential dependency — though the 11%-of-revenue customer is the closest thing to a concentration vulnerability.

Chance of a total loss? Negligible. This is a profitable, cash-generative, IG-rated market leader; the realistic downside is a valuation de-rating and a growth disappointment, not insolvency.

Recent News & Events

Has the business environment changed recently? Materially — the AI-data-center capex cycle turned SNX’s Hyve arm from a rounding error into ~28% of revenue and drove a near-doubling of the stock, alongside vendor go-to-market rationalization (the HPE global-distribution win) and component-cost-driven ASP inflation across hardware.

Significant acquisitions? Only tuck-ins (Apptium, IPsense).

Change in accounting policies? None material; note the growing use of net (agent) revenue recognition as software/cloud mix rises, and the prospective introduction of a distinct “Hyve Solutions” segment in FY2026 XBRL.

Recent changes — new markets, facilities, management? New CEO (Zammit, Sept 2024) and CFO (Jordan, Oct 2025); >1M sq ft of new Hyve manufacturing footprint; a new €650M EU receivables-securitization program; and geographic expansion emphasis in APJ and Latin America.

APPENDIX B — Source Appendix

TD SYNNEX Corporation (NYSE: SNX) — research note dated 2026-07-04

Primary sources first, then aggregated/third-party data, then price/factor and industry sources. All management commentary was treated as hypothesis and validated against filings and external data. Facts are cited to the underlying primary document wherever possible.

1. SEC filings (primary; CIK 0001177394)

• FY2025 Form 10-K (filed 2026-01-27, fiscal year ended 2025-11-30) — business description, three geographic reportable segments (Americas / Europe / APJ), revenue disaggregation, vendor concentration (Apple 12%, HP Inc. 10%), customer concentration (one customer 11% of FY25 revenue), risk factors (tariff/FX/vendor/credit/inventory), non-GAAP reconciliations (intangible amortization $296.3M; non-GAAP diluted EPS $13.19; adjusted ROIC 10.5%), capital-return and debt disclosures.
• Prior Forms 10-K (FY2021–FY2024) — multi-year financials, post-merger balance-sheet evolution, tangible-book history.
• Form 10-Q, Q2 FY2026 (period ended 2026-05-31, filed 2026-07-02) — quarterly financials, working-capital detail, and the new “Hyve Solutions Segment” XBRL member.
• Q2 FY2026 earnings 8-K and press release (2026-06-25) — gross billings $28.9B (+33%), non-GAAP operating income $615M (+49%), non-GAAP EPS $4.85, GAAP EPS $4.15; Distribution vs Hyve splits; Q3 FY26 guidance.
• Q2 FY2026 earnings call transcript (2026-06-25) — CEO Patrick Zammit, CFO David Jordan: Hyve +117% to $5.5B, five-hyperscaler programs, Amazon warrant, HPE global-distribution win, cost-plus/pricing commentary, strategic-inventory 5–10bps benefit, FCF −$330M, net leverage 1.6×.
• Form 424B7 prospectus supplement (2026-06-29, off S-3ASR File 333-278517) and related 8-K (2026-06-29) — resale registration of 3,238,066 shares underlying the Amazon.com NV Investment Holdings LLC commercial warrant (215,871 vested as of 2026-06-18); ~0.27% of shares outstanding.
• 8-K (2026-07-02, “snx-20260626”) — new €650M European receivables-securitization program (BNP Paribas master purchaser; Santander, Crédit Agricole).
• 8-K (2024-06-20) — CEO succession: Patrick Zammit → President & CEO effective 2024-09-01; Rich Hume to board.
• 8-K (2024-04) and 424B7 (2024-04-08) — Apollo/Tiger Parent final secondary (5,309,299 shares at $114.20; ~$606M; company concurrent repurchase of 1,750,000); Schedule 13D/A (2024-04-11) reporting Apollo at 0.0% of class.
• Schedule 13G/A (MiTAC Holdings, 2026-02-04) — ~4.3% (<5%), trimming; index-fund 13G/As (Vanguard ~10.3%, BlackRock ~8.5%, FMR/Fidelity ~6.1%).
• DEF 14A proxy (filed 2026-02-13) — executive compensation: annual incentive 70% worldwide non-GAAP net income + 30% adjusted ROIC / return-on-working-capital; PRSU/RSU structure; CEO Zammit FY25 total ~$10.27M; say-on-pay result.
• Form 4 corpus (~Dec 2024 – Jul 2026) — zero open-market purchases (no code-P); net selling led by retired CEO Hume and a director; routine exercise-and-sell / tax-withholding otherwise.
• 8-K (2026-03-27) — charter amendment (25%-holder special-meeting right; de-control step).
• Debt: $600M 6.100% senior notes due 2034 (April 2024 issuance; refinanced $700M 1.250% notes due Aug-2024).

2. Quantitative data — aggregated/third-party (reconciled to filings)

• Third-party fundamental-data aggregator (accessed 2026-07-04) — income statement, balance sheet, cash flow, per-share data, profitability ratios (ROIC/ROE/margins), enterprise value, and valuation multiples (FY2020–FY2025 annual + quarterly), plus the Q2 FY26 earnings-call transcript. Used for the multi-year trend and cross-check; every material figure reconciled to the 10-K/10-Q.
• Third-party market-data providers — own-history valuation percentile ranks (trailing P/E ~20× near the top of the stock’s decade range; P/S and P/B likewise elevated, as of 2026-07-02) and a news/estimate feed (analyst price-target-raise cluster post-Q2). Third-party scores treated as signal and validated against primary filings.

3. Price action & factor positioning

• Split/dividend-adjusted price history (accessed 2026-07-04) — daily OHLCV since 2003; used for the five-year event map, ATH $296.47 (2026-06-24), current $244.64 (2026-07-02), 52-week range, and daily-move attribution.
• A public quantitative factor model (accessed 2026-07-04) — risk-adjusted track record (y1 +77.7%, Sharpe 2.46, max drawdown −15.7%), factor loadings (Market +0.76, SmallSize +0.30, DividendYield +0.28, Robotics&AI −0.11), factor-similar peers (Sanmina/SANM similarity 0.97), and idiosyncratic volatility (23.8%, R² 39%). Third-party statistical estimates; loadings/returns are facts, “will continue/revert” is interpretation.

4. Industry & peer data

• Comparable-company data (company filings/releases and third-party aggregators, accessed 2026-07-04): Arrow Electronics (ARW), Ingram Micro (INGM), ScanSource (SCSC), WESCO (WCC) — revenue, margins, ROIC/ROE, EV/EBITDA, P/E for the peer comparison. Ingram Micro Oct-2024 IPO and post-LBO deleveraging (ir.ingrammicro.com; businesswire 2024-10-23). Arrow FY25 results and ECS segment (businesswire, 2026-02).
• Mordor Intelligence — IT Hardware Distribution & Channel Market sizing/share (2025).
• Omdia / Canalys — technology-distribution channel structure and outlook (2025).
• CDW Corporation public filings — the downstream-reseller peer, used for value-chain framing and the distributor-vs-reseller distinction (CDW is SNX’s customer, not its peer).

5. Analytical frameworks

• Greenwald & Kahn, Competition Demystified — barriers-to-entry / moat-type taxonomy (scale economies + customer captivity), market-share-stability and ROIC tests.
• Marathon / Chancellor, Capital Returns — capital-cycle analysis (disciplined distribution oligopoly vs. capital flooding into the AI end-market; growth-consumes-cash dynamic).

All URLs and documents accessed 2026-07-04 unless otherwise dated. Primary SEC filings are authoritative; third-party aggregated/estimated data were used for acceleration and cross-check and reconciled to filings where material.