NTGR — NETGEAR, Inc. · 2026-09-06 · Verdict: WATCH · Conviction 2
Price $21.31 (screen row from universe_under2b.csv, 2026-09-06 build; no live prices available) · Mkt cap $579.2M · EV $311.3M (screen says $404.9M; it counts only cash and misses $93.6M of short-term investments) · EV/EBIT n/m (LTM GAAP operating loss $33.8M) · FCF yield -3.0% (LTM FCF -$17.3M) · Net cash $267.9M, no debt · ADV $6.8M Sources read: 10-K filed 2026-02-13 (Items 1, 1A, 7), 10-Q filed 2026-08-06 (Q2 FY26), DEF 14A 2026-04-16, 8-Ks 2026-04-14, 2026-04-16, 2026-04-29, 2026-06-02, 2026-08-06, Form 4s (12 most recent of 30 filed in the trailing 12 months). The file labelled "transcript" in the bundle is the Q2 press release text, not call Q&A, so section 5 rests on written guidance versus reported results.
1. What the business actually does
NETGEAR sells networking hardware, software and subscriptions in two segments (10-K 2025, Item 1). Enterprise sells Pro AV managed switches, SME wireless LAN, the Insight cloud management platform and Exium firewall/SASE software through distributors, direct market resellers and managed service providers. Consumer sells WiFi 6/7 mesh systems and routers plus 4G/5G mobile hotspots through Amazon, Best Buy, Walmart and broadband carriers, and includes a service provider channel management is explicitly harvesting. Manufacturing is fully outsourced to Foxconn, Delta, Senao, Pegatron and T&W. FY2025 revenue was $699.6M, 68% of it Americas, with 784 employees.
2. Why it is mispriced — the edge case
Misunderstood segment, plus a screen artefact. Consolidated GAAP results show an operating loss in every recent period: -$34.2M in FY2025, -$22.0M in H1 2026 (10-K 2025, Item 7; 10-Q Q2 2026). Underneath, Enterprise earned $86.2M of segment contribution on $353.0M of LTM revenue, a 24.4% margin, and Q2 2026 alone delivered 7.7% growth, an all-time-high 54.1% non-GAAP gross margin and a 25.9% contribution margin, the best in over seven years (8-K 2026-08-06). Enterprise is now more than half of revenue and about 69% of non-GAAP gross profit (Q2 release, CEO). Consumer produced $5.0M of contribution on $341.4M of LTM revenue. Two mechanical things obscure this. First, segment contribution excludes roughly $125M a year of costs below the line, so a profitable segment coexists with a loss-making company. Second, the screen computes EV from cash only and flags "debt data missing"; the Q2 balance sheet shows no borrowings at all and $93.6M of short-term investments alongside $174.3M of cash, so real EV is $311M, not $405M. Who is selling: index and quantitative funds screening on GAAP losses and negative free cash flow, both accurate descriptions of the consolidated entity.
3. Unit economics and growth
LTM revenue is $694.4M, roughly flat, and the mix shift is the whole story. Consolidated gross margin went from 29.1% (FY2024) to 38.0% (FY2025) to 40.3% in H1 2026 (10-K 2025, Item 7; 10-Q Q2 2026). Enterprise grew 18.8% in FY2025 and 6.8% in H1 2026 on Pro AV switches at higher selling prices. Consumer fell 7.3% then 9.5%, but splitting it, the harvested service provider line fell to $44.0M in H1 2026 from $56.9M while the rest of Consumer fell only 2.8% to $110.6M (Q2 release). Subscription ARR is about $42M, up from $40M a quarter earlier. Customer concentration is real (Amazon, Best Buy, Walmart, TD SYNNEX, Ingram Micro), and the 10-K warns results depend on large orders from few customers (Item 1A). The problem is everything below segment contribution. LTM segment contribution of $91.2M becomes a $33.8M GAAP operating loss, a $125.1M gap made of unallocated corporate cost (about $71M annualised from H1 2026), stock compensation (about $36M annualised), amortisation and restructuring. Free cash flow was -$18.9M in FY2025 and -$17.3M LTM. This company does not currently generate cash.
4. Balance sheet and capital allocation
Cash plus short-term investments was $267.9M at June 28, 2026, down $55.2M from $323.0M at December 31, 2025 (10-Q Q2 2026). There is no debt. Obligations are $55.3M of non-cancellable purchase commitments and $67.0M of undiscounted lease payments (10-K 2025, Item 7). Buybacks are real but half-eaten by dilution: $33.0M spent in H1 2026 retired 1.5 million shares, yet shares outstanding fell only from 27.94 million to 27.14 million, so about 0.7 million shares of RSU and ESPP issuance offset it. The Board added $75.0M to the authorisation on April 22, 2026. Insiders show zero open-market purchases and $162,531 of sales across the 12 Form 4s parsed. CEO Prober holds 426,294 shares, about $9.1M of skin, but his 2025 total compensation was $10.56M (DEF 14A 2026), roughly 1.8% of the market cap in one year, and his performance units vest on relative shareholder return percentile rather than on segment margin or cash generation.
5. Management: what they said versus what they did
In April, management guided Q2 revenue to $150M-$165M, GAAP operating margin to -8.4% to -5.4% and non-GAAP to -1.0% to 2.0% (8-K 2026-04-29). Q2 came in at $168.6M, -5.0% GAAP and +2.4% non-GAAP: above the high end on all three. They guided service provider revenue to "around $18 million" and delivered $23.8M. The CEO is candid about the weak half, promising to "profitably scale our Enterprise business while preserving optionality for our Consumer business" (Q2 release). Q3 guidance is the honest counterweight: GAAP operating margin of -12.0% to -9.0% on $165M-$175M of revenue, a step backwards driven by an expected 200 basis point second-half gross margin headwind from memory costs.
6. Valuation
Cash and investments are $9.87 a share, so the operating business is priced at $11.44 a share, or $311M. Reverse DCF: at a 10% discount rate and no growth, that price implies about $31M a year of sustainable after-tax free cash flow from a business that produced negative $17M over the last twelve months, a swing of roughly $48M. Bear (30%, about $14): memory inflation runs through 2027, Consumer contribution margin sits at -4% on $310M, Enterprise growth slows to 3% at a 22% contribution margin, corporate cost stays at $71M, buyback is suspended and cash falls to $200M. Market pays 0.5x Enterprise revenue plus cash. Base (50%, about $25): Enterprise compounds 7% to about $404M by FY2028 at a 25% contribution margin, Consumer declines 8% a year at roughly break-even contribution, corporate cost flat. Non-GAAP operating income reaches about $33M, GAAP is near break-even, free cash flow $15M-$20M. Value the operating business at 1.0x Enterprise revenue, cross-checked at 12x pre-stock-compensation operating income, plus $250M of net cash on about 26 million shares. Bull (20%, about $37): Consumer is separated or wound down, Enterprise compounds 10% to $427M at a 26% contribution margin with a lighter $55M standalone corporate load, earning about $56M of operating income at 14x, plus $230M net cash. Probability weighted, roughly $24 against $21.31, about 13% upside. That is thin compensation for a business that currently burns cash.
7. Catalysts and timeline
Q3 results in late October 2026 show whether the memory shock is one quarter or a regime. Q4 and Q1 2027 show whether Consumer contribution margin recovers above zero once price increases flow through. The "preserving optionality for Consumer" language and the deliberate SIC code change to an enterprise classification both point toward a possible structural separation, the bull case trigger.
8. Risks and pre-registered kill criteria
Sole-source memory and chipset supply with lengthening lead times is the live risk (10-K 2025, Item 1A; 10-Q Q2 2026). TP-Link, Ubiquiti, Cisco and Amazon-owned Eero all have more resources. The March 2026 FCC Covered List designation of foreign-produced routers left NETGEAR on a conditional approval granted in April 2026; losing it would halt new consumer router launches (10-Q Q2 2026). Kill criteria: 1. Enterprise segment contribution margin below 20% for two consecutive quarters. 2. Enterprise revenue growth below 3% year over year for two consecutive quarters. 3. Cash plus short-term investments below $220M at any quarter end without at least $40M of cumulative buyback since June 28, 2026. 4. Zero share repurchases for two consecutive quarters while the authorisation is outstanding.
9. Verdict and one-paragraph summary
WATCH, conviction 2. NETGEAR is two businesses inside one GAAP loss. Enterprise is genuinely good and getting better: $353 million of revenue growing 7%, a record 54.1% gross margin, a 25.9% contribution margin and $86 million of segment contribution, now more than half of revenue and 69% of gross profit. The screen understates the balance sheet, which holds $268 million of cash and investments and no debt, so you pay only $311 million for the operating company. But the consolidated entity still loses $34 million a year at the operating line and burned $17 million of cash over the last twelve months, because roughly $125 million of corporate cost and stock compensation sits below segment contribution while Consumer contributes almost nothing. Management has beaten its own guidance twice running, yet guides Q3 GAAP operating margin to negative 12% to negative 9% on memory inflation. At $21.31 my probability weighted value is about $24, a 13% edge on a company that does not yet generate cash and whose buyback is half consumed by dilution. Wait for the Q4 2026 or Q1 2027 print that shows Consumer contribution margin back above zero with Enterprise holding 25%, and this becomes an IDEA. Not yet.
Research for discussion, not investment advice. Positions and sizing are the reader's decision.