Report contents
At a glance
- Business model: Fortinet sells FortiGate next-generation firewalls through two-tier distribution, then bills FortiGuard, FortiCare, and cloud-delivered subscriptions against the installed sockets those appliances create. Service revenue carries the higher gross margin because threat research and engineering costs are largely fixed. The strength is that renewals avoid a fresh competitive bake-off; the obstacle is that a delayed appliance refresh thins the annuity that follows it.
- Financial quality: In FY2025 Fortinet turned $6.80bn of revenue into $2.59bn of operating cash flow and $2.23bn of free cash flow at an 80.9% gross margin. The $9.15bn liability stack is mostly deferred revenue from customer prepayments, not borrowings; interest expense was $20m. Buybacks absorb close to all free cash flow, so total equity stays thin near $1.24bn and ROE is not interpretable.
Business model
I. Business model: a firewall franchise that sells an annuity through the channel
Fortinet is a brand-led security hardware maker with a subscription layer wrapped around it, not a pure software platform. Its core product is the FortiGate line of next-generation firewalls — appliances that inspect and control traffic entering a network — sold with software licenses, and then surrounded by recurring services: FortiGuard security subscriptions, FortiCare support, and cloud-delivered offerings such as SASE (security delivered from the cloud rather than from on-premises boxes). The paying customers are enterprise, mid-market, and small-business IT and security teams, service providers and managed security service providers (MSSPs), and government agencies. Most do not buy directly. Fortinet sells through two-tier distribution, where distributors sell to resellers. Product revenue is recognized when appliances and licenses ship; service revenue is recognized ratably over the contract term. Stage: the firewall core is mature and refresh-driven, while cloud-delivered security is still an expansion effort.
One data caveat up front. The only piece of Fortinet's FY2025 Form 10-K (filed February 25, 2026) visible in the supplied material is the cover page — 739.9 million shares outstanding as of February 20, 2026, and $56.8 billion of non-affiliate market value at June 30, 2025. Segment, product, and geographic mix live in Item 1 and MD&A, which are not in the excerpt. So the mix described below is structure, not FY2025 results.
The appliance is the hook; the subscription is where the profit sits
The direct cause of the revenue pattern is mechanical. An appliance sale installs a socket at a customer site, and each later FortiGuard or FortiCare renewal bills against that socket without a fresh competitive bake-off. The deeper cause is cost structure: threat research, engineering, and support are largely fixed and spread across a growing installed base, so service revenue carries structurally higher gross margin than hardware. Land the box, then monetize it for years.
The real tradeoff is price versus demand. Discounting to win a refresh cycle lowers product gross margin today and shrinks the base off which future renewals get priced. Win the socket cheaply and the annuity is thinner for years. Hold price and a competitor may take the socket — and the annuity with it.
What would overturn this reading: product revenue growing while deferred revenue (cash collected for services not yet delivered) stalls. That would say the installed base is not converting, and the annuity is weaker than the appliance shipment numbers imply.
Competition now looks like consolidation, not point products
Firewall budgets are consolidating toward fewer vendors. Fortinet competes with dedicated firewall vendors, networking incumbents bundling security into broader deals, cloud-delivered secure web gateway and SASE providers, and cloud platforms shipping native controls. Pressure usually shows up first in price, then in large-deal counts and average selling price.
The structural reason is that a firewall refresh is a re-evaluation event in a way a subscription renewal is not. A customer that keeps the appliance keeps the subscription attached to it; a customer who switches vendors takes both, which is why competitive losses are more damaging here than in a typical hardware business. The tradeoff is expansion versus execution — pushing SASE and SecOps products creates a second front against well-funded rivals before the firewall core has been fully defended.
Two-tier distribution buys reach and costs visibility
Distributors and resellers get Fortinet into mid-market, international, and public-sector accounts it could not serve directly. That is the operating benefit. The structural cost is that Fortinet sees sell-in, not end demand. Distributor inventory can build ahead of a slowdown, and a quarter of channel stuffing shows up as a weak following quarter. MSSP relationships create renewal stickiness but also concentrate renewal risk if a large partner standardizes elsewhere, and long equipment refresh cycles mean a lapsed upgrade is often a lost customer rather than a deferred one.
What to watch in the next filings
The cleanest public test of this model is the product-versus-service revenue split, deferred revenue growth, and any disclosure on renewal rates or customer counts in Item 1 and MD&A. Renewal rates would directly show whether installed sockets keep converting; without that line item, deferred revenue and service mix are the closest available proxies. Distributor inventory and large-deal counts are the earliest signs that pricing pressure — rather than demand — is driving the numbers.
Questions worth following
- Can Fortinet keep renewing subscriptions off its installed firewall base as refresh cycles stretch?
Renewals are the profit engine: they bill against sockets already installed and carry much higher margin than hardware. If the base stops converting, service revenue growth fades and the appliance business must be re-won every cycle at competitive prices.
- Can Fortinet hold appliance pricing as cloud-delivered firewall and SASE rivals compete for the same refresh budgets?
Pricing is the first place competitive pressure appears, and it determines both product gross margin and the size of the annuity that follows each socket. Concessions would show up in margin and deal economics well before any customer-count decline.
- Is Fortinet's channel adding mid-market and operational-technology buyers, or mainly reselling to existing enterprise accounts?
Growth funded by deeper penetration of the same accounts has a lower ceiling and greater exposure to competitive displacement. Genuinely new customers expand the installed base that future service revenue bills against, which changes the durability of the annuity.
Financial quality
II. Financial quality: 80% gross margins funded by customers, not lenders
Fortinet's income statement reads like a high-quality software business and its balance sheet reads like a leveraged one. In FY2025 it produced $6.80bn of revenue, a 27.2% net margin, and $2.59bn of operating cash flow, while carrying $9.15bn of liabilities against $10.39bn of assets. It also recorded $20m of interest expense for the year. That last number is the one that reframes everything else: almost none of the liability stack is borrowed money. Most of it is service Fortinet has already been paid to deliver. That is why cash has run ahead of reported earnings for five straight years, and why ordinary leverage and liquidity screens misread this company in both directions.
Last five fiscal years plus the first half of 2026 (USD bn unless noted)
| Metric | FY2021 | FY2022 | FY2023 | FY2024 | FY2025 | H1 2026 | Assessment |
|---|---|---|---|---|---|---|---|
| Revenue | 3.34 | 4.42 | 5.30 | 5.96 | 6.80 | 3.90 | Strong — about 15% three-year CAGR |
| Gross margin | 76.6% | 75.4% | 76.7% | 80.6% | 80.9% | 80.0% | Exceptional |
| Operating margin | 19.5% | 21.9% | 23.4% | 30.2% | 30.6% | 32.6% | Strong — opex leverage |
| Net margin | 18.2% | 19.4% | 21.6% | 29.3% | 27.2% | 29.2% | Exceptional; FY2024 tax-flattered |
| ≈ Effective tax rate | 1.6% | 3.2% | 4.9% | 13.6% | 19.2% | 16.7% | Adequate — normalizing |
| Operating cash flow | 1.50 | 1.73 | 1.94 | 2.26 | 2.59 | 2.12 | Exceptional |
| Free cash flow | 1.20 | 1.45 | 1.73 | 1.88 | 2.23 | 1.98 | Exceptional — positive every year |
| Capex / revenue | 9.0% | 6.3% | 4.0% | 6.4% | 5.3% | 3.6% | Exceptional — asset-light |
| Operating cash flow / net income | 2.46× | 2.01× | 1.69× | 1.30× | 1.40× | 1.86× | Strong |
| Liabilities / assets | 86.5% | 104.5% | 106.4% | 84.7% | 88.1% | 85.7% | Demanding, but not borrowed |
| Current ratio | 1.55 | 1.24 | 1.19 | 1.47 | Not disclosed | Not disclosed | Healthy but thin |
| Receivable days | 83 | 104 | 92 | 88 | Not disclosed | Not disclosed | Adequate and improving |
| Total equity | 0.80 | (0.28) | (0.46) | 1.49 | 1.24 | 1.55 | Thin; moves with repurchases |
| ROE | 74.1% | n/m | n/m | 338.7% | Not disclosed | Not disclosed | Not meaningful |
H1 2026 sums the two supplied quarters. Ratios marked ≈ are computed from supplied line items. Capex is operating cash flow minus free cash flow. n/m means negative or near-zero equity makes the ratio meaningless.
Six supplied quarters; receivables, inventory, and fixed assets go dark after FY2024
| Single quarter, USD bn | Q1'25 | Q2'25 | Q3'25 | Q4'25 | Q1'26 | Q2'26 | Assessment | Trend |
|---|---|---|---|---|---|---|---|---|
| Revenue | 1.54 | 1.63 | 1.73 | 1.91 | 1.85 | 2.05 | +20% and +26% year over year in 2026 | Accelerating |
| Gross margin | 81.2% | 81.6% | 81.5% | 79.6% | 80.0% | 80.0% | Steady near 80% | Flat |
| Net income | 0.43 | 0.44 | 0.47 | 0.51 | 0.53 | 0.61 | Steady | Improving |
| Operating cash flow | 0.86 | 0.45 | 0.66 | 0.62 | 1.08 | 1.04 | Lumpy with billings | Improving |
| Free cash flow | 0.80 | 0.28 | 0.57 | 0.58 | 1.01 | 0.97 | Follows operating cash | Improving |
| Total assets | 10.41 | 10.64 | 9.36 | 10.39 | 9.88 | 10.86 | Falls when cash funds buybacks | Volatile |
| Total liabilities | 8.44 | 8.58 | 8.63 | 9.15 | 8.89 | 9.31 | Grows with the business | Rising |
| Total equity | 1.96 | 2.06 | 0.73 | 1.24 | 0.99 | 1.55 | Set by repurchases | Volatile |
| Cash | 3.60 | 3.37 | 2.00 | 2.50 | 2.22 | 2.94 | $1.4bn drop in Q3'25 | Recovering |
| Current liabilities | 4.73 | 4.78 | 4.79 | 5.03 | 4.62 | 4.84 | Stable | Flat |
| Goodwill | 0.24 | 0.26 | 0.26 | 0.26 | 0.34 | 0.26 | Round trip not explained | Watch |
| Fixed assets | Not disclosed | Not disclosed | Not disclosed | Not disclosed | Not disclosed | Not disclosed | Not in supplied data | — |
Only six quarters are supplied, so this is not the eight-quarter series I would want. Fortinet's fiscal year is the calendar year; these are calendar quarters. Gross margin is computed from supplied gross profit and revenue.
Opex, not gross margin, produced the earnings step-up
Gross margin rose from 75.4% in FY2022 to 80.6% in FY2024, but operating margin moved from 21.9% to 30.2% — an 8.3-point gain from a 5.2-point improvement in cost of revenue. The rest came from operating expenses growing slower than revenue: 53.4% of revenue in FY2023, 50.3% in FY2024. That is real scale on a largely fixed R&D and sales base, with mix as the structural driver. FY2025 shows the limit: gross margin improved another 0.3 point and opex held at 50.3%, yet net margin fell 2.1 points because the effective tax rate moved from about 13.6% to 19.2%. The 29.3% net margin of FY2024 is not a run rate; 27% on a normalized tax rate is the better anchor. The supplied statements do not break out R&D, so the essential check — whether margin is being funded by product development or harvested from it — cannot be completed here.
Repurchases turn ROE into noise and consume nearly all free cash flow
Equity fell from $1.49bn to $1.24bn during FY2025 even though the company earned $1.85bn. It was negative in FY2022 and FY2023, producing ROE of +343% and −308% in consecutive years, and FY2024's 338.7% comes off a base equal to 15% of assets for a company holding $9.76bn of them. ROE here measures the buyback, not the business. The cash test is the real one: FY2025 free cash flow of $2.23bn against $2.37bn of financing outflows, mostly repurchases, with cash falling from $2.88bn to $2.50bn. Distributions slightly outran internally generated cash. H1 2026 reversed that — free cash flow of $1.98bn against $1.49bn of financing outflows, cash back to $2.94bn — but the program is designed to absorb close to 100% of free cash flow, so no cushion builds.
Cash conversion normalized; it did not deteriorate
Operating cash flow ran 2.46× net income in FY2021 and 1.30× in FY2024. That looks like a warning and mostly is not. The upfront cash benefit from deferred revenue scales with the growth rate, not the balance, so as revenue growth slowed from 32% to 12–14%, the gap between cash and earnings narrowed by arithmetic. Collections actually improved: receivable days went from 104 in FY2022 to 88 in FY2024. The ratio stabilized at 1.40× in FY2025 and 1.86× in H1 2026. The test is whether it holds above roughly 1.2× while deferred revenue keeps growing; a slide below 1.0× alongside decelerating revenue would mean the accrual earnings, not the timing, are the problem.
What to check
The composition of $4.84bn of current liabilities against $2.94bn of cash. Interest expense rules out borrowings, but the split between deferred revenue, payables, and tax accruals decides whether that gap is a delivery obligation or a funding gap. The supplied data does not break it out, and it is the single most important missing disclosure.
Goodwill moved from $0.26bn to $0.34bn in Q1 2026 and back to $0.26bn in Q2 2026. Goodwill rises with acquisitions and falls through impairment or disposal. Three months is a short life for a purchase; the 10-Qs should say which happened.
H1 2026 revenue grew about 23% against FY2025's 14%, with gross margin flat near 80%. If deferred revenue is growing at least as fast, the acceleration is contractual. If revenue accelerates while deferred revenue lags, part of it was pulled forward — and that is the difference between a permanently higher earnings base and a cyclical peak.
Generated by DeepVal AI. Data and models may contain errors or be outdated. For research only; not personalized investment advice.
Continue reading the full report
- Key risksWhat could break the investment case
- Valuation and margin of safetyWhat the current price assumes, and how much room for error it leaves
- SummaryOur overall judgement and what to watch next
Free to sign up · 20 research points · unlocking this report uses 10
20 points cover 2 full reports. No card needed.
Research reflects its generation date. Preview the first two modules.
More on Fortinet Inc
Fortinet Inc
Business, moat and key questions
Open company page