[ Inside the business ]
NVDA stock analysis
NVDA · Nvidia Corp
Semiconductors and related devices. Growing 103%. Keeps 65¢ of every dollar.
Watch: a leadership change, growth that has stopped speeding up, an earnings warning score.
The future, in 7 questions
Priced in: 4.0× sales in 5 yrs. Evidence: sales +106% · 65¢ profit/$1 · AI supercomputers rising · scarce.
- Fraud-pattern score high (3.0)
- Growth peaked
- Insiders sold $1.4B outside pre-set plans
- Moves strongly with the S&P 500
From SEC filings, price and Radar. Not advice.
The record
Early signals
Growth peaked.
Sales growth +85%, +106%, then +103% vs. a year ago. A leadership change was filed on Nov 12.
Early signals
Growth peaked.
Sales growth vs. a year ago, by quarter ending
What just happened
Signals in the filings
Fundamentals improving (Piotroski)
Early signals· full write-up
Growth peaked.
Early signals Mixed
What's changing in the filings before it shows up in yearly results. Signals, not predictions.
New since last 10-Q (Aug 26, 2026)
Nov 17, 2026 · Results announced. Revenue $115.8 billion, up 103% from a year ago. Gross margins were 74.8%. Outlook: Revenue is expected to be $130.0 billion, plus or minus 3%; gross margins are expected to be between 71.0% and 72.0%. From 8-KNov 12, 2026 · Executive or director change. The Board of Directors, or the Board, of NVIDIA Corporation accepted the resignation of Jen-Hsun Huang, as NVIDIA's President, Chief Executive Officer and member of the Board, effective immediately. From 8-KSep 3, 2026 · Other event. NVIDIA Corporation ("NVIDIA") entered into a definitive agreement to acquire Hugging Face, Inc. From 8-KGrowth momentum: sales vs. a year ago, by quarter Calculated
Signals in the filings
What could go wrong
Falls harder.
It has moved 1.8× the market. In the 2008 crisis it fell 77% while the market fell 51%. Debt is covered by cash. Guarantees: $108.5B.
What could go wrong
Falls harder.
Red-flag scan, last 3 years
Past shocks replayed stock S&P 500
Cash vs. debt
What could go wrong· full write-up
Falls harder.
8 · What could go wrong
Warning signs in the filings, debt, how it reacts to the economy, and the risks it names
Red-flag scan, last 3 years
Modeled on the SEC's published Accounting Quality Model description, not SEC scores
Debt and leverage as of Jul 26, 2026
Macro risk 118 months to Oct 2026 Calculated
Stress test: past shocks replayed stock vs. S&P 500, month-end prices
Right now: last 3 months Reported
What if the market falls?
Historically, about -36% for this stock
Based on its 1.82× pattern. Real moves vary a lot, as the stress test shows.
Biggest risks it names
- For fiscal year 2026, sales to one direct customer represented 22% of total revenue and sales to another direct customer represented 14% of total revenue, all of which were primarily attributable to the Compute & Networking segment. From 10-K
- Guarantees of up to $108.5B of others' obligations. From 10-K
- Purchase commitments could turn into charges if demand slows. From the 10-Q: "We have significantly increased our supply and capacity commitments from $119 billion last quarter to $279 billion as of July 26, 2026 to meet future demand." From 10-K
Full list: risk factors in the 10-K (Item 1A)
Speculative check
Self-funding.
Passes every condition: profitable, cash-generating and financially sound.
Speculative check
Self-funding.
All 14 tests passed or excused. Price is 10.0× yearly revenue.
Speculative check· full write-up
Self-funding.
Speculative check: does the business pay for itself?
Self-funding: every condition true, across all three statements. Income statement: revenue of $100M or more, a gross profit, an operating profit, a net profit (last 12 months) and a profit in the latest quarter. Balance sheet: current ratio of 1 or more, positive equity, debt manageable (cash at least the debt, or debt no more than 4 years of operating profit before depreciation). Cash flow: positive cash from operations and positive free cash flow. Plus: new shares under 5% a year, no going-concern warning, Altman Z-score out of the distress zone and a Piotroski F-score of 4 or more. Companies with both profits and positive operating cash are excused negative equity and a low Altman score (buybacks), a current ratio under 1 and negative free cash flow (heavy investment). Speculative: an operating loss, cash flowing out of operations and under $1B in sales (however much cash it has raised, even if one-off gains made a net profit); or a going-concern warning; or a penny stock (under $5 a share with under $6M revenue). Not classified: everything else, with the reasons listed. Tests that don't apply (like gross margin for banks) are left out. Calculated
Verdict: Self-funding, 0 of 14 tests red. Numbers to Jul 26, 2026. Calculated
Net profit: $192.9B (last 12 months). Is it making money after all costs? Losses aren't bad by themselves, but they mean someone else pays for now. Reported
Operating profit: $197.6B (last 12 months). Does the business itself make money, before interest, tax and one-off gains? A net profit with an operating loss usually means the profit came from outside the business. Reported
Gross margin: 75% (after the direct cost of what it sells). Does it make money on each sale before overhead? A negative gross margin means every sale loses money. Calculated
Revenue: $303B (last 12 months). Is the business big enough to stand on its own? Under $100 million a year counts as small; under $20 million as tiny. Reported
Cash from operations: $134.4B (last 12 months). Does the day-to-day business bring cash in, or use it up? Reported
Cash runway: Self-funded (free cash flow $127B a year). If it burns cash, how long until it runs out without raising more money? Cash includes short-term investments; burn is cash from operations minus spending on equipment. Calculated
Debt vs. cash: $33.4B debt ($56.6B cash · 0.2× yearly operating profit). Could it pay off its debt? Green when cash covers the debt or debt is under 2 years of operating profit (before depreciation). Reported
Current ratio: 4.6× (short-term assets ÷ bills due within a year). Can it pay the bills due within a year from short-term assets? Below 1× means it can't, without new money. Calculated
Shareholders' equity: $229B (assets exceed debts). Do its assets exceed everything it owes? Negative equity can come from years of losses or from big buybacks, so check which. Reported
New shares issued: −0.9% a year (share count over the last 2 years). Is it selling new shares to survive? Fast-rising share counts shrink each owner's slice. Uses diluted shares from annual reports. Calculated
Going-concern warning: None (in the latest annual report). Auditors must warn when they doubt a company can keep operating for the next year. Any such warning makes the verdict Speculative. From 10-K
Latest quarter profit: $59.7B (quarter to 2026-07-26). The S&P 500 asks for a profit in the latest quarter as well as over the last four. A loss in the latest quarter can be the first sign of trouble, or a one-off. Reported
Altman Z-score: 24.4 (safe zone). Edward Altman's bankruptcy-risk score from working capital, retained earnings, operating profit, market value and sales. Under 1.8 (1.1 for non-manufacturers) is the distress zone. Reported
Piotroski F-score: 4 of 9 (profit, cash, debt, liquidity, dilution, margins, efficiency vs. last year). Joseph Piotroski's nine pass/fail checks, this year vs. last: profit, operating cash, rising return on assets, cash above profit, less debt, better liquidity, no new shares, better margins, better asset turnover. 0–3 is weak, 7–9 strong. Reported
Benjamin Graham's defensive test (for reference, not part of the verdict): passes 6 of 7. ✓ Sales of $700M or more (his $100M in 1972 money) · ✓ Current ratio of 2 or more · ✓ Long-term debt no more than working capital · ✓ A profit every year (last 10) · ✓ Dividends every year (last 10) · ✓ Earnings up a third or more over the decade (3-year averages) · ✗ Price × earnings × book value moderate (P/E × P/B ≤ 22.5)
Price vs. sales: the market values it at 10.0× its yearly revenue. This is about the stock price, not the business, so it isn't one of the tests. Calculated
A label for learning, not a recommendation. Speculative companies can succeed and self-funding ones can fall.
What they sell
51% Hyperscale.
Hyperscale 51% · AI Clouds, Industrial, & Enterprise 42% · Edge Computing 7% (latest quarter)
What they sell
51% Hyperscale.
Latest quarter, to Jul 26, 2026
Full year (its older categories)
What they sell· full write-up
51% Hyperscale.
1 · What they sell fiscal year to Jan 25, 2026
Reported 10-K revenue by product or segment
Latest quarter, to Jul 26, 2026: new categories Reported
Growth vs. the same quarter a year earlier. The company now reports these categories; the yearly split above uses its previous ones.
Rivals it names: Our current competitors include: suppliers and licensors of hardware and software for discrete and integrated GPUs, custom chips and other accelerated computing solutions, including solutions offered for AI, such as Advanced Micro Devices, Inc., or AMD, Huawei Technologies Co. From 10-K
Where each dollar goes
65¢ profit.
25¢ to make what it sells, 9¢ on research and overhead, 65¢ left over. Revenue $303.0B in the last 12 months.
Where each dollar goes
65¢ profit.
Out of each $1 of $303B revenue
Where customers are
Where each dollar goes· full write-up
65¢ profit.
3 · Where each $1 goes last 12 months to Jul 26, 2026
$303.0B revenue Reported
Out of each $1 of revenue Reported
- Making the products25¢
- Research and development8¢
- Sales and admin2¢
- Operating profit65¢
Revenue by product fiscal year to Jan 25, 2026
- Compute75%
- Networking15%
- Gaming7.4%
- Professional Visualization1.5%
- Automotive1.1%
- OEM and Other0.3%
Where customers are last 12 months to Jul 26, 2026 Reported
Its market
1.3× the economy.
The government expects its industry, semiconductor and other electronic component manufacturing, to grow 2.5% a year to 2035, vs. 2.0% for the whole economy.
Its market
1.3× the economy.
Output growth a year, 2025–35, after inflation
Its market· full write-up
1.3× the economy.
2 · Where its markets head (10 years)
U.S. industry output growth after inflation, 2025–35
Projected to grow 1.3× as fast as the economy. Jobs in the industry: +0.8% a year. Gov projection
U.S. Bureau of Labor Statistics, Employment Projections 2025–35 (Table 2.11), released Aug 27, 2026
Market size and share
Your call.
There's no official size for its market. Enter one you trust to see its share, and slide to see what a bigger share would mean.
Market size and share
Your call.
4 · Market size and share
Not available: no market size was found in the company's SEC filings. Companies usually give these in investor presentations, which aren't part of their official filings.
Market capture
Needs a market size (above). Nothing is assumed.
Life cycle
Early maturity.
It earns cash, still invests, and pays back more than it borrows. Unusually, sales are still growing like a young company's.
Life cycle
Early maturity.
Life cycle stage Early maturity
Where the company sits, based only on its financials
NVDA, last 12 months to Jul 26, 2026 Reported
Cash pattern (+ − −) = maturity. Sales are still growing like a growth company, so the dot sits just past the line. Inferred
Path over time
Earnings quality
Flagged.
On paper its numbers look like past manipulators', mostly from sales growth and asset quality. That's a reason to look closer, not proof. Cash covers 70¢ of each $1 of profit.
Earnings quality
Flagged.
Beneish M-score: -1.15
Fraud-pattern scores by year · NVDA
| Year | M-score | M zone | F-score | F risk |
|---|---|---|---|---|
| FY21 | −1.65 | 🔴 Flagged | 1.61 | 🟠 Above normal |
| FY22 | −1.74 | 🔴 Flagged | 1.02 | 🟠 About normal |
| FY23 | −2.69 | 🟢 Clean | 0.98 | 🟢 Normal |
| FY24 | −1.12 | 🔴 Flagged | 1.10 | 🟠 About normal |
| FY25 | −0.95 | 🔴 Flagged | 1.69 | 🟠 Above normal |
| FY26 | −1.15 | 🔴 Flagged | 2.98 | 🔴 High |
M-score: flagged above −1.78, grey zone −2.22 to −1.78, clean below −2.22. F-score: 1.0 = an average company's risk, above 2.45 = high. Patterns to look into, not proof.
Cash checks
Earnings quality· full write-up
Flagged.
5 · Earnings quality
Are the profits backed by cash?
Mostly from sales growth, asset quality, accruals.
How cautious are you? cost of a miss vs. a false alarm
Flagged at cutoff -1.28
Vs. 15 peers (Semiconductors & Related Devices): #2 highest · peer median -2.58 · 3 in the high zone (fastest growers among them: CRDO, MRVL) Calculated
Fiscal years to Jan 25, 2026 and Jan 26, 2025 · Beneish (1999)
Its fiscal 2026 numbers look 3.0× as much like past SEC fraud cases as an average company's.
Mostly from hard-to-value assets (75% of assets, incl. $22.3B of private stakes) and profit ahead of cash (48% of assets).
| Fiscal year | Beneish M-score | M zone | Dechow F-score | F risk |
|---|---|---|---|---|
| FY2021 | −1.65 | 🔴 Flagged | 1.61 | 🟠 Above normal |
| FY2022 | −1.74 | 🔴 Flagged | 1.02 | 🟠 About normal |
| FY2023 | −2.69 | 🟢 Clean | 0.98 | 🟢 Normal |
| FY2024 | −1.12 | 🔴 Flagged | 1.10 | 🟠 About normal |
| FY2025 | −0.95 | 🔴 Flagged | 1.69 | 🟠 Above normal |
| FY2026 | −1.15 | 🔴 Flagged | 2.98 | 🔴 High |
Each year compared with the one before. “Off the scale” usually means a merger, spin-off or restructuring that year. Beneish (1999); Dechow, Ge, Larson & Sloan (2011), model 1.
Competitive advantage
9 of 10 years.
Returns beat its 12.9% cost of capital in 9 of the last 10 years. Gross margin went from 59% to 71%.
Competitive advantage
9 of 10 years.
Return on capital by year, vs. its cost of capital
Competitive advantage· full write-up
9 of 10 years.
6 · Competitive advantage
Can competitors erode its returns? Evidence, not a verdict.
Return on capital by fiscal year; dashed line = 12.9% cost of capital
Source of advantage, per the 10-K: switching costs Inferred · Our currently issued patents have expiration dates from March 2026 to June 2045. From 10-K
Rivals it names: Our current competitors include: suppliers and licensors of hardware and software for discrete and integrated GPUs, custom chips and other accelerated computing solutions, including solutions offered for AI, such as Advanced Micro Devices, Inc., or AMD, Huawei Technologies Co. From 10-K
Future bets
$279B committed.
Promised to suppliers, per its latest 10-Q. Plus $23.6B a year on research (8% of sales). Stakes in other companies: $90.7B.
Future bets
$279B committed.
Where the money for the years ahead goes
Future bets· full write-up
$279B committed.
7 · Future bets
Where is it putting money for the years ahead? Plans and commitments, not outcomes.
“We have significantly increased our supply and capacity commitments from $119 billion last quarter to $279 billion as of July 26, 2026 to meet future demand.” From 10-Q
All figures from Nvidia Corp's SEC filings (EDGAR) unless tagged otherwise. Educational only, not investment advice.