NVIDIA CORP
NVDA · Technology · $5.03T mkt cap · FY2026 filings · Wide moat ·
On the watchlist
The five investor questions
2022–2026 (latest 5 aligned FYs; v3 window cap 5): revenue $26.9B to $215.9B, increasing; diluted EPS 0.39 to 4.90, increasing.
Median gross margin 62.2% over 10y, stable.
Median ROIC 46.1%, above the 12% hurdle in 100% of years.
Net cash position — no leverage risk.
Owner earnings +60.8%/yr, share count growing (dilution).
Price vs. conservative owner-earnings floor
A cautious baseline, not a price target: what the shares may support if normalized owner earnings simply hold steady.
- Owner earnings (normalized)
- $30.2B
- Owner-earnings floor / share
- $13.81
- Capped-growth comparison / share
- $22.37
- Recent price
- $206.84
- Price vs. owner-earnings floor
- 1398% above owner-earnings floor
What today’s price assumes: owner earnings growing ~72%/yr over 5 years. The floor assumes owner earnings stay flat (0% growth).
How to read this
We solve for the constant annual change in owner earnings that would make the five-year floor equal today’s price. This keeps the existing model inputs and calculation unchanged; it is an expectations lens, not a forecast.
This floor assumes trailing owner earnings persist — unreliable for declining businesses. The recent FY revenue series does not trigger the three-consecutive-declines flag.
Cohort context: median current price is 239% above the floor across 48 of the current top 50 companies by Moat Score with both values.
Score records captured Jul 26, 2026.
Exactly which 50 companies?
FTNT, META, MEDP, IBEX, MSFT, GOOGL, LULU, ANF, IDXX, TXRH, FHI, QLYS, FFIV, ADBE, LLY, MANH, VEEV, GRMN, SSD, MNST, DOCS, MCO, CPRT, NTAP, LRCX, NVDA, CORT, SFM, MPTI, ULTA, SN, RMD, ROL, PAYC, IRMD, INTU, DECK, MORN, LGCY, DAL, GWW, ISRG, MPWR, CTAS, LOPE, MRK, CMG, FAST, BKNG, ORLY
Missing a usable price or floor: ANF, BKNG.
Model details under moat-index@3.0.0: zero-growth floor 11.1×, capped-growth comparison at no more than 18×; 9% discount rate; high-end growth 4% (capped at 4%) using min(5y revenue CAGR, 5y diluted-EPS CAGR), clamped to 0%–4%; maintenance capex uses min(capex, D&A); abstain when D&A is unavailable. This is a disclosed model range, not intrinsic value as fact.
20 years of fundamentals
The business, in plain English
NVIDIA CORP booked $215.9B of revenue in FY2026 in the Technology sector and kept 71.1% of it as gross profit — a high-margin business by that measure. After every other cost, 55.6% of each revenue dollar reached the bottom line.
Across the filed record, revenue grew from $4.1B (FY2008) to $215.9B (FY2026) — about 24.6% a year compounded over 18 years.
It earned 71.3% on invested capital in the latest filed year, FY2026. Across the full 19-year measurable filed record, median ROIC was 19.4%. Over the v3 recent window (10 measurable filed years), median ROIC was 46.1%. The Returns on Capital filter above scores it 100/100.
The balance sheet carried $8.5B of total debt in FY2026 against $120.1B of owner earnings — roughly 0.1 years of owner earnings to retire it all. Balance-Sheet Safety scores it 100/100.
Put together: Returns on Capital is the strongest of the five filters (100/100) and Capital Discipline the weakest (72/100), which is how NVDA lands at 92/100 — a Wide moat.
This breakdown is generated from the filed numbers and sub-scores above — no outside narrative, no estimates. Where a filing doesn’t disclose an input, the sentence that would need it is omitted instead of guessed.
Gaps in a line mean that item isn’t in NVDA’s filings for that year — the series is never interpolated or estimated. The Table view lists every filed value, including operating and net margins, total debt, and share count.
Moat Score history
Breaks in the line mark methodology or normalizer upgrades — readings across versions aren’t compared, so a level shift there isn’t a change in the business.
Tier changesSame-methodology crossings of the Wide / Narrow / Shallow bars
- Normalizer upgraded unstamped (version unknown)tomoat-xbrl-normalizer@1.1.0 — readings across versions aren’t comparedJul 26, 2026
- Methodology upgraded moat-index@2.0.0tomoat-index@3.0.0 — readings across versions aren’t comparedJul 23, 2026
- Methodology upgraded moat-index@1.2.0tomoat-index@2.0.0 — readings across versions aren’t comparedJul 20, 2026
Scores are logged append-only and never overwritten — this record can’t be backfilled, which is exactly why it’s worth keeping.
Track record
How NVIDIA CORP’s moat rated in each of the years we can reconstruct from its filings — scored only on what was knowable at the time — and what its price and returns did afterward. The score never saw a price; the two are joined only in hindsight, for education, not as a signal.
How to read this: each dot is what the engine would have scored NVDA on that December 31; the line below is its total-return price path (dividends reinvested) in the years since.
Two tracks, one timeline: the score has its own 0–100 scale (top), the price its own 100-based scale (bottom) — never a shared axis. The price path is a total-return (dividends reinvested) index built from the same data the forward returns use; gaps in the score line are years with no reconstructed rating (see the table for why). The Table view carries every value.
What followed, in the years it rated Wide
In the 6 years NVDA rated Wide-moat (2019, 2020, 2021, 2022, 2024 and 2025), the median forward total return that followed — measured only after each year’s filings were public — was:
- 1 year 119% vs the S&P 500’s 17% (price basis) · median over 5 years
- 3 years 314% vs the S&P 500’s 25% (price basis) · median over 4 years
- 5 years 1779% vs the S&P 500’s 83% (price basis) · median over 2 years
These are medians computed from the data, not a claim about any one year. The company figures are total returns (dividends reinvested); the S&P 500 is the price-only ^GSPC index, which excludes dividends and so understates the index — the gap flatters the company. A quality rating is not a return forecast, and past returns don’t predict future ones. Educational only, not investment advice.
Technology context
#14 of 744 scored Technology companies, ranked by Moat Score.
Nearest peers by Moat Score
- #12MCO MOODY'S CORPORATION92.6 out of 100, Wide moatWide moat
- #13NTAP NetApp, Inc.92.1 out of 100, Wide moatWide moat
- #15MPTI M-tron Industries, Inc.91.9 out of 100, Wide moatWide moat
- #16ROL ROLLINS, INC.91.5 out of 100, Wide moatWide moat
Compare NVDA with its nearest peers →All Technology companies on the Index →
Common questions about NVDA
- Does NVIDIA CORP have an economic moat?
- Based on its FY2026 SEC filings, the Moat Index scores NVIDIA CORP (NVDA) 92.0 out of 100 — a Wide moat. The five questions behind that score begin at 20% each; if exactly one genuinely lacks data, the other four weights are renormalized and disclosed. revenue and EPS trend 95, pricing power 93, returns on capital 100, balance-sheet safety 100, capital discipline 72.
- Is NVDA trading below the conservative owner-earnings estimate?
- The conservative owner-earnings floor is $13.81 per share. It capitalizes normalized owner earnings at a 9% rate and assumes 0% growth. Versus a recent price of $206.84, that is 1398% above owner-earnings floor. The model also publishes a capped-growth comparison, but the floor is the cautious baseline. It assumes trailing owner earnings persist and is unreliable for declining businesses. This is an educational estimate from primary SEC filings, not intrinsic value as a fact or investment advice.
- How has NVDA's Moat Score changed over time?
- The record logs 16 readings since Jul 17, 2026; the latest reads 92.0 out of 100 (wide moat). No tier changes on record yet. (Methodology or normalizer upgrades on Jul 20, 2026, Jul 23, 2026, Jul 26, 2026 re-based the score; readings across versions aren't compared.) The history is append-only — readings are only ever added, never rewritten.