Equity due-diligence report

NVDA

Version date
July 10, 2026
Generated
Jul 10, 2026, 13:12 UTC
Freshness
Stale · 1mo

LiveMarket quote

NVDA$323.32−$2.71 (−0.83%)today$306.97$324.46day rangeretrieved Aug 20, 2026, 05:31 UTC

Market state at retrieval — independent of this report’s as-of research inputs recorded below.

Source data

4 inputs · freshness

  • companyfacts.json1 month oldas of Jul 10, 2026, 11:14 UTC
  • filings.json1 month oldas of Jul 10, 2026, 11:14 UTC
  • market.json1 month oldas of Jul 10, 2026, 11:14 UTC
  • history.json1 month oldas of Jul 10, 2026, 11:14 UTC

TL;DR

TL;DR

At $202.78 the market values NVDA at a reconciled EV of $4,906.8B (mcap $4,911.5B + total debt $8.470B − cash&equiv $13.237B, at 2026-04-26; the provider EV $4,543.0B is rejected as a $368.5B-phantom-net-cash artifact). That EV capitalizes NVDA's already-printed $214B annualized-Q1 operating-income run-rate at ~22.9× — the price is not paying for an acceleration, it is paying full multiple for the current run-rate. The whole debate reduces to one variable: the terminal multiple a decelerating-but-still-growing NVDA commands, first testable at the Q2 FY27 print.

  • Bull. Strip the entire $15.936B Q1 investment mark (27.3% of net income, which pushed net income above operating income) and the case does not weaken: operating income $53,536M grew +147.4% YoY and sits below the mark, so the operating engine is already clean of it and out-earns even the gain-stripped net income. On the metric that discounts growth NVDA is the cheapest of its AI/data-center-semi peers (fetched forward P/E 20.40 vs AMD 62.24 / AVGO 25.47 / TSM 22.30; PEG 0.45 lowest), the China cliff has already fallen to 5.6% of Q1 revenue and been absorbed while total revenue grew +85.2%, and the concentrated buyers are the best-capitalized capex programs on earth (Big-Four 2026 capex ~$630B, +~62%).
  • Bear. Concede all three — supercycle, China-immaterial, clean operating growth — and the residual thesis still stands: the price carries no growth cushion at ~22.9×, so conceding continued operating-income growth, a single-variable multiple normalization prices NVDA −20.5% (decelerate to +21% at 15×) and a digestion year −47.6% (flat at 12×), versus +10.0% upside that needs +40% growth and an 18× premium to both hold. The $15.936B mark is not diversification but correlated leverage — the marketable book ($30,237M) and the customers' buying power move on the same AI-sentiment variable, and NVDA's own 54%-of-revenue three customers are financing the TPU/Trainium/ASIC capacity that attacks its ~74% gross margin.
  • Priced-in break-even. At the adopted EV $4,906.8B the market underwrites $245.34B/yr of forward operating income at a 20× terminal multiple — +14.6% above the $214B run-rate. It is priced-in-fair iff forward operating income clears that threshold and the multiple holds; rich if either fails. The single assumption the verdict is most fragile to is the investment-mark contribution to earnings: on clean operating income the EV is 22.9×, bank the $15.936B mark and it looks 21.0×, strip it and it is 28.9× — a ~6-turn swing from composition alone. Versus the pre-print baseline the price is −8.1% (drawdown has conceded part of the premium); versus the autumn-2025 base +7.8% (the multiple has compressed through earnings growth). Below the last print, above the pre-breakout base — a two-sided fact, not a call. No buy/sell/hold.

Watch next — a cold reader can act on these:

WindowEventIf it lands bullishIf it lands bearish
~2026-08-26, after market (est.; NVDA-unconfirmed)Q2 FY27 earnings (8-K + 10-Q) — both cases' falsifier #1; resolves the growth vector, the reflexivity ordering test, and gross margin in one printOperating income grows YoY, gross margin ≥ ~72%, operating income ≥ net income (no fresh mark carrying the quarter)Operating-income growth decelerates, GM slips below ~72%, or a fresh GainLossOnInvestments again inverts the ordering
~late-Jul / early-Aug 2026Big-Four 2026 capex guidance (MSFT / AMZN / GOOG / META Q2 prints) — the supercycle-vs-digestion testAll four raise FY2026 capex; none flags a pause — the ~$630B demand pillar holdsAny of the four cuts capex guidance below the ~$630B trajectory — the single-customer air-pocket risk crystallizes
Open; checkpoint at the Q2 printH200-to-China revenue realization — licenses granted Feb 2026 but $0 realized for ~5 months (25% import tariff + a ~15% USG revenue-share, un-codified)First China H200 revenue $ disclosed in a 10-Q at workable margins — un-modeled upside turns onRevenue stays near zero, or a fresh BIS/Federal-Register B-series restriction lands (live-sourced) — China inverts to downside
~late Feb 2027FY2027 Form 10-K (FY ends 2027-01-31) — full-year read on margin/share through the AMD MI400 and Broadcom/Anthropic-TPU ramp, and on China termsGross margin ≥ ~72% and DC-GPU share ≥ ~80% held through the competitive rampMargin compresses / share slips below ~70%; concentration % rises; China terms confirmed low-margin

What the numbers sayLocked

This section is locked. Unlock NVDA to read it.

Bull caseLocked

This section is locked. Unlock NVDA to read it.

Bear caseLocked

This section is locked. Unlock NVDA to read it.

Cross-examination ledgerLocked

This section is locked. Unlock NVDA to read it.

Is it priced in?Locked

This section is locked. Unlock NVDA to read it.

Catalyst calendarLocked

This section is locked. Unlock NVDA to read it.

Sources appendixLocked

This section is locked. Unlock NVDA to read it.