Wednesday, May 20, 2026 · Afternoon
Methodology note: One-off pre-earnings setup brief. Clean pre-print snapshot before the 4:00pm ET print. Post-earnings reaction in the 7pm night briefing. All numbers are pre-print only.
Wednesday afternoon — NVDA Q1 FY27 earnings at 4pm ET, conference call at 5pm ET.
Setup Verdict — HOLD / STEP ASIDE — AWAIT THE PRINT
This is not a directional call. This is a setup frame.
NVDA is trading at $224.86, +1.7% on the day — day range $220.50–$225.30, opening at $219.62. The market has already embedded a pre-earnings bid. The options market implies ±12.9% (~$29) on the print, with front-month IV at 44–50% and IV rank at 61. Straddles are expensive and near fairly priced given the 8–12% historical average earnings move. The right posture is wait-and-confirm before sizing a directional position.
This briefing is a clean pre-print snapshot. Compare it against the 7pm night briefing to score the reaction.
Setup numbers:
- Q1 Rev consensus: $78.8B (~+79% YoY) — company guided $78.0B ±2% in February
- Q1 EPS consensus: $1.77–$1.78 adj. (+120% YoY)
- Q1 Data Center rev consensus: ~$73B
- Q1 Gross margin: consensus 74.5% vs company guide 75% ±50bps
- Q2 rev consensus: $86.08B; whisper $90B — the guide is the whole story
- Options-implied move: ±12.9%; IVR 61; positive call skew (calls at premium to puts)
Pre-Print Market Snapshot
| Level | Change | Note | |
|---|---|---|---|
| NVDA | $224.86 | +1.7% | Earnings-day bid; range $220.50–$225.30 |
| S&P 500 | ~7,464 | mixed | Softening from 7,488 morning open |
| VIX | ~18 | elevated | At/near the 18 bearish trigger |
| SOXX | ~$496 | — | Up 64% YTD; extended into the print |
| SMH | ~$537 | — | May 19 close; semi basket holding |
| 10Y UST | 4.51–4.65% | — | Range today; 16-month highs mid-week |
| DXY | 98.91 | -0.27% | Risk-on signature |
What's Priced In
The street consensus at $78.8B implies ~79% YoY revenue growth — an extraordinary number — but the bar is high because NVDA guided $78.0B itself in February. A $78B print is an in-line, not a beat. The market needs $79B+ to score it a beat; buy-side whispers are in the $80–81B range.
The more important number is Q2 guidance. Street consensus sits at $86.08B, +9.3% sequential growth. Whisper is $90B. The activation levels:
- ≥$88B: Bull acceleration narrative re-engages; above-consensus by $2B required for meaningful upside
- $85–87B: In-line; stock digests; no incremental story
- <$85B: Bear narrative activates; "peak AI capex" thesis runs
Gross margin is the quiet watch item. NVDA guided 75% ±50bps; consensus is 74.5%. Blackwell NVL configurations carry elevated CoGs during ramp. A sub-74% GM print is a structural story that overrides a revenue beat.
China remains zeroed out. The April 2025 H20 export ban produced a $5.5B charge and NVDA's Q1 guidance assumed no DC compute revenue from China. H200 re-engagement with ByteDance, Alibaba, and Tencent is approved under a 15% US revenue-share, but no material revenue has been booked. Any incremental China path is upside to numbers; any negative commentary is already in guidance.
What the Bulls Need
- Revenue beat: $79.5B+ — beat the $78.8B consensus by $700M–$1B
- Data Center: >$74B — confirms hyperscaler demand acceleration, not deceleration
- Gross margin: At/above 75% — Blackwell mix accretive, not dilutive
- Q2 guide: ≥$88B — above the $86.08B consensus by $2B; the number that re-engages the acceleration trade
- China commentary: Any incremental H200/GB200 visibility — ByteDance, Alibaba, Tencent cadence
- Blackwell: >70% of DC compute rev; GB300 Ultra moving from sampling to production confirmed
- Supply language: No supply-constraint narrative; demand-led, not supply-capped
Bull reaction: +8–12% $242–$252 zone
What the Bears Need
- Revenue miss or inline: ≤$78.5B — "priced to perfection" narrative with no beat; sell-the-news setup
- Data Center deceleration: <$72B — sequential slowdown triggers fundamental rethink
- Gross margin compression: GM <74% — Blackwell CoGs eating into margin structure
- Q2 guide light: <$85B — decelerating sequential growth; bears claim the capex cycle is peaking
- China negative: H200/GB200 path blocked or management explicitly excludes China from forward framework
- Demand pull-forward concern: Any hyperscaler commentary suggesting compute consolidation or pause
Bear reaction: -6–10% $202–$211 zone
Three Scenarios
| Scenario | Probability | Conditions | NVDA Move | NVDA Target |
|---|---|---|---|---|
| Beat & Raise | 35% | Rev $80–81B; DC $74–76B; Q2 guide $88–92B; China progress | +8–12% | $242–$252 |
| In-Line | 40% | Rev $78.5–79.5B; Q2 $85–87B; GM at guide; no China unlock | ±3% | $218–$232 |
| Miss / Light Guide | 25% | Rev ≤$78.5B OR Q2 ≤$84B; GM slip; China excluded | -6–10% | $202–$211 |
Cross-Asset Reaction Map
Beat & Raise scenario:
- SOXX / SMH: +3–6% (lagged; partial NVDA discount already in semi basket)
- NDX / SPX: +0.5–1.5% (NVDA is ~5% of NDX; megacap tailwind lifts indices)
- AMD: +2–5% (rising-tide on AI capex confirmation, though competes on margin)
- AVGO: +2–4% (custom silicon narrative tempered; NVDA win = less diversification urgency)
- 10Y yields: marginally higher (risk-on; AI capital spending supports growth premium)
- BTC: +1–3% (risk-on correlation holds in beat scenario)
Miss / Light Guide scenario:
- SOXX / SMH: -4–8% (full semi flush; AI capex uncertainty re-priced across the basket)
- NDX: -1–2% (NVDA drag pulls growth complex; 5% NDX weight amplifies the move)
- AMD: -5–8% (custom silicon story only works if AI capex is accelerating; miss hits both)
- AVGO: -2–4% (partially cushioned — custom silicon bid as NVDA alternative gets reactivated)
- 10Y yields: drop 3–8bps (risk-off; growth premium deflation)
- BTC: -2–5% (follows NDX in miss scenario; risk-off correlation intact)
Comp Setup — How the Cycle Reads Into Tonight
Hyperscaler capex in 2026 has now crossed $700B combined: Alphabet $185B, Amazon $200B, Meta $135B, Microsoft $190B — trending toward $1T in 2027. The demand signal for NVDA is structurally intact. But the comp earnings reactions were mixed:
- Alphabet +10%: Cloud +18% beat; confirmed hyperscaler AI demand bullish read for NVDA
- Meta — sold on capex: Fundamental beat but punished for $135B capital allocation market wants returns, not just spend
- Amazon — flat capex guide: No raise; management comfortable at current pacing mildly less bullish for NVDA incremental demand
- Microsoft +24% capex raise: $190B guide, explicit NVDA relationship direct demand signal
- AVGO — custom silicon growing: Meta/Broadcom 1GW+ partnership deepening the competing narrative to NVDA GPU dominance is real but not yet at scale
Net read: Capex is growing, but the market now demands ROI visibility, not just spend commitments. NVDA needs the beat and the guide to validate that the capex is flowing to GPUs, not away.
Pre-Print Trade Setups
1. NVDA Straddle — Step aside unless directional (conviction: low)
- Thesis: Implied move ±12.9% (~$29 straddle cost on $224 stock). Historical average earnings move 8–12%. The straddle is near fairly priced — not cheap, not catastrophically expensive. IV crush of 40–55% is the post-print norm. Buying a straddle here requires the actual move to exceed the implied move to profit; at IVR 61, the odds are against it. Directional bulls should buy calls only, not straddles.
- Structure: May 23 or June 6 $225 calls for bull thesis; step aside otherwise
- Invalidation: Do not buy straddles when IVR > 60 without directional edge
2. SOXX Hedge — SMH puts or short notional (for existing long exposure) (conviction: medium)
- Thesis: SOXX is up 64% YTD heading into the most binary catalyst of the semi cycle. A NVDA miss cascades through the entire basket. Buying SMH puts or shorting a small notional is cheap relative to the event risk and does not require a bearish directional view on NVDA — only a hedge against tail risk.
- Entry: SMH $535 June puts; or short 1/4 notional SMH into the print
- Invalidation: NVDA beats and SOXX gaps above $510 on Thursday open — cover the hedge immediately
3. Long NVDA / Short AMD pair — post-print activation only (conviction: medium — only valid post-print in beat scenario)
- Thesis: A NVDA beat confirms AI capex concentration at NVDA over AMD. The pair trade activates post-print: long NVDA / short AMD captures spread compression that follows NVDA beats. AMD typically rallies less and gives back faster after NVDA beat quarters. AMD is trading at a premium that assumes custom silicon share capture; a NVDA beat invalidates that premium.
- Entry: Post-print ONLY if beat confirmed
- Invalidation: AMD custom silicon revenue beat that demonstrates structural NVDA share capture
5 Questions This Briefing Wants Answered by the Print
- Q1 Revenue: Does the $78.8B consensus get beaten by ≥$1B? Or does the print miss the $80B whisper number entirely?
- Data Center revenue: Does DC beat the $73B consensus? Is the sequential trajectory accelerating or decelerating?
- Gross margin: Does NVDA print at/above its 75% ±50bps guidance? Or does Blackwell ramp CoGs compress margin to sub-74%?
- Q2 Guidance: Where does it land vs the $86.08B consensus? ≥$88B keeps the bull case alive; <$85B activates the bear deceleration thesis.
- China commentary: Any incremental H200/GB200 visibility for ByteDance, Alibaba, Tencent? Or does management permanently exclude China from the guidance framework?