$80B+ Q1, $88B+ Q2 guide, DC q/q re-acceleration, GM at or above 75.5%.
$236-245 first reaction; sustained break above $240 needs guide quality, not just headline EPS.
NVIDIA 这次 FY27 Q1 的核心问题:Blackwell / Blackwell Ultra 需求是否继续把 Data Center 拉到新的斜率,Q2 guide 能否覆盖市场已付出的高期待,以及管理层能否消除 China、rack-level deployment 和 2027 capex digestion 的尾部疑虑。
从 Q1 FY26 的 $44.1B 到 Q4 FY26 的 $68.1B,收入台阶已经非常陡。FY27 Q1 guide $78B 隐含继续加速,但投资者会把注意力前移到 Q2。
用这个表判断 AH 第一反应是否值得追。左列是 bullish surprise,右列是会触发 sell-the-news 的区域。
| Metric | Bullish | Acceptable | Weak | Why it matters |
|---|---|---|---|---|
| Q1 revenue | >= $80B clean beat | $78-80B acceptable | < $77B weak | Company guide midpoint was $78B, +/- 2%. |
| Q2 guide | >= $87-90B | $84-87B | < $84B | This is the real stock-moving line: demand visibility into Blackwell Ultra / Rubin bridge. |
| Data Center | >= high-teens q/q growth | mid-teens q/q | low-teens or worse | At $62.3B last quarter, Data Center is almost the whole debate. |
| Gross margin | >= 75.5% | 74.5-75.5% | < 74.5% | GM must show mix strength and no new China/export-control drag. |
| Narrative | Rubin on-track + inference monetization | Blackwell supply constrained but stable | Cooling/rack delays or capex fatigue | Market needs confidence that the 2027 cycle is not just pulled-forward capex. |
价格区间不是目标价,是财报后 1-3 个交易日的交易语言:先看 guide,再看 IV crush 后是否还能维持趋势。
$80B+ Q1, $88B+ Q2 guide, DC q/q re-acceleration, GM at or above 75.5%.
$236-245 first reaction; sustained break above $240 needs guide quality, not just headline EPS.
$78-80B Q1 and $84-87B guide; management confirms demand but does not add new upside to the 2027 curve.
$215-236 range; IV crush can hurt late weekly calls even if the print is fine.
Q1 below high end or guide under $84B, with China/export-control or rack-level bottleneck language.
$200-210 first support; supply-chain beta likely sells first: memory, optical, power/cooling, high-beta AI infra.
以 $222.59 spot 和 6.0% implied earnings move 粗略计算,市场给出的第一层反应区间约为 $209-$236。超过这个区间才算真正清掉期权定价。
过去四次财报显示:公司连续强劲,但股票反应越来越受“预期已满”和 guide 质量约束。
| Report date | Quarter | Next day | 5 trading days | Read-through |
|---|---|---|---|---|
| 2025-05-28 | Q1 FY26 | +3.25% | +5.28% | Revenue beat, but H20 charge made GM quality the debate. |
| 2025-08-27 | Q2 FY26 | -0.79% | -5.47% | Good print, market shifted to guide quality and China/H20 uncertainty. |
| 2025-11-19 | Q3 FY26 | -3.15% | -3.36% | Strong Blackwell language, but elevated expectations absorbed the beat. |
| 2026-02-25 | Q4 FY26 | -5.46% | -6.40% | Beat and raise still sold: China exclusion and capex digestion mattered. |
NVDA 这份财报会直接影响 AI infra basket 的风险偏好,但不同分支的 beta 机制不一样。
Guide quality sets the risk appetite for the whole AI accelerator complex.
Bull case helps HBM, high-IOPS NAND, nearline storage and AI server storage hierarchy.
Listen for Ethernet, Spectrum-X, NVLink, CPO/LPO adjacency and data-center networking bottlenecks.
If guide is strong but rack deployment is constrained, these names can outperform semis on capex realism.
High convexity: reacts to capex confidence, but can be punished by any hint of digestion.
| After release | What to check | Interpretation |
|---|---|---|
| 1. Headline | Revenue, Data Center, GM, EPS vs guide and consensus | 先判断有没有越过 $80B / 75.5% 这两条心理线。 |
| 2. Guide | Q2 revenue guide and margin guide | 决定 AH 反应是否能延续到第二天。 |
| 3. CFO commentary | Blackwell Ultra, Rubin, China, supply, customer concentration | 真正决定 2027 multiple 的部分。 |
| 4. Cross-asset | SOXX, SMH, AVGO, MU, WDC, LITE, COHR, VRT, CRWV | 判断是单名反应还是 AI infra beta 重定价。 |