FEATURED

NVDA Is Down 17% From Its High

On May 14, Nvidia hit $236.54. A new all-time high. The company had just reported $81.6 billion in quarterly revenue, up 85% year-over-year, in what may be the most staggering single quarter in semiconductor history. CEO Jensen Huang called it the arrival of agentic AI. The board authorized an $80 billion buyback. The quarterly dividend was raised from $0.01 to $0.25 per share, a 25-fold increase. And the stock fell.

Six weeks later, NVDA is trading near $195. That is roughly 17% below the peak. The next earnings date is August 26, confirmed after the close.

So what is actually going on here?

Three things are compressing the stock simultaneously. First, GPU rental pricing. Across cloud listings, B200 hourly pricing has been quoted around a median of $6.11 across available configurations, but pricing varies widely by provider, configuration, and whether capacity is on-demand or spot. That metric functions as a directional proxy for GPU availability and near-term demand signals, not a standardized market price. Treat it as a signal, not a verdict.

Second, export controls. In April 2026, the U.S. government imposed new restrictions on H20 chip exports to China, a move management acknowledged would cost Nvidia roughly $8 billion in Q2 revenue. Then, in early June 2026, the Commerce Department moved to close a loophole that had allowed advanced AI chips to reach Chinese companies through offshore subsidiaries. This is not a new blanket ban, but the direction is clear. Nvidia’s Q1 FY2027 earnings confirmed it directly: no shipments of Data Center products to China occurred in the quarter, compared with $4.6 billion in Q1 FY2026. Jensen Huang stated publicly that Nvidia has largely conceded China’s advanced AI chip market to Huawei. That is a real number that has permanently exited the revenue model.

Third, the Vera Rubin transition. According to Nvidia’s official announcement on May 31, 2026, the Vera Rubin platform is ramping into full production across more than 350 factories in 30 countries, with production shipments set to begin in Q3 2026 and volume ramping in Q4 2026. That is ahead of schedule, and it is bullish. But it also means customers are deciding right now whether to keep buying Blackwell at current prices or wait for what is next. Vera Rubin delivers up to 10x agent throughput compared to Grace Blackwell, with initial deployments expected at AWS, Google Cloud, Microsoft, and Oracle. The spending gap that creates in the near term shows up in short-term demand signals before it resolves in the quarterly revenue line.

What Q2 Guidance Actually Says

Nvidia guided Q2 FY2027 revenue at $91 billion, plus or minus 2%. That number explicitly excludes any contribution from China’s data center compute market. Every analyst building a model around that figure is working with a China-zero baseline. Data Center revenue in Q1 reached $75.2 billion, up 92% year-over-year, driven by Blackwell architecture demand. Getting to $91 billion from $81.6 billion requires roughly 11% sequential growth on top of what was already a record quarter.

That math is achievable. But it leaves almost no room for execution risk on the Vera Rubin ramp, on HBM4 supply from SK Hynix and Micron, or on hyperscaler capex holding firm into the back half of 2026. Microsoft, Google, Amazon, and Meta have all guided substantial AI spending increases. Any reversal in that commentary would be felt directly in Nvidia’s Q2 result. The supply chain for Vera Rubin is reportedly twice the size of what Blackwell required, according to Huang at GTC Taipei, which adds a new layer of coordination risk that did not exist a year ago.

Slight tangent, but it matters: the analyst community is uniformly bullish. According to data from S&P Global covering 62 analysts, the consensus rating is Strong Buy with an average 12-month price target near $299. The high sits at $500 (Baird), and the low end ranges from $180 to $215 depending on the source. The direction is clear. The magnitude of the gap between where NVDA trades today and where analysts expect it to go is the part worth sitting with.

Options Market Read

With earnings confirmed for August 26 after the close, the options market is pricing a meaningful move. Implied volatility heading into that date is running elevated relative to recent realized volatility. The stock has declined in the majority of sessions throughout June and is now down roughly 12% for the month, a move CNBC described as entering correction territory. Volume has remained heavy, suggesting this is institutional repositioning rather than retail panic.

What is interesting is the structure. The $200 strike carries significant gamma concentration. A sustained break above $200 shifts market maker hedging flows meaningfully to the bullish side. A failure to hold $190 support risks accelerating the move toward the $170 area where longer-dated put open interest clusters.

For traders expecting the $91 billion Q2 guide to be met or exceeded, a defined-risk bull call spread targeting the $210 to $230 range for the August expiration captures the earnings event while defining maximum loss to the debit paid. If you believe the GPU rental pricing decline signals something more structural, a bear put spread anchored near current levels with a strike ladder below $185 defines the downside without the unlimited risk of a naked short.

The neutral case is actually the most crowded. Elevated implied volatility heading into August 26 makes long premium expensive, but short premium carries event risk the market is explicitly not treating as routine. Straddles centered around the $195 to $200 zone are implicitly pricing a move of roughly 8 to 10% in either direction. That is where the options market’s real opinion lives right now.

The Risk Checklist

• If B200 GPU rental prices continue declining toward $3 per hour, the revenue trajectory supporting the $91 billion Q2 guide comes under real pressure
• If any major hyperscaler softens its AI capex commentary in upcoming earnings, the non-China demand story weakens directly and immediately
• If the Vera Rubin production ramp encounters supply constraints given its supply chain is twice the scale of Blackwell’s, FY2028 revenue models face downward revision at precisely the moment the market is pricing in the full upgrade cycle
• If export controls expand beyond the current China restrictions to cover additional jurisdictions or next-generation Rubin chips, Nvidia’s total addressable market contracts in ways that are difficult to model in advance
• Consensus analyst targets cluster near $299 on average, with the stock currently trading near $195. That is a wide gap. Wide gaps cut both ways.

The stock is down 17% from its high. The fundamentals have not deteriorated. The question the market is actually asking is whether the Blackwell-to-Vera Rubin transition creates a temporary demand air pocket, or whether the upgrade cycle compounds demand rather than displacing it. August 26 will answer part of that question. The options market is charging real premium to find out. Whether the answer justifies that premium is the only trade left to make before then.