Key insights
- Despite NVIDIA reporting record revenue and strong guidance, the stock has declined post-earnings. This is attributed to extremely high ownership among active funds and its significant S&P 500 weighting, limiting incremental buyers. While consensus estimates increased, the market awaits the official inclusion of Vera CPU in financial filings for a substantial model revision and potential stock catalyst. The growth split between ACIE and Hyperscale also presents a potential bear case if custom silicon displacement falters.
The short answer: everyone who can buy it already owns it, and the revenue lines that would force a model revision haven't printed in a quarterly filing yet.
The long answer: On May 20, NVIDIA reported $81.6B revenue, up 85% YoY. $91B Q2 guide. $48.6B free cash flow in a single quarter. Stock was at $220 pre-earnings. Sitting at $200 now.
Before anyone says priced in, I think this one is more specific than that.
On the call Jensen said verbatim:
"The $20 billion is for stand-alone CPU."
Then separately when asked what sits above the $1 trillion Blackwell and Rubin demand signal:
"we didn't include any Vera CPU, stand-alone CPU in that number. And so I expect that to be the second largest."
FY2028 consensus moved from $491B to $551.7B after the print. That captured the $91B guide extrapolated forward. It did not build a discrete Vera CPU line because sell-side models require a historical 10-Q line to anchor from. Vera CPU has never appeared in one. A blank cell requires an assumption with no data behind it.
The ownership structure makes the stock unresponsive even to this. Consensus moved up $60B. 36 of 37 analysts have Buy ratings. Price targets raised across the board. Stock still down 10%. At 8.3% of the S&P 500 and 78% active fund ownership, every fund that can own it already does. The incremental buyer is index flows and retail. Neither reads the transcript.
On the bear case: the Q1 10-Q introduced the ACIE vs Hyperscale split for the first time. ACIE $37.4B, Hyperscale $37.9B. ACIE grew 31% QoQ, Hyperscale grew 12%. The custom silicon displacement thesis requires Hyperscale dominant and accelerating. Two numbers say the opposite.
So what actually moves the stock? The upgrades since May 20 are guide extrapolation. The upgrade that moves the stock is when Vera CPU appears in a 10-Q and every analyst rebuilds their model around a segment that's never existed before. That's a structural revision. Q4 FY2027 at earliest, Q1 FY2028 more likely.
Full writeup with primary sources here: https://substack.com/home/post/p-203242044
Pushback welcome.