Gene Munster

Big-picture takeaway from $GOOG, $MSFT, and $META: The mojo of the AI trade took a step back, while the fundamentals took a step forward. https://t.co/ogxRthLtX3

What does $GOOG Search want to be when it grows up? The question came up at the end of the call and is relevant given the company is juggling four surfaces Search, AI Overviews, AI Mode, and Gemini. Sundar describes this as a “dynamic, expansionary moment,” because Search is evolving and engagement is increasing.. He adds "over time" changes are coming across Search, AI Overviews, AI Mode, and Gemini, users can expect the experience to become simpler. My take: The Search results page still feels like a high school science project gone amok. It sounds like Google recognizes that, and over the next couple of years wants to slowly make it si…

On the call, $GOOG management said AI Mode has 75M users in the U.S. They’re testing ads, but it’s too early to share details about those tests. My take: It’s only a matter of time before AI Mode ads are fully turned on. When they are, it will likely be a net positive for revenue growth even though some users may feel the ads clash with the spirit of generative AI providing a “real” answer.

Sundar says Waymo will add new experiences next year, possibly around personalization. It’s a good example of an area where Waymo has a head start over Tesla’s robotaxi offering. $GOOG $TSLA

$GOOG first question is what impact does agentic AI have on search? Answer: They think agentic is additive to search. My take: They're doing a great job of turning chat bots into search revenue today. Down the road, the playbook will get re written with agentic which is another hurdle.

$GOOG still expects tight capacity for Google Cloud in December and throughout 2026. That means whatever growth they report could have been higher if they had more capacity.

Sundar kicks off $GOOG, saying growth rates of queries accelerated in the September quarter compared to June. "AI is driving meaningful query growth because people are increasingly realizing they can get their questions answered."

$META call just wrapped up. I think I figured out why $META stock is down 9%. Expenses are growing faster than revenue, driven by third party cloud. The last two years it’s been the flip story. March revenue up 16% vs. expenses up 9%, June revenue up 22% vs. expenses up 12%. September revenue up 26% vs. expense up 32%. Next year it's going to be more like 18% revenue growth and 35% expense growth.

It looks like Zuck is most bullish on AI compared to Nadella, Jassy and Pichai based on Capex and spend on third party cloud for next year. $META, $GOOG, $AMZN, $MSFT

$GOOG guided 2025 Capex up by 7%. As expected. Looks like Google is setting up to grow Capex next year by 25% vs. $META up 60%.

Good news for $NVDA. Street is looking for $META to increase capex next year by 40%. My guess is it will be up 60%. They will give more detailed guide for next year in three months. All they said was it would be will be "an aggressive and significant" increase.

$META guided revenue up 3% for December (looking at the high end of the range). The guide would have been up around 4% if not for the revenue timing around Reality Labs.

$META Capex is going up for next year. CFO says compute needs have expanded meaningfully, and we will aggressively invest through both our own infrastructure and third-party providers. Capex growth will be notably larger, and total expenses will rise more quickly given increasing cloud costs. The stinger is the expense growth rates will increase because of cloud. Not sure who they will purchase the cloud from.

Metric on how addictive $META products are. They say there was double digit growth in US time spent on Facebook/Instagram y/y.

A small but powerful example of why $META is next to impossible to slow. Daily users were up 7.6% y/y reaching 3.54B people. Thats 43% of the world population. . More impressive is that y/y growth has been accelerating. Sep- 7.6% Jun - 6.4% Mar- 5.9% Dec 24 - 5.0% Sep 24 - 4.8%

$META call is live. Zuck says he will overbuild for the most optimistic case around Superintelligence. That means they will likely spend at a rate faster than Google, Microsoft and Amazon. That’s also how the Street is modeling our Capex for those companies with Meta up around 40% next year, Microsoft up 20%, Google up 13% and Amazon up 10%.

All of the important updates for $MSFT will come on the conference call guide at about 6p ET tonight.

I also don't understand why $MSFT is down 3%. Azure is up 39% (constant currency) vs Street at 37.5%. Unchanged from June which was 39%.

I don't’ understand why $META is down 7%. They effectively guided Dec revenue up by 3%. The actual mid point was $57.5B vs Street at $57.25. More importantly, the high end of the guide was $59B, which is likely where the revenue will come in for December, that’s where the 3% guide up comes from. That means growth in December will be around 22% compared to 26%. That's the same top line growth in June.

$GOOG Search was up 14.5% vs. 11.7% in June.

$GOOG Cloud up 33.5% vs. 29.5% estimate. Also an acceleration from 31.7% in June.

5 Keys to $GOOG, $META, and $MSFT Earnings Tonight. 1. GCP growth of 29.5% vs. 31.7% in June. 2 Azure growth of 37.5% vs. 39% in June. 3. GOOG Search growth of 13% y/y (whisper), up from 11.7% in June. 4. GOOG Capex for CY26 above current expectations of 13%. Meta is already at 42%. 5. MSFT & META Capex growth for CY26 remains around current estimates of 22% and 42%.

Looks like the more affordable $TSLA may end up being the Cybercab with a steering wheel. Bloomberg interviewed Robyn Denholm, the chair of Tesla’s board. https://t.co/WMfvefaQVo

We’ve been tracking AI Overview ad load, and its going up. Good news for $GOOG Search revenue. https://t.co/7cqV814c6v

I believe Jensen is suggesting about 15% upside to revenue over the next 5 quarters. The stock’s up only 4% on that likely because there are enough moving parts to keep investors from taking his comments to the bank. $NVDA https://t.co/NBF92YLpvm

Here's the math on how much $NVDA believes they'll beat the Street over the next year. Bottom line: Jensen said that over the next five quarters, Nvidia already has enough Blackwell and Rubin demand to exceed Street estimates by 15%. This assumes capacity is available and adding back the networking business of $41B. That upside will likely grow as they continue selling. My take: Supply is a critical factor here. if they can make them, it sounds like they're on track to beat the current Street numbers for CY26 by more than 20%. If they don't make enough, that's good news for CY27. Here's the math: 1. Jensen says cumulative demand for Black…

$NVDA is likely going higher. Jensen says: “Seeing extraordinary growth in GB, and now have visibility into $500B in cumulative Blackwell and early ramps of Rubin revenue through the end of CY26, business on the books.” Have already shipped 6m Blackwell's in the first 3.5 quarters of production.

Jensen makes his case for Nvidia vs. competition, highlighting a chart what shows GB vs. H20 cost per token is 10x lower. My take: For the past two years investors have been concerned about a cheap competitor competing with Nvidia and we just don't see it because the hyperscalers think about total cost of ownership. $NVDA

Jensen says the most important AI dynamic to understand is AI is not just a tool; it's work that will augment labor. My take: Jensen is pushing us to rethink the paradigm of how transformative AI can be. This is similar to how Elon talks about Optimus. $NVDA, $TSLA

Quantum: Nvidia’s GPUs will help advance quantum computing by improving control and error correction, two major barriers to quantum scalability. Most quantum companies are participating, which likely explains why those stocks are up slightly. My take: Quantum has been a recurring headache for Nvidia, given the narrative that it’s only a matter of time before quantum computing renders GPUs obsolete. Today’s announcement signals to investors that Nvidia intends to play a role in the quantum ecosystem rather than be displaced by it. $NVDA