Capex is not bad. It's good. These companies have been printing money for years. They see the huge AI opportunity and are investing to control this future and its profits. This is what good businesses do... $GOOG
Today was about China, beans, and a lack of any hint/sniff/announcement US Ag purchases. 🌱 down nearly 3%, slipping back below $12 after tagging a 2 yr high yesterday.
The culprit? 💨 The US-China summit delivered zero fresh ag deals…
No new Chinese purchases. No surprise commitments. Just silence.
🇨🇳 China signaled it won’t go beyond its existing 25MMT pledge and why would they? Brazil is cheaper and loaded with supply.
📉 US ag exports to China already collapsed from $24.4B in 2024 to just $8.4B last year. The trend is not your friend.
⚠️ Funds were massively long coming in. When the summit headline fizzled, the exit doors got crowded fast.
USDA still sees ending stocks tightening to 310 mbu vs 340 supportive longer term. But right now?
📣 The market bought the rumor. Thursday it sold the news hard.
#ThursdayNotes 📝 #Fluff
$NVDA Conference Call Summary: AI bubble? We see it differently. Everything we’ve been saying is coming true, and we are more successful this year at this point than we were last year. There are three massive platform shifts occurring at once, and for the first time since the dawn of Moore’s Law, we are uniquely addressing each of them. CPU general-purpose computing is transitioning to GPU-accelerated computing. Generative AI is transitioning existing applications while enabling entirely new ones. Also, new AI systems, capable of reasoning, planning, and using tools, mark the next frontier of computing. We are still in the early innings, and we are preparing for significant growth ahead. What Nvidia is doing has never been done before, and we’ve created a whole new industry. We feel good about our ability to execute against our opportunity set. We had a record quarter of sequential revenue growth of $10 billion, and we currently have visibility to half a trillion dollars in Blackwell and Ruben revenue in 2026. There is probably upside to those numbers, too. We expect to capture much of the $3 to $4 trillion in annual AI infrastructure build by the end of the decade. Demand for AI infrastructure continues to exceed expectations, and all the clouds are sold out. We aren’t counting on any revenue contributions from China data centers, and despite input costs rising, we expect to maintain margins in the mid-70s range next year. The world doesn’t have an excess of power, so choosing the right architecture is incredibly important, and your performance per watt translates directly to your revenues, which is why so many choose us. Nvidia provides the best performance per watt. Once people recognize that the cost-reduction from GPUs because Moore’s Law has slowed, the revenue-boosting from better recommender systems, and the net-new revenue from gen AI applications, they’ll have more appreciation.
https://t.co/xXHIY4WTCg