NVIDIA Corporation (NVDA)

The poster child. Gone up a lot, and its valuation is probably 40x forward PE. Quite expensive, unless you believe that it can maintain the treadmill of that kind of growth. He's not saying the growth is over, just that maybe the growth slows down from here. Perhaps the valuation has to stay here while earnings catch up, or it has to come down a little bit.

See his Top Picks.

If the company continues to execute, it could double its $4.4 trillion valuation. Forget the skeptics.

It just reported a top and bottom line beat and reiterated its full-year forecast, but data centre sales came in a little light, so the stock got dinged after hours. Mostly, the stock was up 35% for the year going into the print and it was priced to be perfect. Overall, a good quarter, but didn't address the conundrum that megatech companies are getting the bang from their buck in buying NVDA stuff. China is another issue. Own this, don't trade it.

The AI hype is way overextended, like the dotcom era. NVDA's market cap is bigger than all of France or the UK or Canada. Ridiculous. If you own it, trim and take profits. NVDA is only one component of AI with software being another. Why is NVDA soaring and the other components are lagging? Is a recipe for disaster.

They report later this week. In some ways, they will hit it out of the park, but how much sentiment is built into the stock already? NVDA has had a huge run, doubling since April. He suspects some people are over-invested and few who are under-invested. No, he won't buy it ahead of earnings, unless you're a daytrader looking at options.

Sure, there are concerns including their demand in China, and will the hyperscalers cut back on buying NVDA chips? You can't have gen-AI without NVDA chips. The new NVDA chips allow chatbots to reason--reasoning will be the holy grail of the AI generation. Selling ahead of this week's report is wrong. 

She has taken profits many times. It reports earnings on Wednesday and she doesn't like to buy before earnings. Expectations are high and valuation is a sticking point. She sees an upside potential of 8% and analysts are changing to a hold. She still likes the AI theme. Be patient if buying.

He'd still buy here. Always fireworks around quarterly earnings. It's run up, but fundamentals are outstanding. Clear technical superiority in AI chips. AI infrastructure buildout will keep order books full for a long time. Major beneficiary of AI capex.

No contest. This is the one to own.

Marries hardware and software and applying massive improvements to push product cycle forward. AI wouldn't exist if not for NVDA. Capex cycles do end, so it won't be the honeymoon period forever. But we're only in 4th-5th inning, much more growth to go on the capex front. Still underbuilt, undersupplied, AI still taking off. 

Earnings are coming up, and he thinks they'll beat and raise again. Yield is 0.02%.

They report Aug. 27. AI will remain a tailwind for productivity and economic growth. However, he liked NVDA when it was below $90 last April, but not at twice the price. It costs too much now, though he likes the company. Technically, it needs to fall to at least $150, in the range of $125-150.

They have a moat in terms of chips that run AI software, which keeps the #2 far behind.

It's good that they can now sell those chips to China (though pay a tariff and will result in an earnings hit). NVDA has had a lot of momentum.