NVIDIA Corporation (NVDA)

Shares tanked today on a report that their next-gen AI chips will be delayed, but he has heard nothing to confirm this and believes that it's business as usual, that the chips will ship as scheduled. Even if the chip is delayed, NVDA is a buy at these levels because demand for the current ship is so strong.

Buy this at $105 and sell the October $120s at $8.

AI negative press is just noise. META, for example, said that it's already strategically using AI to benefit bottom line and free cashflow. Multi-decade disruption from AI, and NVDA is right at the centre. Paying ~40x for a name that's growing around 40%.

Don't buy right at the top, and don't put all your money in at once. But when there's a pullback, as there was 2 days ago, you can put some money to work. Still buyable at certain times.

Overbought. Now pulling back, and maybe that's over or maybe it's not. Can't argue with the trend, but a pullback is expected.

Since its Q1 earnings and stock split: it flirted with a $3 trillion market cap after getting too expensive, with a PE too high. June 20 was the start of a correction. Its PE has fallen to that of P&G, Coke and Colgate, when it should trade higher. Eventually, their chips will make businesses more profitable. Own, don't trade it.

He should own this. This pullback is temporary. It could go sideways for a bit, but you could see a spike as early as today if we get positive news out of the Fed.

The real, bleeding-edge innovator in the ecosystem. 

It's interesting that TSM makes NVDA chips, and MU memory goes into NVDA chips. All part of the same value chain. NVDA is at bleeding edge of technology, no direct competitor in the conventional sense. If AI is real, NVDA is the winner. And if we're not at peak capex, NVDA is the winner.

He's adding in all these caveats because the valuation's OK, but you have to believe that revenue from here and 2.5 years from now is going to double. Is that likely? He thinks it's more of a 50/50 bet.

As an investor, you should look at whether you think MSFT Copilot is going to be successful, and will it have 10-30M subs willing to subscribe? If yes, then the killer app for AI is here, and NVDA will continue to rally.

He's more neutral on NVDA today, and still owns it today, but has been trimming as capex risks starting to build up. Cost of admission to NVDA is +/- 30% at the very least. He's not chasing the recent 15% dip. Worth owning over a 3-5 year horizon, but be mindful that you can't double or triple up the weight right now by adding on tertiary AI stocks that are associated with it.

Own it, but own it in a size that you'd be willing to tolerate big, significant drawdowns and moves higher. Believe in it, but don't be over-invested.

Great opportunity to pick up 4 pillars. MU on the manufacturing, TSM for the foundry, LRCX or KLAC or ASML as the equipment suppliers, NVDA is a gift down here as a designer. And (he can't believe he's going to say this) even INTC; come 2025, it will be competitive with NVDA.

The best is behind it, now fairly valued and potentially overvalued. You can't argue with the chart. Still lots of risk in this business. You never know if there are 2 guys in a garage who can come up with better software. So many companies like AMD and INTC are working to try to compete. At the same time, big techs are trying to reduce their reliance on it.

Better ideas in software or big tech.

Likely, their strength won't last that long, because competition will emerge. For now, NVDA is cooking, doing very well. You can hold this, but don't bet the farm on it. A very good company that's growing into its valuation. Be careful.