Micron Technology (MU)

It's too late to buy it. It up 219% this year. MU is seeing a massive uptick in earnings growth due to price increases. This situation rarely lasts after a couple of years when peers compete and become suppliers. Also, the LLMs and data centres will eventually find ways to use less memory. All it takes is for someone to say that they have found 20% more memory efficiency and MU stock could be $700 in two seconds.

Buying calls is likely the right strategy, for the near-term return. Don't buy long term or you will face the inevitable correction. It's trading at 60 million share daily, well above its average of 15 million. Options interest continues to grow. There's excessive speculation here, for sure.

It took Nvidia years to reach a trillion dollars, but Micron only 48 days. This screams it's meme-ish, is in the wrong hands, not in thoughtful hands. You could still make money buying calls, but MU has no investment thesis, just a speculation one. Still trades at 10x 2027 PE because there's a backlog. Speculation, not an investment.

Huge shortage. Sold out for next 1-2 years and can charge what they like. Wildly cyclical. Hard to sell something when the story's so good, but he'd be reducing over the next little while.

Chips, and mostly memory (which are in short supply). Making tons of money, but it's cyclical. You go from supply shortage to supply glut, and the stock rolls over. We're not there yet, but he'd be cautious. Good company, but skewed toward risk at this point.

He prefers more broadly diversified companies.

Continues to work, but a bit extended. If owned, perhaps trim a bit. Anticipating a consolidation phase. Overall, positive technical profile. To add, wait for a pullback.

It trades under 12x PE, but pulled back 6% today which may be the time to buy this. Buy some now, and more after a 2-3% decline.

A lot of these charts are tough to buy, but lots more room to run. When you think about the AI compute, it's not just the GPU anymore. We're getting all these optimizations across different parts that have to roll into data centres.

Focused on revolution of agentic AI. To reach solutions, it has to be heavily integrated and remember the different components. Its high-bandwidth memory is sold out through 2026, with supplies remaining tight potentially into 2028. 

You have to decide if this is still a cyclical company, or is it structural as we need a lot more memory going forward. His vote is for structural. Yield is 0.08%.

He's taken some shares off the table. Is trading at 7x PE, but historically it's 6x. They have contracts through 2028. At some point, this will reset, but it's too early for it.

Another beneficiary of AI. Look at that chart -- what's the likelihood that it'll double from here? Probably not very high. Take profits. Risky if you're looking to squeeze out another 10-15%.

This can keep rallying because there's endless demand from the data centre companies and not enough supply. He targets $1,000. It remains cheap on a PE basis, now only 6x PE.

He just added more shares. Next year, this looks like a commodity valuation. You can't ignore memory. Is up 133% in the last 6 months.