Taiwan Semiconductor MFG. (TSM)

Attractively positioned in the value chain, as it's on the foundry side producing chips for designers. Benefits no matter who gains traction. Valuation a bit ahead of itself. Better value in other spots across the value chain. Still upside to earnings, but not enough for her to be interested.

Are the most important semiconductor factory. They don't design, but make those chips. They are always sold out for manufacturing. Has dominant market share.

It reports Thursday, and it should be a monster quarter. Problem is, everyone knows that. Shares have been weak despite a steady stream of good news. 

The biggest risk is a Chinese invasion, but he owns Alibaba which operates in China. TSM is growing cash flow as EBITDA, revenues and operating income all are growing at 20% while EBITDA multiple is 11-12x (much lower than Nvidia) and PE is 20x. Earnings are much more predictable than Nvidia's. It's his largest position.

The winner in AI is really TSM, because they make the chips for both companies and Broadcom. TSM sells at 19x PE and 13x EBITDA whereas Broadcom's multiples are ridiculously high. He doesn't see China taking over Taiwan.

Backbone of chip and AI industry. Commands about 70% of global market share. Very strong forecast of ~30% earnings growth. Leadership in the 3-nanometer production space, and progressing towards 2 nanometers (just mind-blowing). That will really sustain partnerships with customers. 

Very strong structural demand for chip uses across industries, leading to long-term secular growth. Additional fabrication plants built around the globe enhances supply-chain security. Yield is 1.35%.

It is the leader in the space with about 38% of semis. The semi space is very cyclical - the demand goes way up and then starts to dry up with big inventories. It is therefore not a long term hold. 

Really benefited with the data centre buildout and AI investment. She doesn't typically invest in the semi space, as it's very cyclical. But now it's viewed as a growth sector. Valuations in the sector are quite high, not sure how long growth is sustainable.

Enables all the other companies to do what they do. Monopoly on leading-edge fabrication of chips. Margins are high. Today announced price increases. Very high ROIC. Cheap valuation of 17x PE (half that of NVDA) for the growth ahead. If you believe that AI is even modestly real, there's no debate that this name will have a role to play for many years. 

Why take the risk with an INTC, who's betting that it can leapfrog into success? TSM is already there. His anchor position in technology. Yield is 1.09%.

Not sure where it's going to go in the short term. This name is probably #1 in its industry, controlling ~2/3 of chip production. Good, solid company in a leadership position in a leadership sector.

Stock's come up on the back of growth plus multiple expansion. Now trading around mid-20x PE. Rich for him, he'd look for a pullback. If you buy now, significantly more valuation risk. He hasn't been selling, but not buying anew either. Numbers tomorrow will likely be quite strong. Releases data monthly, so there are fewer surprises.

Typically, he doesn't buy into a release. But in this case, the last monthly release was quite strong, and demand is the same. So numbers are likely to be strong. However, it also gets into what was the market expecting? If the market was expecting 35% growth, but it comes in at "only" 30%, will the stock sell off? So directionally he expects quite strong topline growth, but will that satisfy the market?

Performed very well, up 62% in last 12 months. Very important part of the AI infrastructure space. Expected 25% earnings growth rate, at only 29x forward PE. Only bad thing about it is that he doesn't own it :(  Likes it, but there needs to be a sizable pullback. Need to be aware of its geographic and regulatory issues.

In the top 10 (if not the top 5) in his fund. Already surpassed average analyst price target. With current market momentum, he's not selling, but is writing calls. Biggest foundry in the world, you have to stick with it.

Think of this name as the unsung hero of the AI conversation. NVDA gets the credit, but TSM does the hard work of making these chips at the atomic layer. The very last AI stock he'd sell. It's also the cheapest, with the most latent pricing power. Will benefit from deepening of semiconductor technology in our day-to-day lives, including in AI data centres.

Not a screaming bargain at these levels. A hold, but still decent returns ahead.