TSMC is the Hormuz Strait of semiconductors. I moved 30% of my portfolio over today.

Yesterday I moved 30% of my portfolio to Taiwan Semiconductor, and I plan to hold it for 3-5 years. I am convinced this stock is under priced relative to other megacap stocks, for a few key reasons. 

* First, and most importantly, the geopolitical risk from China is overweighted, and currently suppresses TSM’s valuation in a meaningful way. TSM is the Hormuz strait of semiconductors. China has postured about Taiwan since the 1970s. For 50 years now, they’ve done nothing. I am not convinced China can mess with TSM (Taiwan more generally) in a way that could force the stock to drop a significant amount (50%). TSM’s global integration means that a lot of players are relying on Taiwan. It would be geopolitical suicide for China, who also rely on TSM’s chips. TSM is also diversifying locations with plants in Arizona and Europe, which are expected to be completed in the next 5 years.
* Second, TSM’s MOAT is greater than any other company in the world. TSM manufactures around 90% of the world's advanced nodes. They have decades of accumulated process knowledge that even well-funded competitors with government backing have been unable to replicate, such as Intel, who is 2-3 generations behind TSM. On top of what they’ve already achieved, they are also investing \~$40 billion in annual capital expenditure, with approximately $6 billion dedicated to R&D.  What this has led to is generating a higher portion of usable wafers, which no other manufacturer is capable of. With so much invested in RD, and their partnerships working with NVDA and Google to create these chips, they will continue to be at the bleeding edge for years to come.
* Third, and related to the second point, TSM is the backbone of the AI industry. NVIDIA, Google, AMD, and most major competitors are all paying TSM to produce their chips. There may be winners and losers, but TSM owns the sport. Apple alone generates \~25% of TSM’s revenue. They’re not going back to intel, they can’t go in house, and Apple itself is more dominant than ever. After moving to their M chips, Apple’s thesis paid off extraordinarily. I wouldn’t be surprised if other players follow.
* Fourth, relative to other Mega cap stocks (MSFT, APPLE, NVDA), the potential upside for TSM is still huge. They have the 6th largest market cap at around 2.0 Trillion. With NVDA worth \~5.2 trillion, it’s easier to see TSCM 3x than it is to see NVDA go to 15 trillion. The TTM PE of \~34x dropping to a forward PE of \~22x implies analysts expect \~35% earnings growth in the next year alone. For a $2 trillion company that's extraordinary.
* Fifth, TSM has not yet attracted significant retail investor attention relative to its fundamental quality. While dell and sandisk have gone parabolic, TSM has a lot of room to go vertical in the short term, and while it's up 2x in the last year, this treatment hasn't happened yet.

The primary risk is geopolitical. However, as TSM becomes more globally integrated, the cost of Chinese interference rises for everyone including China itself,  who also relies on TSM's chips. A Chinese move on Taiwan crashes the entire global stock market, not just TSM. This is closer to mutually assured economic destruction than a targeted risk. Relatedly, while the US has increased spending for global chip manufacturing, they are arguably a decade behind, and any movement from China in the near-term is offset by global interests from world superpowers, including the USA. 

The secondary risk is an AI capex slowdown from hyperscalers. Given that Microsoft, Google, Meta and Amazon have all raised 2026 capex guidance significantly, and inference demand is growing independently of training, I believe the AI buildout has sufficient runway over my 3-5 year horizon.