Taiwan Semiconductor Q4 FY12/2025 Earnings Analysis
DISCLAIMER: This note is intended for US recipients only and, in particular, is not directed at, nor intended to be relied upon by any UK recipients. Any information or analysis in this note is not an offer to sell or the solicitation of an offer to buy any securities. Nothing in this note is intended to be investment advice and nor should it be relied upon to make investment decisions. Read our full disclaimer, here.
It’s Not About Chips
There are two easy answers to the question of “should I own TSMC stock”. The first is “of course, because AI capex”. And the second is “of course not, because the chart”. At the risk of being distinctly mid-curve on this, I think it’s quite a difficult question. And it’s nothing to do with AI capex and it’s nothing to do with valuation. It has to do with geopolitics and it has to do with Intel.
Since ever, if you wanted at-scale small feature size dies produced at high yield, you got your chips made by TSMC. The more that capital favored the fabless model, the more manufacturing edge that TSMC gained. Capital markets made TSMC a global monopolist. This was fine in a globalizing world. In a world which has resurrected the primacy of the nation-state, this is not fine. Semiconductors, the basic building block of modern economies, cannot of course be allowed by nation states which are not China to be built in Taiwan.
The attempted decoupling of the US from China across technology, manufacturing, materials and more has led to capital markets funding Intel to become a viable competitor to TSMC as a merchant fabrication shop. Success has been a long time coming, but the new “18A” plant appears to be having some success. Apple has assigned some production to take place there and there are technical achievements which look promising too. As a quasi-nationalized industry, Intel is going to see a river of capital flow its way until it can rival TSMC in both feature size and yield. Intel may fail in this, but it won’t fail tomorrow.
TSMC just printed another strong quarter. Revenue growth fell but the company is guiding to an acceleration next quarter, and they don’t tend to miss their numbers. Should the company hit a snag they have some $71bn of net cash on hand to buy their way out of trouble. (Intel is highly leveraged but it also has a kind of equity ATM facility that it can tap when needed, investors being NVDA, the Federal Government and others).
The TSMC stock chart is extended to say the least. In our normal technical analysis framework the stock is way above where one would expect it to have corrected. So between rising competition from Intel and rising threat of annexation by China, the stock should be sold or ignored. Right?
Ha. As if it were that easy.
Let’s consider the bull case.