
(SeaPRwire) – By: Lucas Caldwell
A company prints a 69% profit jump, guides for 58% growth, and the stock drops 2%. That is not a contradiction. That is the market telling you the TSMC story is already fully priced, fully owned, and one macro headline away from a crowded exit. Thursday’s close at $422.25, down $7.75, came on a day when Nvidia slid 2% and Broadcom slipped 1%. Nobody was selling TSMC’s numbers. They were selling the entire chip complex because a hot inflation print revived rate-hike fear. When your stock trades as a macro proxy instead of a company, fundamentals become background noise.
The numbers themselves were absurd. Sixth consecutive record quarter. Profit of NT$763.1 billion, up nearly 69%. Revenue of NT$1.49 trillion, up roughly 50%. In dollar terms, $4.60 per share on $46.72 billion in revenue, beating FactSet estimates of $4.46 and $45.67 billion. Gross margin came in at 66.4%, down from 67.7% in the second quarter, with fourth-quarter guidance of 65% to 67%. Q4 revenue is guided to $52.3 billion to $54.3 billion, about 58% year-over-year growth. CEO C.C. Wei called leading-edge demand “extremely robust” and pointed to agentic AI as the driver.
Now the forward picture. Management expects 44% revenue growth in 2026, slightly above consensus. They also signaled AI accelerator revenue could outrun the current mid- to high-50s CAGR guide. CFO Wendell Huang held the 2026 capital budget at $60 billion to $64 billion, but the real tell was his comment that 2026-to-2028 spending should be far above the past three years. Needham’s Charles Shi raised his target to $560 from $530 and kept a Buy, calling it a beat-and-raise quarter. The stock is still up more than 41% this year and touched record highs this month.
Here is the game underneath the game. TSMC is the only foundry on earth that can actually fulfill the AI order book, and every major customer knows it. That monopoly on execution is why the AI accelerator CAGR keeps getting revised upward. But it also concentrates systemic risk in one company, one island, one supply chain. The margin dip from 67.7% to 66.4% looks trivial until you realize overseas fabs structurally dilute profitability. Arizona, and whatever comes next, will cost more per wafer than Tainan ever will. Wei can talk about robust demand all day. The gross margin line tells the quieter truth about geographic expansion.
Then there is the Texas wildcard. The recent run to record highs followed Elon Musk’s comments about early talks on his Terafab chip project in Texas. TSMC confirmed it is exploring the state, and Wei said teams are there doing evaluations, naming Terafab directly as a customer trying to build a fab. Read that carefully. TSMC is entertaining a role in a project that could eventually compete with its own customers’ ambitions, while already stretched across Arizona and evaluating more US sites. Every dollar of that $60-plus billion budget is a bet on where geopolitics lets them build, not just where demand is.
The next twelve months will decide whether TSMC is a company being valued or a hostage being traded.
Author bio: Lucas Caldwell is a tech opinion leader with millions of followers on X/Twitter, known for sharp semiconductor market analysis, earnings-season breakdowns, and blunt commentary on the AI hardware supply chain.