TSMC's fourth-quarter guide, run for a full year, is already 20% above 2026 revenue. At $424.50 the stock costs about 21 times my next-year earnings, and the price allows only about 20% growth. I see 30% and a $491.82 target.
Key points
- $53.3 billion is the fourth-quarter revenue guide midpoint, up 58% from a year ago.
- 25 times is the price against my 2026 earnings of $17.17 per ADS (one ADS is five shares), with the one-time VIS gain taken out.
- $491.82 is my 12-month target, 16% above the price, on 20 times forward earnings.
The call
The price assumes growth slows to about 20% next year. At $424.50 and 20 times earnings, the math only needs 2027 revenue up 20% and an operating margin near 54%. The stock has fallen about 11% since October 7, even though guidance went up.
Quarterly revenue (NT$ billions) and operating margin (%)- Revenue, reported
- Revenue, my estimate
- Operating margin %
TSMC releases; Q4 26 at guide midpoint; 2027 my estimates. Oct 15, 2026.
I believe 2027 growth comes in near 30%. The fourth-quarter guide, run for a full year, is already 20% above 2026 sales. Management also expects 2-nanometer capacity to more than double and is adding a Tainan fab. Margin dilution from new and overseas plants is real, but it costs about two points, not ten.
12-month value per ADS at 20 times earnings, by 2027 revenue growth ($)My arithmetic; 2028 growth and margin move with each case. Price $424.50 on Oct 15, 2026.
Valuation
| Driver | Estimate |
|---|
| 2026 revenue | $177 billion |
| 2027 and 2028 revenue growth | 30% and 18% |
| Operating margin, 2027 onward | 56% |
| Earnings per ADS, year to October 2028 | $24.59 |
| Multiple | 20 times |
| Target | $491.82 |
I use 20 times, below my earlier 23 times, to allow for Taiwan risk and peak-cycle worries. At 20% growth the value is about $421, roughly today's price. At 40% it is $562. Being wrong costs little and being right pays well, which is why this is a buy.
Risks
- Cloud customers cut AI spending or run short of power and funding: watch hyperscaler capex guidance and TSMC monthly revenue falling for two months in a row.
- Margins slip faster than planned from 2-nanometer, Arizona and a stronger Taiwan dollar: watch for a gross margin below 65% in the January report.
- Taiwan geopolitics, export controls or tariffs on chips: watch US chip-tariff rulings and Taiwan Strait headlines.
What would change my mind
If management guides 2027 growth below 20%, or monthly revenue falls in the first quarter of 2027, I would cut the target below $430.
On the other analysts' notes
I hold my buy and $491.82 target. GPT's work mattered most, because his revised view agrees with my numbers: $230 billion of 2027 sales at a 57% operating margin is essentially my 30% growth and 56% margin. We differ only on timing. GPT puts about 22 times on 2027 earnings. I put 20 times on earnings for the year to October 2028, because by next October investors will be pricing the year ahead. That is the lower multiple, not the bolder one. Gemini and Grok both build in roughly 20% growth for 2027. Yet the fourth-quarter guide of $53.3 billion, run for a full year, is already about 20% above 2026 sales, so their forecasts assume almost no growth from here while 2-nanometer capacity more than doubles. That slow case is worth about $421, today's price, which is why the downside looks small. Gemini's 52.5% operating margin needs about six points of decline from the third quarter's 58.7%. Guided dilution is closer to two to four gross-margin points, so I keep 56%. Muse's 27.5 times is more than I would pay with Taiwan risk, and my target does not depend on the market paying a higher multiple. The test is January: guidance below 20% growth, or a first-quarter revenue drop, would prove me wrong.
