Naksha Market intelligence
All research
TSMTaiwan Semiconductor ManufacturingLeading-edge foundry

TSM at $451.25

TSMC is the most important manufacturer in AI. The valuation assumes its technological and geopolitical constraints remain profitable for a very long time.

TSMC may be the best semiconductor manufacturer in the world.

That is not the same as saying the market's expectations are modest.

At $451.25 per ADR, the market is paying for years of leading-edge dominance, exceptional margins, and AI-driven demand while discounting less of the capital intensity and geopolitical risk.

This is where investors must separate business certainty from valuation certainty.

What TSMC actually controls

TSMC manufactures the advanced chips designed by Nvidia, AMD, Broadcom, Apple, and many others. It combines leading process nodes, advanced packaging, scale, yield knowledge, and customer trust.

That combination is extraordinarily hard to reproduce.

In fiscal Q2 2026, revenue reached NT$1.270 trillion, or about $40.2 billion, up 36% in New Taiwan dollars. Gross margin was 67.7%, operating margin was 60.3%, and high-performance computing represented 66% of revenue.

Seven-nanometer and more advanced processes produced 77% of wafer revenue. Two-nanometer contributed its first 3% during the ramp.

Why does that matter?

Because TSMC is not simply adding wafers. It is capturing the economics of the hardest wafers.

Official August 2026 revenue reached NT$514.8 billion, up 53.3% year over year. Near-term demand remains exceptional.

The moat requires constant rebuilding

TSMC's advantage is deep, but it is not free.

The company spent NT$496 billion on capital expenditure in Q2 alone. First-half capital expenditure was about $26.8 billion. New fabs, two-nanometer ramps, advanced packaging, and overseas expansion consume cash before they generate it.

Overseas fabs also dilute margin. Customers and governments want geographic resilience, while shareholders want the economics of concentrated Taiwan production. TSMC must satisfy both.

That tension is the investment case.

What $451.25 requires

For TSMC I use a 10-year free-cash-flow-to-equity model rather than a terminal P/E. The moderate case starts with NT$1.4 trillion of normalized 2026 FCFE, applies growth that fades from 18% to 4% over ten years, uses a 3.5% terminal growth rate, and discounts at 11.1%.

That produces an estimated value of about $210 per ADR.

The current price is more than twice that result.

Holding the moderate growth structure constant, the market price would require normalized starting FCFE of roughly NT$3.1 trillion. That is about 2.4 times the annualized first-half 2026 free cash flow.

The valuation is not merely assuming more AI demand. It is assuming that much more of the demand becomes distributable cash after the capital program.

Three possible outcomes

Conservative: about $138 per ADR

Normalized 2026 FCFE starts at NT$1.25 trillion. Growth fades from 12% to 3% over ten years, terminal growth is 3%, and the cost of equity is 11.6%.

This case assumes the business remains dominant but overseas fabs, capital intensity, and a normal semiconductor cycle absorb more cash.

Moderate: about $210 per ADR

Normalized FCFE starts at NT$1.4 trillion. Ten-year growth begins at 18% and fades to 4%, terminal growth is 3.5%, and the cost of equity is 11.1%.

This is an excellent operating outcome. It is still far below the checked price.

Aggressive: about $328 per ADR

Normalized FCFE starts at NT$1.6 trillion. Growth begins at 24% and remains elevated for longer, terminal growth is 4%, and the cost of equity is 10.6%.

Even this case remains about 27% below $451.25.

That does not prove the market is wrong. It shows how much growth duration and cash conversion the market is assuming.

The balance sheet is strong. Geography is the unhedgeable risk

TSMC ended Q2 with roughly NT$3.52 trillion of cash and marketable securities against NT$1.03 trillion of interest-bearing debt. Net cash was approximately NT$2.49 trillion.

The financial balance sheet is not the main concern.

The physical concentration in Taiwan is.

No spreadsheet can assign a precise probability or clean discount rate to a severe cross-strait disruption. Hiding that uncertainty inside a small adjustment to the cost of equity creates false precision.

I handle it through scenario ranges and a separate margin-of-safety decision. A concentrated geopolitical risk deserves a wider margin of safety, not a prettier decimal.

What to watch out for

  • Monthly revenue - because it provides the fastest official demand signal.
  • Quarterly gross margin, operating margin, and free cash flow - because revenue growth must survive capital intensity.
  • Two-nanometer yield, customer breadth, and wafer mix - because process leadership must continue into volume.
  • CoWoS and other advanced-packaging capacity - because packaging can be the bottleneck even when wafer capacity exists.
  • Overseas-fab dilution and utilization - because geographic resilience has a real economic cost.
  • Taiwan and customer concentration - because the two largest strategic risks are not visible in a simple P/E ratio.

What could change the thesis

The valuation case would strengthen if normalized FCFE establishes a substantially higher base, overseas fabs approach Taiwan economics faster than expected, and advanced packaging expands cash returns rather than only capital needs.

The thesis would weaken if two-nanometer yields disappoint, capital spending rises faster than operating cash, customer concentration increases, or geopolitical risk reduces TSMC's ability to operate and distribute cash.

The fact that the business keeps beating estimates is important. The question is how much of each beat reaches owners after reinvestment.

Bottom line

TSMC is the indispensable manufacturer behind the AI boom.

Business quality: world class.

Balance sheet: strong.

Capital intensity: enormous and necessary.

Valuation: above even my aggressive cash-flow scenario.

At $451.25, I would not call TSMC a bad company. I would call it a stock that leaves very little room for ordinary greatness.

The practical principle: the most important company in a supply chain can still be a poor purchase if the price assumes the constraint lasts forever.

Sources

Scenario values are analytical estimates, not company guidance or price targets. Educational only - not financial advice.

Continue the library

Compare the assumption, not just the ticker.

Browse all valuation research