How Exposed Is My Portfolio to TSMC?
September 12, 2026
Altsets
Research by Altsets Research
Measure TSMC dependence beyond direct TSM stock weight by separating portfolio holdings with direct relationships, second-order paths, structural exposure, and Taiwan-specific geographic risk.
Data used:Altsets Supply Chain Intelligence: 90k+ entities, 400k+ relationships, 20+ years of history.
Key findings
- Owning no TSMC security does not imply zero TSMC portfolio exposure because holdings can depend on the foundry directly or through second-order economic relationships.
- A defensible portfolio view keeps direct quantified relationships, structural paths, and Taiwan geographic exposure separate instead of multiplying percentages through a network into one false precision number.
Measure portfolio exposure to TSMC in separate layers: direct TSMC security weight, first-order customer or supplier relationships, second-order network paths, and Taiwan geographic risk. A portfolio can have material TSMC dependence even when it owns no TSMC stock, but those different exposure types should not be collapsed into one false percentage.
That is why the useful question is not simply, "How much TSM do I own?" The safer investing question is, "How much of my portfolio depends on TSMC continuing to function as expected?"
Direct stock exposure is only the first layer
If the portfolio owns TSMC shares or its NYSE-traded ADSs, the direct portfolio weight is easy to calculate. TSMC says its common shares trade in Taiwan under 2330 and its American Depositary Shares trade in New York under TSM.
That position weight still says nothing about indirect exposure inside other holdings. A portfolio with no TSMC security can have meaningful economic dependence on the foundry through customers, suppliers, and products that sit elsewhere in the network.
The second layer is portfolio companies that depend on TSMC
The most important dependency question is whether a portfolio holding relies on TSMC for manufacturing capacity, process technology, advanced packaging, or another economically important function. That requires company-level relationship evidence rather than an assumption based on the fact that a company sells semiconductors.
A dependency-aware portfolio screen would resolve each holding, retrieve its mapped TSMC relationship if one exists, preserve the direction of that relationship, and then decide whether the edge is quantified or structural only. The result is a portfolio-specific answer rather than a generic list of companies that are "probably exposed to Taiwan."
The third layer is second-order dependence
A company can depend on another company that itself depends on TSMC. That creates a second-order path even when the portfolio holding never buys directly from TSMC.
Those paths matter because diversification can fail quietly. Two holdings can have different sectors, customers, and tickers while still relying on the same upstream manufacturing node. The second-order exposure guide explains why those paths are useful while also warning against inventing economic weights for unquantified hops.
Do not multiply percentages through the chain
If a portfolio holding has a quantified relationship with a TSMC customer, that percentage cannot simply be multiplied by another relationship percentage to manufacture a "TSMC exposure" number. Each percentage has its own denominator, period, direction, and estimation context.
A safe portfolio analysis should therefore keep direct quantified exposure separate from structural second-order exposure. The investor can rank and flag those paths without pretending that a multi-hop graph produces one exact portfolio percentage.
TSMC itself shows why the node matters
TSMC's 2025 annual report says the company served hundreds of customers and produced 12,682 products across 305 process technologies. Its 2025 Form 20-F also says the ten largest customers accounted for about 78% of net revenue, while the largest and second-largest customers represented 19% and 17%, respectively.
Those facts do not reveal your portfolio's exposure to TSMC. They show why TSMC sits inside a large and concentrated commercial network whose disruptions, capacity decisions, and technology transitions can matter well beyond the TSMC security itself.
Geographic diversification does not automatically solve TSMC dependence
A U.S. stock, a European stock, and an Asian stock can still rely on the same foundry. Country diversification can therefore coexist with manufacturing-node concentration.
TSMC is also expanding its manufacturing footprint outside Taiwan, including Arizona, Japan, and Europe. That means "TSMC exposure" and "Taiwan exposure" are related but not identical questions. A portfolio risk model should distinguish dependence on the company from dependence on one geographic production location.
A useful TSMC portfolio view has several buckets
A dependency-aware portfolio can classify holdings by direct security exposure, direct company relationship exposure, second-order network exposure, and uncertain or unquantified structural exposure. Those buckets should remain separate because they carry different evidence quality and different implications for portfolio construction.
The goal is not to collapse every path into one dramatic number. It is to discover whether several supposedly independent holdings rely on the same foundry node and whether that concentration is large enough to change position sizing, hedging, or stock selection.
The safer investing decision comes after the map
Once the TSMC-connected holdings are visible, the investor can decide whether to reduce one position, add holdings with different manufacturing dependencies, cap exposure to the shared node, or simply monitor the concentration more deliberately. The relationship map does not prescribe the trade.
This is where dependency awareness differs from ordinary geographic diversification. The investor is diversifying the economic machinery underneath the portfolio, not just the labels attached to the securities.
The supply-chain diversification guide explains why sector and country labels can miss common dependencies. The custom-index risk-limit guide shows how relationship constraints can be incorporated into portfolio construction once the dependencies are measured.
For relationship definitions and evidence limits, read the Altsets methodology. The Altsets documentation explains the available interfaces for portfolio-scale relationship research.