How to Stress Test Tariffs With Supply-Chain Data
August 10, 2026
Altsets
Research by Altsets Research
Combine relationship size with explicit tariff assumptions while keeping company country, product origin, policy scope, and pass-through separate.
Data used:Altsets Supply Chain Intelligence: 90k+ entities, 400k+ relationships, 20+ years of history.
Key findings
- Relationship size provides an economic base for tariff scenarios, but company domicile alone does not establish product origin or tariff eligibility.
- A tariff calculation should keep observed relationship data separate from hypothetical rates, qualifying shares, exemptions, and pass-through assumptions.
A tariff scenario should begin with a clearly defined policy assumption, then apply that assumption only to relationships that plausibly fall inside its scope.
Supply-chain relationship data helps with the economic part of that process. It can show which supplier relationships are large enough to matter and which customers have meaningful mapped cost exposure. The Micron supplier network provides a worked example:
| Supplier | Represented country | Relationship size | Micron cost percentage |
|---|---|---|---|
| ASML | Netherlands | 3B USD | 11.91% |
| Applied Materials | United States | 990M USD | 3.84% |
| Shin-Etsu Chemical | Japan | 366M USD in the company-specific view | 1.34% |
These rows are not tariff liabilities. They are relationship inputs that can be combined with a hypothetical tariff assumption after country of origin, product classification, and policy scope are established.
Start with a scenario, not a headline
A useful tariff model needs an explicit rule. For example: What would the gross cost impact be if a 10% tariff applied to the entire qualifying value of a defined supplier relationship?
That is a scenario question, not a claim about current law. If the full 990M USD Applied Materials to Micron relationship were assumed to qualify, a simple gross scenario would be: 990M USD multiplied by 10% = 99M USD
That 99M USD is not an Altsets estimate of tariff cost. It is arithmetic applied to an Altsets relationship-size estimate under a hypothetical assumption. Keeping those two layers separate is essential.
Why relationship size is the starting metric
Tariff calculations usually require a monetary base. Relationship size provides an estimated economic scale for the supplier-customer connection. The same scenario rate applied to different relationships can therefore produce very different gross exposure.
A 10% scenario on 3B USD produces 300M USD. The same rate on 990M USD produces 99M USD. On 366M USD it produces 36.6M USD.
Those examples show why a supplier count is not enough. Three country-linked suppliers do not create equal potential cost exposure when their mapped relationship sizes differ materially.
Why company country is not enough
The most important limitation comes before the arithmetic. ASML being associated with the Netherlands does not prove that the relevant product shipped to Micron originates in the Netherlands. Applied Materials being a US company does not prove every qualifying component, subassembly, or service comes from the United States.
Shin-Etsu Chemical being associated with Japan does not prove every wafer or material tied to Micron originates in Japan. Tariff rules usually depend on product origin, classification, and policy scope. Company domicile is therefore a research lead, not a customs determination.
What additional data a real tariff model needs
A serious tariff stress test should add country of origin, tariff classification, applicable rate, exemptions, effective date, whether services are included, contractual pass-through, sourcing alternatives, inventory, and currency assumptions. The supply-chain graph does not replace those inputs. It identifies the relationships where obtaining them is worth the effort.
Customer cost percentage helps prioritize the work
Relationship size answers the gross-dollar question. Customer cost percentage helps decide which relationships may matter more inside the customer's cost base. In the Micron examples, ASML is associated with 11.91% of Micron COGS, Applied Materials with 3.84%, and Shin-Etsu Chemical with 1.34%. If all three relationships were exposed to the same policy shock, the ASML relationship would deserve earlier investigation based on displayed customer-cost share.
That still does not mean its actual tariff burden is largest. Policy scope and origin can reverse the ranking. The percentage is a prioritization signal.
Gross cost is not net earnings impact
A gross modeled cost can be absorbed in many ways. The customer may negotiate with the supplier, switch sourcing, pass cost into prices, redesign the product, reduce volume, use inventory, shift production, or receive an exemption. The supplier may also absorb part of the burden.
A 99M USD gross scenario therefore should not be described as a 99M USD earnings hit. The scenario measures first-order mapped cost pressure before behavioral responses.
Use ranges instead of one false-precision number
A better framework can show several assumptions. Using the 990M USD Applied Materials relationship as an example:
| Assumed qualifying share | Tariff rate | Gross modeled cost |
|---|---|---|
| 25% | 10% | 24.75M USD |
| 50% | 10% | 49.5M USD |
| 100% | 10% | 99M USD |
The range makes the dependency on assumptions visible. It is usually better than presenting one point estimate that hides uncertainty about product origin and policy coverage.
A repeatable tariff workflow
- Define the policy scenario precisely.
- Map supplier relationships into the affected customer.
- Rank by relationship size and customer cost percentage.
- Resolve product origin and tariff classification.
- Estimate the qualifying share of each relationship.
- Apply the scenario rate.
- Model pass-through and substitution separately.
- Keep observed relationship data distinct from policy assumptions.
- Present ranges when origin or qualifying share is uncertain.
The geographic supplier exposure guide explains the country-screening step that comes before tariff modeling. The supplier-ranking guide explains how to prioritize relationships before adding policy assumptions.
For relationship metric definitions and limitations, read the Altsets supply-chain data methodology. Browse Supply-Chain Data Use Cases for other methods.