How Investors Can Use Supply-Chain Data Before Buying a Stock
July 20, 2026
Altsets
Research by Altsets Research
Use quantified supplier and customer relationships to add upstream dependency, downstream demand, substitution, geography, and contract checks to individual-stock due diligence.
Data used:Altsets Supply Chain Intelligence: 90k+ entities, 400k+ relationships, 20+ years of history.
Key findings
- Micron's visible network shows material exposure on both sides of the company, including ASML at 11.91% of Micron's cost base and Nvidia at 17.62% of Micron revenue.
- Supply-chain due diligence is most useful when the relationship changes the next research question, such as supplier replacement difficulty, customer dependence, end-market overlap, geography, or contract durability.
Before buying a stock, use supply-chain data to identify what the company depends on to operate, which customers drive demand, and whether those relationships repeat dependencies already in the portfolio. That exposes risks and concentration a sector label or the company's own financial statements may not show.
Micron provides a useful two-sided example. In the supplied Altsets data:
- ASML is associated with 11.91% of Micron's cost base.
- Nvidia is associated with 17.62% of Micron's revenue.
Those percentages describe different sides of the same business. The first is an upstream dependency. The second is a downstream dependency. That is the core value of supply-chain research before buying a stock: it reveals where operating and demand concentration can sit outside the company's own financial statements.
Start with the two sides of the company
A normal stock screen often starts with revenue growth, margins, valuation, sector, and country. A supply-chain screen adds an upstream question about equipment, materials, components, and services, plus a downstream question about customers and end markets. For Micron, the visible network contains semiconductor-manufacturing suppliers upstream and Nvidia plus server-hardware companies downstream.
That makes the company more than a ticker inside the semiconductor sector. It is a node between capital-intensive manufacturing inputs and several computing-demand channels.
Question 1: Is one supplier economically important?
Customer cost percentage helps answer whether a supplier relationship is large enough to deserve attention. The displayed ASML relationship represents 11.91% of Micron's cost base. Lam Research is 5.52%.
Applied Materials is 3.84%. KLA is 2.84%. Shin-Etsu Chemical is 1.34%.
This does not tell you which supplier is easiest to replace. It tells you where the known economic exposure is concentrated. The supplier-ranking guide goes deeper on that distinction.
Question 2: Can a customer move the supplier's results?
Supplier revenue percentage answers the opposite-side question. Altsets associates Nvidia with 17.62% of Micron revenue. A demand change at a customer with that level of mapped exposure belongs in the investment thesis.
It still does not mean Micron revenue moves one-for-one with Nvidia demand. Pricing, product mix, capacity, contracts, and other customers all matter. The Micron-Nvidia demand analysis focuses on that relationship specifically.
Question 3: Is the dependency operationally hard to replace?
Economic size alone is not enough. ASML sells lithography systems. Lam Research sells equipment used in deposition, etch, cleaning, and related wafer-processing steps.
KLA focuses on inspection, metrology, and process control. Shin-Etsu sells semiconductor materials including silicon wafers and photoresists. Those functions are not interchangeable.
A smaller relationship can create a larger operational problem if the input has few substitutes or takes a long time to qualify. The operational-criticality guide separates financial size from replacement difficulty.
Question 4: Does the company have hidden end-market exposure?
A customer list can hide the economic activity underneath it. Micron's downstream graph includes Nvidia and structural relationships with Inventec, Quanta Computer, and ASUS. Those companies participate in AI, server, cloud, and computing infrastructure.
That does not mean every Micron sale to those companies is an AI sale. It does show why a relationship network can reveal end-market clusters that are not visible in a flat sector label. The Micron downstream-network analysis treats that as a separate research problem.
Question 5: Does geography change the risk?
Company headquarters is only one geographic layer. A supplier can be headquartered in one country, manufacture in another, and serve an end customer in a third. Before buying a stock because it appears geographically diversified, an investor should ask where production occurs, where critical suppliers operate, where the direct customer is based, where final demand is generated, and whether trade rules depend on product origin. The customer-headquarters versus end-demand guide and geographic supplier guide handle those questions separately.
Question 6: Are the relationships durable?
A relationship percentage is a snapshot of economic exposure. Contract structure can change how durable that exposure is. A long-term volume agreement can improve visibility.
It can also lock the supplier into capacity, pricing, and customer concentration. Customer deposits can fund capacity before revenue is recognized. Purchase commitments can support suppliers even when short-term demand changes. The contracted-demand guide explains why relationship size and contract quality are different layers.
Supply-chain data is most useful when it changes the next question
The goal is not to collect more supplier names. The goal is to identify which research questions become material. For one stock, the network might reveal customer concentration.
For another, it might reveal a hard-to-replace supplier. For another, it might reveal that several supposedly diversified holdings depend on the same economic cycle. That is why the useful unit is not a generic supply-chain score. It is a specific dependency that changes what the investor should investigate next.
Visual map, LLM, and API are different ways to interrogate the same data
The underlying relationship should not change with the interface. A visual workspace is useful when the investor needs to see the surrounding network and discover unexpected counterparties. An LLM connected through MCP is useful for iterative questions such as comparing several relationships, tracing a disruption, or turning a network observation into follow-up research.
An API is useful when the analysis needs to be reproducible across many companies or historical dates. The Altsets documentation covers the available data interfaces. The free workspace is useful for seeing how a quantified relationship sits inside a network before building a larger research workflow.
A compact pre-purchase checklist
Before buying an individual stock, supply-chain research can add six checks:
- Which known suppliers represent the largest share of customer costs?
- Which customers represent meaningful supplier revenue?
- Which relationships are difficult to replace?
- Which end markets sit behind the named counterparties?
- Which geographic layer matters to the risk being modeled?
- Which dependencies are supported by contracts, commitments, or other public evidence?
The answers do not replace valuation or financial analysis. They reveal economic dependencies that those analyses can otherwise miss. For metric definitions and evidence limits, read the Altsets supply-chain data methodology. Browse Supply-Chain Data Use Cases for narrower research questions built from the same relationship data.