Do You Need to Be a Quant to Invest With Supply-Chain Data?
August 6, 2026
Altsets
Research by Altsets Research
Usually not. Most investors can use supply-chain data without factor models, backtests, or optimizers; simple relationship comparisons, overlap checks, and directional exposure analysis are enough for many decisions.
Data used:Altsets Supply Chain Intelligence: 90k+ entities, 400k+ relationships, 20+ years of history.
Key findings
- The displayed HPE customer set can be ranked with simple arithmetic: Microsoft represents about 59.84% of the four visible USD relationship sizes without requiring a predictive model.
- Advanced methods become useful for backtests, factors, and portfolio optimization, but they are optional extensions rather than prerequisites for using dependency data.
Usually not. Most investors can use supply-chain data without building factor models, backtests, or portfolio optimizers. Simple customer and supplier comparisons, overlap checks, and directional exposure analysis are enough for many investment decisions.
The first useful questions are usually directional: who matters to this company, how much does the relationship matter, what other holdings share the same dependency, and what event could make that relationship important?
The simplest use case is ranking attention
Consider the four HPE customer relationships visible in the supplied Altsets data: Microsoft at 561M USD, Swisscom at 203M USD, Home Depot at 109M USD, and Volkswagen at 64.5M USD. An investor does not need regression analysis to see that Microsoft deserves more research attention within that displayed set.
The four visible relationships total 937.5M USD, and Microsoft represents about 59.84% of that displayed total. That is a simple descriptive conclusion, not a forecast of HPE's stock price and not an estimate of Microsoft's share of total HPE revenue.
The value comes from knowing where to look next.
Basic arithmetic can be enough to improve a decision
Many useful portfolio questions only require comparison. Which customer relationship is largest? Which supplier appears across the most holdings? Which new stock repeats a dependency already in the portfolio? Which external company is connected to several positions?
Those questions can change position sizing, diversification, and monitoring without requiring a factor model.
The data becomes quantitative only to the extent the question benefits from measurement.
Network direction matters more than mathematical complexity
A supplier revenue percentage and a customer cost percentage answer different questions because they use different denominators. Understanding that direction is more important than applying a sophisticated formula to the wrong metric.
The same principle applies to structural edges. A relationship with no displayed metric can identify a path worth researching, but it should not receive an invented weight simply because the investor wants every line in a spreadsheet to contain a number.
Good reasoning beats unnecessary math.
Quant methods become useful when the question changes
There are cases where more advanced analysis is appropriate. A systematic investor may want to create point-in-time factors, backtest customer momentum, build exposure-constrained baskets, estimate partial concentration indices, or compare network features across thousands of securities.
Those are legitimate quantitative workflows, but they are extensions of the same relationship data rather than prerequisites for using it.
A fundamental investor can stop much earlier and still gain useful information.
The data should make research easier to explain
One advantage of a simple dependency analysis is that the conclusion remains legible. "Three of my largest holdings share the same customer" is easier to audit than a composite risk score built from several hidden assumptions.
That transparency matters because supply-chain data already contains uncertainty. Metrics can be missing, relationships can be structural, and public evidence can vary in quality.
A simple conclusion that preserves those limitations can be more trustworthy than a mathematically elaborate one.
An LLM can handle the mechanics without owning the judgment
An investor can also use an LLM or agent to query the network, compare relationships, calculate simple shares, and gather relevant filings. The investor still decides whether the relationship changes the thesis.
This is where structured tools are valuable. They let the model retrieve known relationships instead of guessing which companies are connected.
The user does not need to write code to ask a quantitative question.
The conclusion is that sophistication should match the decision
Supply-chain data can support advanced quantitative research, but its first layer of value is much simpler. It can tell an investor which outside companies deserve attention, where a portfolio repeats dependencies, and which relationships should influence a decision.
If the question can be answered with a comparison, do not turn it into a model. If the decision requires a backtest or portfolio optimizer, the same data can support that too.
You do not need to be a quant to use supply-chain data. You need to understand what the relationship means before deciding how much math the question deserves.
The quantitative-factor guide covers the systematic end of the spectrum. The relationship-size versus relative-exposure guide explains why even simple comparisons need the correct denominator.
For relationship definitions and evidence limits, read the Altsets methodology.