TreasureX Researchecon.GNYou Don't Need the Whole Supply Chain Map to Make a Good Call
Shaowen Luo, Kwok Ping Tsang, Zichao Yang · September 24, 2026
The problem: Companies and governments rarely see the full map of who supplies whom. They see fragments — industry totals, regional shipments — and must still decide which suppliers or plants to watch.
The finding: Luo, Tsang and Yang show that a production network can stay largely unknown while the decision it supports is fully identified. Using Japan's interregional input-output tables, they picked a monitoring set from 1995 data and tested it against the 2005 benchmark. At about half the statistics needed for full disclosure, the set lost only 0.07 percentage points of average exposure versus the best possible choice. But its worst-case regret across all networks consistent with the published totals was 4.83 points — a real gap between "good on the benchmark" and "guaranteed."
In U.S. coal deliveries around a 2005 Wyoming rail disruption, extra shipment data pinpointed which plants to monitor for inventory risk, even though four of those plants' exposure to the affected coal supply stayed unidentified.
Why it matters: Data collection is expensive. This work says you can design cheaper monitoring around the decision you need to make, and that decisions can be pinned down before individual exposures are. For investors, it reframes supply-chain risk: ask what decision the data must support, not whether the map is complete.
Source: arXiv paper by Shaowen Luo, Kwok Ping Tsang and Zichao Yang — https://arxiv.org/abs/2609.29905
Educational content only, generated with AI assistance from public research abstracts. Not investment advice. Thank you to arXiv for use of its open access interoperability.