A single crude-oil shipment can need 30+ original documents passed between 20+ parties — banks, exporters, shippers, insurers, customs. The paperwork locks up working capital for over a week.
This is the case where blockchain's shared ledger property does the heavy lifting — and where the first wave of projects taught a hard lesson.
In a letter of credit (LC), a bank promises to pay the exporter once documents prove the goods were shipped. The catch: the bill of lading, inspection report, and insurance certificate are physical or siloed digital documents couriered between parties who each keep separate records.
Two expensive problems follow. First, time: reconciling and verifying documents by hand takes days, and the goods (and capital) sit idle. Second, fraud: because no one sees a shared record, the same cargo can be pledged to two different banks for two loans — a classic trade-finance fraud that has cost banks billions.
The enemy isn't the document. It's that every party trusts only its own copy of it.
For a $200m copper trade, shaving nine days off the cycle frees roughly $493,000 in working capital — just from the time value of money, before counting fraud losses or staff hours.
Unlike Bitcoin, these are permissioned ledgers: only vetted banks, traders, and inspectors can join. But the core trustless property still applies — every member holds the same tamper-evident copy, and no single party can quietly alter or cancel a record without the others seeing it.
This is the same hashing + shared-ledger idea from the fundamentals course, applied to documents instead of coins.
On the left, the old world: two banks, two private files, no shared view. On the right, a shared ledger. Press the button to pledge the same bill of lading to a second bank and watch what each system does.
The commodity-finance platform komgo moved the whole LC workflow — application, inspection report, KYC pack — onto one shared permissioned ledger. In early production runs, LC turnaround fell from about ten days to roughly one hour. Step through the lifecycle and compare the two timelines.
The same shared-ledger property powers supply-chain provenance. Walmart famously cut the time to trace a package of sliced mangoes from its source from nearly 7 days to 2.2 seconds. In a contamination scare, that's the difference between recalling one farm's output and dumping an entire product line.
Each handoff — farm, packer, shipper, distributor, store — writes a hashed, timestamped record. No party can backdate or quietly edit the chain of custody.
Click a stage to inspect its immutable record. Then try to tamper with one.
Be honest with your students: TradeLens (Maersk/IBM), we.trade (11 European banks), Marco Polo, and Contour all ceased operations between 2022 and 2024. Yet none failed on the cryptography. They failed on economics and governance:
The second wave learned this. Focused tools — electronic bills of lading, with legal backing from the UN's Model Law on Electronic Transferable Records (adopted by Singapore, Bahrain, Abu Dhabi and a growing list) — are gaining real, measurable traction precisely because they are narrow, neutral, and legally enforceable.
Lesson for any blockchain project: solving the trust problem technically is necessary but not sufficient. Governance and incentives decide adoption.
Trade finance is a textbook fit for a shared ledger: many distrustful parties, duplicate documents, fraud from siloed records. The technology demonstrably works — 10 days to 1 hour, fraud caught instantly. The binding constraint is human: neutral governance and a viable commercial model.
Next case → Tokenizing real-world assets
Jan Erik Meidell