Why Trade Finance Blockchains Keep Failing

1. An Unopened Market: On-Chain Trade Finance

Tokenization has spread quickly, led by Treasuries, money market funds, and private credit. On-chain RWA value grew from about $2.9B in 2022 to over $30B in 2026, expanding more than tenfold in a little over three years. US Treasuries make up the largest share at around 40%, followed by commodities in the low teens, private credit at about 20%, and equities at about 5%.

Trade receivables, by contrast, sit at close to zero. That absence is hard to chalk up to weak market appeal, because the underlying market is far too large for that.

• Global trade in goods and services moves about $33T a year.
• Trade finance underpins as much as 90% of cross-border commerce.
• The receivables finance market alone runs $3 to $3.5T a year, and US corporations are estimated to hold about $6T in receivables on their balance sheets.

Trade receivables also carry a distinct profile as an asset class. They are short-dated instruments with defined maturities and payment amounts, and because they arise from actual commercial transactions, they turn over quickly and carry clear payment obligations that make cash flows easy to forecast. Above all, they are a core working capital asset, bridging a company's sales and its cash inflows.

In short, one of the largest real-world financial assets, and a market that shapes how efficiently corporate working capital moves, has so far stayed outside the shift on-chain.

2. Why Earlier Trade Finance Blockchains Collapsed

Applying blockchain to trade finance is not new. From 2017 to 2022, major banks and logistics firms poured substantial capital into the problem, and there were some results along the way.

• we.trade: A Hyperledger Fabric-based platform backed by 12 banks. It automated open-account trade finance for SMEs, executing payment automatically once conditions such as delivery confirmation were met. It began commercial transactions in 2019 and expanded to 15 member banks across 16 countries, but was wound down in 2022.
• Marco Polo: A Corda-based platform with 30 to 45 participating banks, offering receivables discounting and payment commitments. It entered live service in 2020 and processed real discounting transactions, but filed for insolvency in 2023.
• Contour: Backed by ANZ, BNP Paribas, HSBC, Standard Chartered, and others, it focused on digitizing letters of credit. It launched in 2020 but shut down in 2023 with monthly throughput stuck at 60 to 70 transactions.
• TradeLens: A Hyperledger Fabric platform through which IBM and Maersk digitized container trade data, discontinued in 2022.

What deserves attention in these efforts is less the failure itself than its causes, which narrow down to three.

1. Closed networks: Each consortium built its own private chain with proprietary rules and contracts. The chains could not communicate with one another, so a company had to join a separate network for each of its trading partners. Compliance and rules were implemented per service rather than at the chain level, so integration costs grew as participants were added.
2. Custom development costs: Earlier platforms rebuilt permissioning, transfer restrictions, and custody integrations from scratch each time. Development and maintenance costs kept accruing even while transaction volume stayed below critical mass.
3. A structure premised on multi-party participation: Earlier attempts were consortium models that only became useful once many banks and companies joined at once. But differing regulatory standards, internal systems, and interests across participants slowed decision-making and integration. Consensus among many institutions and operating costs were required before enough volume had accumulated, so the cost burden mounted before network effects could form.

In the end, the failure of these efforts involved both technical choices and business structure. The infrastructure was fragmented, and transactions never accumulated enough to form network effects. What was needed was not another closed platform, but standardized financial primitives that can be reused on shared infrastructure.

3. Injective Opens a PoC with POSCO International and LG CNS

Into that gap in on-chain trade finance, Injective has recently stepped forward with a notable initiative. POSCO International and LG CNS have adopted Injective as their base network for a proof of concept that handles the full lifecycle of receivables generated through international trade on-chain.

• POSCO International is South Korea's largest trading company, spanning steel, energy, food, and materials. It supplies the real trade data and transaction flows for the PoC.
• LG CNS designs the trade finance workflow. It has built digital currency infrastructure for more than 220 local governments across South Korea, and has taken part in the Bank of Korea's CBDC pilot and in security token infrastructure for Mirae Asset Securities.
• Injective provides the chain that handles post-issuance administration and settlement of the receivables at the protocol level.

The PoC brings a shared ledger, permissioned real-world assets, and agentic AI together into a single trade finance workflow. POSCO International issues receivables on Injective, and participating institutions verify each receivable's owner, transfer history, and settlement status in real time on the shared ledger.

In conventional trade finance, the seller, buyer, banks, logistics providers, insurers, and overseas subsidiaries each record the same transaction on their own books. That delays document checks and ledger reconciliation before settlement, and while it plays out, the seller's working capital stays locked between shipment and collection. The shared ledger reduces duplicate records and reconciliation, so every participant sees the same state throughout the asset's lifecycle.

LG CNS's agentic AI reviews letters of credit and trade documents to flag omissions and mismatched conditions in advance. By automating document review that has been largely manual, it moves up receivables processing and payment collection.

Compliance is embedded in the asset itself. Approved-holder requirements, jurisdictional restrictions, transfer controls, asset freezes, and issuance and redemption authority are configured through Injective's Permissions framework. When an asset is transferred, those rules apply along with it, and institutions issue and administer assets through Injective Mint, the issuance platform that unifies these functions.

4. Why Injective, and How It Avoids the Earlier Failures

Injective's PoC appears designed around the points where earlier trade finance blockchains stalled. Its responses to each of the three causes above run as follows.

First, fragmented closed networks → permissioned assets on a public chain.

The PoC issues permissioned assets on a public chain. Regulatory requirements are set directly on the asset through Injective's Permissions framework. Because control follows the asset and the underlying infrastructure is shared, assets with different rulesets can run on the same infrastructure without each participant standing up a separate closed network.

Second, repeated custom development → a standardized single issuance workflow.

Here, permissioning, transfer controls, and asset administration are handled within Injective Mint's single issuance workflow. There is no need to build new custom smart contracts and operational workflows with each issuance, which cuts the construction costs and development time that recurred every time a new asset was brought on-chain.

Third, a structure premised on multi-party participation → starting from one company's real transactions.

Starting from POSCO International's actual trade flows offers a way around this. Without first assembling many banks and companies, the PoC can secure initial volume and use cases from within a single corporate group's overseas subsidiaries and counterparties. It sidesteps the cold-start problem of the consortium model.

These three answers are made possible by one thing: compliance, issuance, and administration are already built into the protocol. On most general-purpose blockchains, these functions have to be implemented as separate smart contracts for each asset.

On Injective, by contrast, the holding eligibility, jurisdictional restrictions, and issuance and redemption authority that regulated assets require are provided at the chain level, through modules like the Permissions framework and Injective Mint. Sub-second block production and instant finality then close out actual payment settlement without delay. Carrying both axes, regulation and settlement, on a single chain is what sets it apart from general-purpose chains.

5. The Starting Point and What Remains

There is a reason POSCO International makes a persuasive starting point. It is South Korea's largest trading company, with 2025 revenue of about 32.4T won (about $22.7B), and it has spent half a century building a global trade network dating back to its Daewoo predecessor.

Because a single company's existing trade flows already generate meaningful volume, the scale that earlier consortia struggled to reach by gathering outside participants can plausibly come from one large trading house.

This is still an early stage, of course. Actually securitizing receivables, distributing them to outside investors, and connecting that to real financing remains work for the future.

Even so, across an on-chain finance arc that has run from Treasuries to private credit to repo, trade finance is still an empty market, and the room for the market as a whole to expand is just as large. That is the reason to watch what this proof of concept delivers.