CrossLedger Whitepaper

A settlement layer for physical commodity trade

A cargo of diesel, sugar or copper concentrate is paid for against paper. A bill of lading, an inspection certificate, an invoice and a dozen supporting documents travel by courier between banks in different countries while the goods travel by sea. Payment waits for the paper. When the paper is late, wrong or forged, everyone waits longer, and sometimes the money goes to the wrong party.

CrossLedger is infrastructure to change the order of operations. Trade documents are fingerprinted and registered on a public ledger the moment they are issued. Payment is held in a programmable escrow that releases when the registered documents and a signed inspection result satisfy the conditions both parties agreed in advance. The buyer knows the money moves only on proof; the seller knows the money is already there.

None of this needs a new cryptocurrency for settlement. Trades settle in dollar stablecoins. The CLXT token has a narrower job: it pays platform fees and, in a later phase, bonds the inspectors and verifiers whose signatures release funds. This paper describes the problem, the design, the token, the current state of the deployed contracts, and the risks.

How commodity trade settles today

Letters of credit and documentary collections

Cross-border commodity sales are usually paid in one of three ways. On open account the seller ships and trusts the buyer to pay. With a documentary collection, banks exchange documents for payment but give no guarantee. With a documentary letter of credit (LC), the buyer's bank undertakes to pay the seller if, and only if, the seller presents documents that comply with the credit's terms, under the ICC's Uniform Customs and Practice for Documentary Credits (UCP 600).^[1]

The LC is the workhorse of bulk commodity trade because it replaces counterparty risk with bank risk. It is also slow and expensive. Banks examine documents, not goods; a discrepancy in a date or a port name can block payment even when the cargo is perfect. Settlement commonly takes 5 to 14 business days after presentation, and LC issuance, confirmation and negotiation fees for emerging-market trades often total 1 to 3 percent of transaction value.

For small and mid-sized exporters the problem is access, not just cost. The Asian Development Bank estimates the global trade finance gap, meaning requests for trade finance that banks decline, at US$2.5 trillion, with small firms bearing a disproportionate share of rejections.^[2]

The document problem

A single international shipment can involve dozens of parties and documents. The ICC Digital Standards Initiative estimates that one transaction can require up to 36 original documents and 240 copies.^[3] Most are still paper or unstructured PDF, issued by different organisations on different systems. Each party reconciles them by hand against its own records.

The bill of lading matters most, because for most of trade history possession of the paper original has meant control of the goods. That is why it travels by courier, why it can be lost, and why the industry has relied on letters of indemnity to release cargo without it, shifting risk rather than removing it.

Fraud and duplicate financing

Because no shared record says which documents exist and who has relied on them, the same cargo can be financed more than once. In the 2014 Qingdao port case, metal stocks were pledged to multiple lenders using duplicate warehouse receipts. In 2020 the Singapore oil trader Hin Leong collapsed owing banks about US$3.5 billion, and criminal charges of forgery and cheating followed.^[4] In both cases each lender held documents that looked valid in isolation. A common registry that every financier checks before advancing funds would have shown the duplication.

GDN Group meets the everyday version of this problem on its own trade desk: counterparties offering cargoes they do not control, supported by documents that do not survive verification. CrossLedger began as the tool the desk needed.

Earlier attempts: consortium ledgers

The idea is not new, and the history is instructive. Between 2017 and 2023 several bank- and carrier-led consortia built permissioned blockchain networks for trade: we.trade, Marco Polo, Contour and the Maersk and IBM TradeLens platform. All four have since closed or become insolvent.^[5]

Their technology broadly worked. What failed was the network: each needed every bank, carrier and trader to join one private system, on terms set by competitors, before it delivered value to anyone. CrossLedger draws three lessons from this. Use a public, neutral chain that nobody has to be invited onto. Deliver value to a single trade between two parties, before any network effect. And start from a trading desk with real cargoes rather than a consortium with a mandate.

CrossLedger design

Design principles

• Fingerprints, not documents. Only cryptographic hashes and minimal metadata are written on-chain. Commercial documents stay private.
• Stablecoin settlement. Trades settle in USDT or USDC. Neither party takes price exposure to CLXT.
• Conditions agreed up front. Release rules are fixed when the escrow is created and cannot be changed by one party afterwards.
• Named, accountable signers. Releases depend on identified inspectors and parties whose signing keys are bound to verified identities, not on anonymous oracles.
• Exit to the real world. Every trade keeps a governing-law contract and a dispute path that works without the platform.

Participants

Participants are onboarded through know-your-customer and know-your-business checks run by a regulated third-party provider. The contracts record only that a wallet has passed verification, never who it belongs to.

The document registry

When a document is issued, its SHA-256 hash is written to the registry with a document type (contract, invoice, bill of lading, certificate of origin, inspection report and so on), the trade it belongs to, the submitting wallet and a timestamp. Anyone holding the original can later prove it is the registered version by hashing it again. Any change to a single character produces a different hash.

Two properties follow. First, uniqueness: the registry refuses a hash that is already registered, so the same file cannot be presented twice without the conflict showing. A hash only matches an identical file, though: a rescanned or edited copy of the same bill of lading produces a different hash. Stronger protection therefore keys each entry to issuer-authenticated identifiers, such as the bill of lading number and the carrier that issued it, and records each financing against them. That identifier layer is part of the planned design and is not yet built. Second, provenance: each entry shows which identified party submitted it and when, and verifiers can mark entries verified, rejected or revoked.

Conditional escrow

Each trade has its own escrow. The seller creates it with the amount, the stablecoin, the required documents, the inspection requirement, a delivery deadline and the arbitrator. The buyer funds it. From that point neither party can withdraw funds alone.

Release happens when every required document is registered and verified and the inspection attestation reports results inside the agreed tolerances. If the deadline passes without the conditions being met, the buyer can reclaim the funds. Either party can raise a dispute before release, which freezes the escrow until the named arbitrator rules.

The platform fee, targeted at 0.2 to 0.4 percent of trade value, is charged when the escrow is funded, with a discount when paid in CLXT. These are design targets, not rates CrossLedger has achieved in operation.

Inspection attestations

The weakest point of any on-chain settlement system is the bridge to physical reality: something has to tell the contract that the goods exist and match the contract. CrossLedger does not pretend this can be made trustless. Instead it makes it accountable. The inspector already trusted by both parties signs a structured attestation (quantity, quality parameters, location, time, certificate reference) with a key registered to their verified identity. The escrow checks the signature and compares the reported values against the trade's tolerances.

In a later phase, inspectors and verifiers will post a CLXT bond that can be forfeited if an attestation is shown to be false. That gives the token a security role tied to real activity, and gives the parties recourse beyond reputation.

A trade, end to end

Consider a 30,000 metric tonne in-tank sale of EN590 diesel at a storage terminal, paid on a dip test with title passing at injection into the buyer's tank, a structure GDN's desk uses in practice.

1. The seller creates the trade: quantity and tolerance, price, USDT as settlement asset, required documents (sale contract, tank receipt, commercial invoice, inspection certificate), a 10-day deadline, and a named arbitrator. The contract's hash is registered.
2. The buyer reviews the terms and funds the escrow. The seller can now see the money is committed.
3. The seller registers the tank receipt and invoice. The registry confirms neither has been registered before.
4. The independent inspector dips the tank, tests the product against the EN590 specification and signs an attestation of volume and quality.
5. The escrow verifies the inspector's signature and checks the results against the contract. All conditions are met, so funds release to the seller in the same transaction, and the terminal is instructed to transfer title on injection.

The same sequence under a letter of credit involves an issuing bank, an advising bank, often a confirming bank, a courier and a document examination that can take days after the product has already moved.

Disputes

A dispute freezes the escrow. The arbitrator named in the trade terms, initially an independent trade arbitration professional and later a panel of bonded verifiers, reviews the registered documents and evidence and directs a full release, a full refund or a split. Arbitration is backed by the trade's governing-law contract, so the on-chain ruling and the legal position point the same way.

Privacy and data

Prices, counterparties and volumes are commercially sensitive. CrossLedger writes no documents or party names to the public chain. Originals are stored encrypted off-chain and shared only with the parties to the trade, and registry entries are hashes, document types, trade identifiers and wallet addresses. Personal data from identity checks stays with the regulated verification provider.

Settlement itself is not private. Escrow funding and release amounts, and the wallet addresses involved, are public on Ethereum and can be linked by anyone who learns which wallet belongs to which party. Participants who need payment confidentiality should use dedicated settlement wallets, and a confidential settlement option is a later-phase research item, not a current feature.

Legal effect of electronic records

A registry entry is evidence; it is not, on its own, a negotiable document of title. The law is moving to recognise electronic equivalents. UNCITRAL's Model Law on Electronic Transferable Records (2017) sets out when an electronic record can function like a paper bill of lading, and the United Kingdom's Electronic Trade Documents Act 2023 and Singapore's amended Electronic Transactions Act give that effect in law.^[6] CrossLedger is designed to work alongside paper and electronic bills of lading issued on recognised platforms, registering their fingerprints rather than replacing them, until the legal framework in each corridor supports more.

The CLXT token

On-chain facts

The following figures were read from Ethereum mainnet on 13 October 2026.

Intended utility

CLXT is designed for three uses on the platform. None is live yet.

• Fees. Platform fees can be paid in CLXT at a discount to paying in stablecoins.
• Verifier bonds. Inspectors, verifiers and arbitrators post CLXT as a bond that can be forfeited for false attestations or rulings.
• Governance (later phase). Holders may vote on fee schedules and corridor priorities once the platform is operating.

CLXT is not a settlement currency for trades, does not represent a share of GDN Group or of platform revenue, and carries no right to dividends, interest or redemption.

Allocation

60% of supply is locked on-chain. The ecosystem, team and liquidity allocations, 600,000,000 CLXT in total, sit in three Sablier vesting streams that cannot be cancelled by GDN or anyone else. The schedules were set on 6 October 2026, and the streams were re-created on the new token on 13 October 2026 with the same amounts and dates. They pay back to the owner address only on the schedules above. The treasury, compliance, operations and strategic allocations are held, unlocked, in the owner address. Holdings and locks are summarised on the Token facts page.

Token migration and supply

The original CLXT contract, 0xDa23…0A6d, included a staking function that paid a 10 percent annual reward by adding newly created tokens to a holder's balance, while its totalSupply() returned a hard-coded 1,000,000,000 and no transfer event was emitted. Had staking been used, the real number of tokens could have risen above the reported supply. The contract also carried a trading switch controlled by its owner. Wallet security scanners flagged both. Staking was never used: the contract emitted no staking event in its lifetime.

That contract could not be upgraded, so on 13 October 2026 CLXT was reissued as a new contract, 0x64B3…c5CC: the unmodified OpenZeppelin ERC-20, with all 1,000,000,000 tokens created once at deployment, no owner, no mint function, no staking, no pause and no fees. Its total supply is the actual supply.

How balances were handled. The three vesting locks were re-created on the new token with the same amounts and release dates (Sablier #1797, #1798 and #1799, replacing #1783, #1784 and #1785, which remain locked on the retired token). The presale moved to the same verified contract code on the new token and was funded with the same 20,000,000 CLXT. No other balances were carried over: wallets outside the owner address that held the original token, including the strategic allocation wallet, received no new CLXT, and the equivalent amount is part of the owner address's unlocked balance. No purchase was ever completed through a presale contract for the original token.

Every step was rehearsed on a copy of Ethereum mainnet before it was signed, and the resulting state was checked on-chain afterwards. That was an internal review, not an independent audit.

Presale

The public presale sells CLXT for USDT on Ethereum through a contract that delivers tokens to the buyer in the same transaction. The published plan has four stages:

Be clear about what the contract enforces. It sells at a single rate set by the owner; it does not enforce stage allocations or move between stages automatically. The owner can change the rate, minimum purchase and treasury, pause the sale and withdraw unsold tokens. The website applies a 200 USDT minimum, and the V2 contract also enforces a minimum on-chain. Stage pricing is a commitment by GDN, not a property of the code. The four stages total 17,000,000 CLXT, while V2 was funded with 20,000,000 CLXT. The additional 3,000,000 CLXT has no allocation in the published stage plan. The strategic allocation (125,001,000 CLXT, now in the owner address) and the V2 funding total 145,001,000 CLXT, compared with the planned 150,000,000 CLXT strategic allocation; the allocation of the remaining 4,999,000 CLXT is not specified here.

Each stage price is a sale price, not a valuation. CrossLedger publishes no listing price, target price or expected return, and nothing guarantees that CLXT will be listed on any exchange or trade at or above the price paid.

The V1 presale defect

The first presale contract, 0xABCA…2B35, cannot complete a purchase. It declares USDT's transferFrom as returning a boolean:

interface IUSDT {
    function transferFrom(address from, address to, uint256 amount)
        external returns (bool);   // mainnet USDT returns nothing
}

Mainnet USDT predates the final ERC-20 standard and returns no value from transfer or transferFrom. Since Solidity 0.8, a call declared to return a value reverts if the callee returns nothing. Every buyWithUSDT call therefore reverts, even for a buyer who holds enough USDT and has approved the contract. No buyer funds can be lost to this defect; the transaction simply fails and only gas is spent.

GDN reproduced the failure against live mainnet state on 3 October 2026, confirming the contract was active, funded and given full approval, and that the revert carries none of the contract's own error messages. The corrected contract, CLXPresaleV2, keeps the same purchase flow but makes token calls through a wrapper that accepts both standard tokens and USDT's empty return. It also adds an on-chain minimum and a buyer-set minimum output that protects a pending purchase against a rate change. In a mainnet-fork test it completed a 200 USDT purchase, paying the treasury and delivering 2,000 CLXT to the buyer, and correctly rolled back every failure case tested. V2 was deployed on 5 October 2026 at 0x04cC…6Fc3, its source verified on Etherscan as an exact match, and it was funded with 20,000,000 CLXT. The website checkout now uses V2.

Business model

CrossLedger earns revenue from use, not from token sales.

• Settlement fees on escrowed trade value, targeted at 0.2 to 0.4 percent.
• Document registration fees for financiers and parties registering documents outside a CrossLedger escrow.
• Registry checks for banks and trade financiers who query the registry before funding a cargo.
• Enterprise integration for traders and terminals connecting their own systems.

GDN Group's trade desk, active in refined petroleum products, sugar and agricultural commodities across Asia-Pacific, the Middle East and the Americas, provides the first trades. The platform has to earn its place on that desk before it asks anyone else to use it.

Status and roadmap

Phases are sequenced by readiness, not by date. Each starts only when the previous phase's deliverables exist and can be verified.

Team

Guilherme (Gui) Di Nardo, Founder and CEO, Brisbane. Leads GDN Group's trading and advisory work across petroleum products, agricultural commodities and sovereign advisory engagements.

Tom Young, Co-founder and CTO. Leads smart-contract development and platform engineering across the token, presale and verification layers.

Fernando Nicola, United States commercial representative, Orlando. Commodity trading background with international energy trading houses, focused on North and Latin American markets.

Mathew Dunn, UAE and Gulf representative, Dubai. Leads Gulf relationships and blockchain integration for CrossLedger.

Marcos "Kito" Vianna, Brazil and South America representative, São Paulo. Over twenty years in trading, freight forwarding and port logistics.

Regulatory position and restrictions

CrossLedger is a product of GDN Enterprise Pty Ltd (ACN 666 495 263), an Australian proprietary company registered with ASIC. GDN does not hold an Australian Financial Services Licence and is not an authorised representative of a licence holder.

Australia is reforming the regulation of digital assets and digital asset platforms. GDN obtained Australian legal advice on CLXT and the presale and, on that advice, opened the presale to Australian residents on 6 October 2026. GDN still does not hold an Australian Financial Services Licence, and the classification of digital assets in Australia may change as the reforms proceed.

CLXT is not offered to residents of the United States, Canada, the People's Republic of China, North Korea, Iran, Syria, Cuba, or any jurisdiction subject to comprehensive sanctions or where the offer would require a licence or registration GDN does not hold. The purchase interface is disabled for visitors whose IP address resolves to a restricted jurisdiction. That screen is not proof of residence, and each participant remains responsible for confirming eligibility under their own law. At present no identity verification is applied beyond that IP-based screen. GDN intends to introduce identity verification through a regulated provider and will publish the provider and thresholds before it applies.

Risk factors

• Total loss. CLXT may become worthless. Commit only what you can afford to lose entirely.
• Development risk. The platform described here may never be completed, or may work differently.
• Smart-contract risk. Two defects were found in earlier contracts, and both contracts have been replaced. Others may exist. No independent audit has been completed.
• Wrong-contract risk. The retired original CLXT contract still exists on-chain and can still be transferred. Only 0x64B3…c5CC is CLXT.
• Concentration and control. 60% of supply is locked in vesting, but the owner address holds 38% unlocked, and locked tokens return to the owner address over time. The presale owner can change presale terms.
• Regulatory risk. Laws may change or be applied in ways that restrict CLXT, the presale or the platform, including in Australia.
• Liquidity risk. There may be no market in which to sell CLXT. No exchange listing is guaranteed.
• Adoption risk. Traders, inspectors and financiers may not use the platform.
• Key and wallet risk. Lost keys and mistaken transfers cannot be reversed.

Conclusion

Physical commodity trade does not need a new kind of money. It needs a shared, neutral record of which documents exist and who relied on them, and a way to commit payment that releases on proof rather than on paper arriving. Those are narrow, achievable goals, and public blockchains are now mature enough to meet them without asking an industry to join a consortium first.

CrossLedger is early. This paper tries to be precise about that: what is deployed, what is broken and being fixed, and what is still a plan. The test of the project is simple and public. Watch the registry and the escrow contracts. When real cargoes settle through them, the design is working.

Notes and references

1. International Chamber of Commerce, Uniform Customs and Practice for Documentary Credits, ICC Publication No. 600 (2007). ↩
2. Asian Development Bank, 2023 Trade Finance Gaps, Growth, and Jobs Survey (September 2023). ↩
3. ICC Digital Standards Initiative, estimate of documents per cross-border transaction. ↩
4. Public reporting on the 2014 Qingdao port metals financing fraud and the 2020 insolvency of Hin Leong Trading (Pte) Ltd. ↩
5. we.trade ceased operations in 2022; TradeLens was discontinued in early 2023; Marco Polo entered insolvency and Contour ceased operations in 2023. ↩
6. UNCITRAL Model Law on Electronic Transferable Records (2017); Electronic Trade Documents Act 2023 (UK); Electronic Transactions Act 2010 (Singapore), as amended in 2021. ↩

This whitepaper is general information about a project in development. It is not financial, legal or tax advice, not a prospectus or product disclosure statement, and not an offer of securities or financial products in any jurisdiction. © 2026 GDN Enterprise Pty Ltd. Questions: info@crossledger.trade.