A trade takes microseconds. Settling it takes days.

In that gap the buyer holds an obligation, the seller holds a receivable, and both are exposed to the other failing. Markets do not fix the gap — they insure it.

• A clearing house steps in as counterparty to both sides, and must be capitalised against that risk.
• Margin is posted against settlement failure — capital that cannot be deployed anywhere else.
• Every participant keeps its own record of the same event, so reconciliation is a permanent cost.
• Corporate actions are applied to positions that are still in flight, which is why entitlements are disputed.

Everyone pays for the interval.

The gap is not an inconvenience, it is a line item — in capital that cannot be deployed, in headcount that exists to make records agree, and in the intermediaries a market must capitalise.

• Margin. Capital posted against settlement failure earns nothing and finances nothing.
• The clearing house. An institution that must be capitalised because it stands between both sides.
• Reconciliation. Four records of one event, and a function whose whole job is making them agree.
• Fails and entitlements. Trades that do not settle, and distributions paid to a holder who has since changed.

None of that machinery serves the trade.

Margin, novation and reconciliation all exist to manage the interval after the trade. Close the interval and they have nothing left to manage — which is a far larger prize than making any one of them cheaper.

Both legs, or neither.

A settlement instruction carries the asset leg and the cash leg. They apply inside one transaction against one ledger: the security moves to the buyer, the cash moves to the seller, and the transfer is final at block inclusion — or nothing moves at all.

Cube Mainnet · EVM · chain ID 7526 · zero base fee

Inside one block.

A settlement instruction is a single transaction carrying both legs. It applies against one ledger, and the market is either moved completely or not at all.

• The venue matches a trade and submits one instruction naming the asset leg and the cash leg.
• Both legs apply against the same state, under authorisations each side has granted.
• If either leg cannot apply — balance, authorisation, a halted instrument — the whole transaction reverts.
• On inclusion it is final. There is no probabilistic state for a downstream system to interpret.

Specified in sections 7 and 9 of the whitepaper.

The hard part is the first venue.

Infrastructure like this is not won on code. It is won when a supervised venue is willing to run on it, and that is the work in front of us: ContiSX, a Nigerian exchange, is the venue we are building the integration with. It is not settling on CubeCore today.

• The engineering is the smaller half; the regulatory conversation is the moat.
• A venue that integrates brings its whole book, not a single trade.

We never ask a venue to stop being a venue.

The layer takes the mechanical movement of title and cash. Everything requiring a licence and a judgement stays exactly where a regulator already expects to find it.

One engine, five markets.

The layer is indifferent to the instrument. What changes between a warehouse receipt and a listed share is the rules its issuer attaches, not the settlement path — so each new market is an integration, not a rebuild.

• LOYEX — Digital Value Units, a new asset category native to the layer.
• Public equities — listed instruments, fractional and cross-border.
• Commodities — metals, agriculture and energy, provenance recorded with title.
• Private market — placements, pre-IPO and venture transfers.
• Secondary market — peer-to-peer transfer by atomic swap.

The same path, at any ticket size.

A layer that settles only equities is a utility. One that settles a share, a warehouse receipt and an in-app purchase on the same path is infrastructure — and the ecosystem around it is where that gets proven.

• Identity. Every account carries a Cubeline — a private number derived from its own keys, addressing encrypted messaging and calls, so counterparties reach each other without exchanging personal details.
• The device. A mobile operating system whose account is the chain account: identity, inbox, wallet and Cubeline as one credential, keys held by the holder.
• The DappStore. An application store whose listings and distribution are on-chain, and whose purchases settle through the same atomic path as a trade.
• Why it matters commercially. Small-ticket settlement proves the path is general, and every product on it deepens the same ledger.

The pieces only recently existed together.

• Cash legs are on-chain. Without xNGN and Cube USD the payment side would have to be instructed elsewhere, reintroducing the gap.
• Instruments are issuable natively, with their transfer rules expressed as code rather than paperwork.
• EVM tooling is standard, so auditors, integrators and supervisors read the settlement logic without a new toolchain.
• A licensed venue is running on it, which converts "regulators might allow this" into a reference.

Why not just fix what exists?

• Shorten the cycle. T+2 to T+1 halves the interval and keeps every institution built to manage it. The costs are structural, not schedule-dependent.
• A private consortium chain. Solves the ledger problem for its members and rebuilds reconciliation at its edges, with governance to match.
• Stablecoin rails alone. Move the cash leg on-chain and leave the asset leg in a register elsewhere — which is the gap, restated.

Atomicity is the property that removes the machinery, and it needs both legs on one ledger. Every other choice here follows from that one.

The layer is paid per settlement.

Every settled trade consumes gas, and every venue that integrates brings its whole book rather than a single trade. Value compounds along three lines:

• Settlement throughput — the volume the layer clears for venues that have already integrated.
• Instrument issuance — assets created on the layer stay on it for their whole lifecycle.
• Venue integrations — each licensed operator added is a durable, high-switching-cost relationship.

Fee schedule and unit economics: see the accompanying financial model.

Each venue makes the next one easier.

The honest list.

This is a regulated-market bet, not a technology bet. The engineering risk is the smaller half.

• A supervisor has to be comfortable with settlement finality on a public chain. That conversation is the critical path.
• The cash leg must exist on-chain in the currency the market trades, or atomicity stops at that boundary.
• Issuers have to represent instruments natively rather than mirror a legacy register.
• The first venue has to go live and stay live; until then this is architecture, however good.

Built, not yet live.

The settlement layer is built and testable. No market settles production traffic on it yet, and this deck carries no volume figures because there are none to carry.

• Chain, settlement path and integration surface: built.
• First venue integration: in progress with ContiSX.
• Live production settlement: not yet.
• Audited operating figures: when there are any, with their method.

More venues, more instrument classes.

• Additional licensed venues, following the pattern ContiSX established.
• Deeper instrument coverage — issuance, corporate actions and redemption handled end to end on the layer.
• Cross-border settlement between venues sharing the same ledger and cash legs.
• Integration surface: SDK, contract templates, webhooks and a testnet mirroring production.

Raise size, use of funds and milestones are covered in the accompanying materials.