CubeCore — settlement for digital capital markets.

1Introduction

Securities settlement is the practice of moving an asset to its buyer and cash to its seller after a trade has been agreed. Almost everywhere, those two movements are separated by days. The interval is not a technical necessity; it is an artefact of batch processing between institutions that do not share a ledger, and an entire apparatus — central counterparties, margin, novation, reconciliation — exists to manage the risk that the interval creates.

CubeCore closes the interval. It is a settlement layer on which both legs of a trade execute as a single atomic operation: the asset moves, the cash moves, and the transfer is final in one block, or neither leg moves at all. This paper sets out the problems that follow from the settlement gap, the mechanisms that remove them, the chain those mechanisms run on, and the shape of the first venue integration.

Who this is for

• Market operators evaluating whether a supervised venue can settle on a public chain, and what would remain their responsibility if it did.
• Regulators and auditors asking what a continuously readable register changes about supervision.
• Issuers weighing what it means for an instrument to carry its transfer rules as code.
• Engineers integrating settlement, who will find the interface in section 13.

What this paper is not

It is a description of a design and its current state, not a track record. At the time of writing no market settles production traffic on CubeCore. Section 15 states plainly what is built, what is in progress, and what does not exist yet, and nothing earlier in the document should be read as a claim that a licensed market is already running on it.

2The settlement gap

A trade is agreed in microseconds and settles in days. In that window the buyer has an obligation rather than an asset, the seller has a receivable rather than cash, and both are exposed to the other failing to deliver. Markets manage this with intermediaries: a clearing house steps between the two sides, becomes the counterparty to each, and collects margin against the possibility of default.

The arrangement works, and it is expensive. Margin is capital that cannot be deployed. Novation concentrates risk in an institution that must then be capitalised against it. Reconciliation exists because each participant keeps its own record of the same event and the records disagree. Corporate actions — dividends, splits, redemptions — are processed against positions that are still in flight. Every one of these costs is a function of elapsed time between trade and settlement.

3The problems

The settlement gap is one fact with many consequences. Each of the following is a cost a market pays because ownership and payment are not simultaneous; each is addressed by a specific mechanism in section 5.

3.1 Settlement risk

Between execution and settlement the buyer holds an obligation rather than an asset and the seller holds a receivable rather than cash. If either fails, the other is exposed. Markets do not remove this exposure; they insure it, by inserting a central counterparty that becomes the buyer to every seller and the seller to every buyer, and capitalising it against default.

3.2 Capital locked as margin

The insurance is paid for with margin — capital posted against the possibility that a trade fails to settle. That capital earns nothing for its owner and finances no economic activity. It exists solely to cover an interval.

3.3 Reconciliation

A venue, a clearer, a custodian and a regulator each keep their own record of the same trade. Because the records are separate they disagree, and an industry of matching, breaks investigation and end-of-day files exists to make them agree again. Reconciliation is not a failure of diligence; it is the unavoidable cost of several ledgers describing one event.

3.4 Disputed entitlements

Dividends, splits, rights issues and redemptions are applied to positions that may still be in flight. Who is the holder of record when a distribution falls between execution and settlement is a question that would not exist if the two were simultaneous.

3.5 Cross-border friction

Where a trade crosses currencies or jurisdictions, the gap widens: correspondent banking, separate settlement cycles and separate registers turn a single trade into a chain of dependent transfers, each with its own failure mode.

3.6 Opacity between reports

A supervisor sees the market as it was at the last cut-off, not as it is. Between reports, exposure is inferred rather than observed, and questions that should be answerable in seconds require a request to the supervised firm.

3.7 Fractional and small-ticket exclusion

Fixed per-trade settlement costs make small orders uneconomic, which is one reason fractional ownership and micro-investment remain awkward in conventional infrastructure — the machinery costs the same whether the trade is large or small.

4Design principles

• Atomicity before speed. A fast settlement that can half-complete is worse than a slow one that cannot. Both legs commit together or the transaction reverts.
• One ledger, many participants. Buyer, seller, venue and regulator read the same record rather than reconciling four of them.
• Finality is a property, not a promise. Once a settlement transaction is included, it is not reversible by the venue, the operator or the chain.
• The venue keeps its rules. CubeCore settles; it does not decide who may trade, at what price, or under which licence. Market rules stay with the licensed operator.
• Auditability by default. A supervisor should be able to verify the state of the market without asking anyone for a file.

5The solutions

Every mechanism below follows from one decision: settle both legs of a trade in a single transaction on a ledger all participants read. The sections that follow describe each in technical detail; this is the map from problem to answer.

5.1 Simultaneity removes settlement risk

A settlement instruction carries the asset leg and the cash leg and applies both inside one transaction. There is no interval in which one party has performed and the other has not, so there is no exposure to insure, no central counterparty to novate through, and no default waterfall to capitalise. See section 7.

5.2 No settlement interval, no settlement margin

Margin posted against settlement failure is margin against a risk that no longer arises. Capital that a clearing member would have pledged stays deployable. Market risk and credit risk are untouched — this addresses the settlement component only, and section 7.2 is explicit about what remains.

5.3 One ledger removes reconciliation

Where there were four records there is one. The settlement is the record: a double-entry movement written at the moment of transfer, readable by the venue, the members and the supervisor without a file being produced for anyone. See section 9.

5.4 Exact holders make corporate actions mechanical

Because ownership is exact at every block, a distribution is a transaction to the holders of record rather than an interpretation of who ought to have been the holder. Splits rewrite units; redemptions burn the instrument against delivery. See section 12.

5.5 One chain for both legs removes cross-border chains

When the cash leg is an on-chain instrument — xNGN or Cube USD — a cross-currency trade is still one transaction rather than a sequence of correspondent transfers. See section 8.

5.6 Continuous readability replaces periodic reporting

A supervisor with read access sees positions and settlements as they occur. Reporting becomes a view of the ledger rather than an extract from it, and the gap between what is true and what has been reported closes.

5.7 Uniform cost admits small trades

Settlement is a transaction whose cost does not scale with the size of the position it moves, and the network runs a zero base fee so that cost is predictable rather than auctioned. Fractional units settle on the same path as blocks. See sections 6.1 and 8.

6Chain architecture

CubeCore settles on Cube Mainnet, an EVM chain (chain ID 7526). EVM compatibility is a deliberate constraint rather than a convenience: settlement logic is written in a language auditors, regulators and integrators already read, and existing tooling — wallets, signers, indexers — works without translation.

Gas and fee policy

The chain runs a zero base fee and submits legacy (type-0) transactions. The intent is that the cost of settling a trade is not a variable the market has to price: a venue integrating the layer should be able to quote a fee schedule to its members without exposing them to fee auctions.

Finality

Settlement is confirmed at block inclusion; the layer exposes no probabilistic settlement state to applications. A trade is unsettled, or it is settled and irreversible. There is no intermediate state for a downstream system to interpret, which is what makes the ledger safe to treat as the single record of ownership.

7Atomic settlement

Delivery versus payment is the requirement that the transfer of an asset occurs if, and only if, the corresponding payment occurs. On CubeCore it is not a procedure coordinated between systems; it is a single transaction.

The mechanism

A settlement instruction names the asset leg (instrument, quantity, seller) and the cash leg (currency, amount, buyer), together with the authorisations that permit each. Both legs are applied within one transaction against the same state. If either leg cannot be applied — insufficient balance, a failed authorisation, an instrument that is halted — the transaction reverts and the market is exactly where it was before the instruction was submitted.

• No settlement window: the asset and the cash change hands in the same block.
• No counterparty exposure between execution and settlement, because there is no interval between them.
• No fails-to-deliver to process, because a leg that cannot be delivered aborts the whole settlement.
• No margin against settlement risk, because the risk being margined does not arise.

What it does not remove

Atomic settlement addresses settlement risk. It does not address market risk, liquidity risk, or the credit standing of a participant before a trade is struck. A venue still needs pre-trade controls, and those remain the venue's responsibility.

8Assets and the cash leg

An asset settles on CubeCore when it is represented on the chain as a transferable instrument with a known issuer and a defined lifecycle. The layer is deliberately indifferent to the instrument's category — what varies between a commodity and an equity is the rules its issuer attaches, not the settlement path.

Asset legs

• Digital Value Units (DVUs) — the native instrument of the LOYEX market, a new asset category traded on the layer.
• Tokenised commodities — instruments with verified provenance, redeemable against physical delivery.
• Public equities — instruments issued by listed companies, including fractional units.
• Private instruments — placements, pre-IPO holdings and venture positions, transferred under the issuer's transfer restrictions.

Cash legs

The cash side settles in xNGN or Cube USD. Both are on-chain instruments, which is what allows the payment leg to be part of the same transaction as the asset leg. A payment that had to be instructed through an external system would reintroduce the interval this design exists to remove.

9The ledger

Every settlement writes a double-entry record: the instrument debited from one account and credited to another, the cash leg mirrored in the opposite direction, both carrying the reference of the trade that produced them. The record is the settlement, not a report about it.

This is the property that removes reconciliation. Where a conventional market has a venue's record, a custodian's record, a clearer's record and a regulator's file — all describing the same event and periodically compared — CubeCore participants read one ledger. A supervisor with read access sees positions as they are, continuously, rather than as they were at the last cut-off.

10Market structure

CubeCore is a settlement layer, not an exchange. Order entry, matching, market surveillance and membership stay with the venue; the layer takes the matched trade and settles it. This separation is what lets a licensed operator run a regulated market without asking its regulator to accept a new trading venue at the same time.

ContiSX

ContiSX, a Nigerian exchange, is the first venue the layer is being fitted to. Section 11 describes that integration in full — including its status, which is that ContiSX is not settling on CubeCore today.

Markets the layer is built for

• LOYEX — the DVU market, a new asset category native to the layer.
• Private market — placements, pre-IPO and venture transfers with full audit trails.
• Secondary market — peer-to-peer securities transfer by atomic swap.
• Commodities — metals, agriculture and energy, with provenance recorded alongside title.
• Public equities — listed instruments, including fractional and cross-border settlement.

11ContiSX: the reference integration

ContiSX is a Nigerian exchange. It is the first venue the settlement layer is being fitted to, and it is a useful reference precisely because it is an ordinary regulated market: it has members, listing standards, a matching engine, surveillance obligations and a supervisor. Nothing about the integration asks it to stop being those things.

11.1 The division of responsibility

The venue keeps everything that requires a licence and a judgement; the layer takes the mechanical movement of title and cash.

• ContiSX keeps: membership and admission, listing standards, order entry, matching and price formation, pre-trade risk controls, market surveillance, halts and its own rulebook.
• CubeCore provides: settlement of matched trades, the register of title, movement of the cash leg, the audit trail, and the execution of corporate actions against the register.
• Neither takes the other's job. The layer never decides who may trade or at what price; the venue never has to operate a settlement system.

11.2 The path of a trade

• A member submits an order to ContiSX, under the venue's own membership and risk rules. Nothing about that step changes.
• ContiSX matches it and produces a trade: instrument, quantity, price, buyer, seller.
• The venue submits one settlement instruction carrying both legs — the instrument from seller to buyer, the cash from buyer to seller — under an authorisation each member has granted for its own matched trades.
• The instruction applies atomically. Either both legs commit in the same block, or the instruction reverts and the market is exactly where it was.
• The register and the members' balances are updated by that same transaction. There is no separate settlement run and no end-of-day file.

11.3 What it removes for this venue

Read against section 3, the integration is an attempt to delete costs rather than optimise them.

• Settlement risk (3.1): no interval, so no exposure between execution and settlement and no central counterparty required for it.
• Settlement margin (3.2): capital that would have been posted against settlement failure is not posted.
• Reconciliation (3.3): the venue, its members and the supervisor read the same register; there is no matching exercise between separate records.
• Entitlements (3.4): holders are exact at every block, so distributions and splits are executed against the register itself.
• Opacity (3.6): supervision becomes a read of current state rather than a request for a periodic file.

11.4 What it does not remove

A market still needs pre-trade credit controls, surveillance and the power to halt. Members can still be wrong about price and can still fail for reasons that have nothing to do with settlement. The layer guarantees that a trade either settles completely or not at all; it does not guarantee that the trade was a good idea, and it makes no representation about an issuer honouring what an instrument represents.

11.5 The cash leg in a naira market

Atomicity requires both legs on the same ledger, so the payment side settles in an on-chain instrument — xNGN for a naira market, Cube USD where the trade is dollar-denominated. A payment instructed through an external system would reintroduce exactly the interval this design exists to remove, so the boundary of the atomic guarantee is the boundary of the on-chain cash leg.

11.6 Supervision

A supervisor granted read access sees the register and the settlement history continuously: positions by member, every transfer with the trade that produced it, and every corporate action applied. The intention is that a question about exposure is answered by looking, not by asking the supervised firm to produce a report about itself.

11.7 Where the work stands

• Built: the chain, the atomic settlement path, the instrument and cash-leg model, and the integration surface described in section 13.
• In progress: the ContiSX integration itself — instrument onboarding, member account model, the venue's authorisation flow, and the operational rehearsal that has to precede any regulated go-live.
• Not done: production settlement of a live order book. Until that is true, this section describes intent and architecture.

12Identity and the Cubeline

An account on the chain is more than a balance. Alongside the instruments and cash it holds, every account carries a Cubeline — a private number derived from the account's own keys rather than issued by a carrier — which addresses private messaging and calls between the people and institutions the ledger already knows.

12.1 Why a settlement layer issues an identity at all

Settlement is between counterparties, and counterparties need to reach each other: a member and its venue, an issuer and its holders, a buyer and a seller resolving an instruction. Conventionally that means exchanging phone numbers and email addresses — a second identity system, maintained separately from the one that actually holds the assets, and one that leaks personal contact details to everyone you transact with.

The Cubeline collapses the two. The same account that holds an instrument is reachable as an identity, so a counterparty can be contacted without either party disclosing a personal number, and without a directory that has to be kept in step with the register.

12.2 What it is

• A number derived from the account, not allocated by a telecommunications operator — there is no SIM, no porting and no carrier able to reassign it.
• Portable by construction: it belongs to the keys, so it moves with the account across devices rather than with a handset.
• An address for private messaging and voice and video calls, carried by the ecosystem clients rather than by the settlement path itself.
• Private by default: conversations are end-to-end encrypted between devices, and the chain holds the identity rather than the content.

12.3 What it means in a market

• A venue can reach a member, and an issuer its holders of record, addressed by the same identity that appears in the register.
• Counterparties can communicate about a settlement without exchanging personal contact details.
• Because the identity is derived rather than assigned, there is no separate directory to reconcile against the ledger — the failure mode described in section 3.3, avoided in a second place.

12.4 Boundaries

The Cubeline is an identity primitive, not part of the settlement guarantee. Messages and calls are not settlement instructions, nothing said over one moves an asset, and a compromised device compromises the conversation in the ordinary way. Key loss carries the same consequence here as it does for custody: the identity is the keys, so losing them loses the number with the account.

13The device layer and the DappStore

Settlement is authorised by keys, and keys live somewhere. The ecosystem around CubeCore takes that seriously enough to reach the device: a decentralised mobile operating system in which the account is the chain account, and an application store whose catalogue and distribution are on-chain rather than administered by a platform owner.

13.1 The account is the device

On a conventional phone the operating system account and the wallet are separate things, and the wallet is an application that borrows the device's trust. Here they are one: the account that signs a settlement instruction is the account that signs into the device, holds the inbox, and carries the Cubeline described in section 12. Keys are held by the device rather than by a custodian, so authorisation happens where the person is.

• One account: identity, inbox, wallet and Cubeline are the same credential rather than four.
• Self-custodial by construction — the device holds the keys, and no operator can move an instrument on the holder's behalf.
• Portable: because identity derives from keys, moving to another device is a key operation, not a support request.

13.2 An on-chain application store

The DappStore is the ecosystem's application store, and it is a market like any other: publishers list, users install, and value moves. What distinguishes it is that the listing and the distribution are recorded on the chain rather than held in a platform's private database, and that anything paid — a purchase, a subscription, a reward — settles through the same layer as a securities trade, on the same terms described in section 7.

• Publishing is permissionless in the sense that matters: the record of what was published, by whom and when is on-chain and not deletable by a gatekeeper.
• Payments and rewards settle atomically like any other transaction — the store is not a special case in the settlement path.
• Installs are addressed to the account rather than to a handset, so an application follows the person across devices.

13.3 Why this belongs in a settlement paper

Two reasons. The first is custody: a settlement layer whose users hold their own keys needs somewhere safe for those keys to live, and a device that treats the chain account as its primary account is a better answer than an application sitting on an operating system that does not know what a key is. The second is that a store is a market — the same atomic movement of an asset against a payment, at a much smaller ticket size — which is a useful proof that the settlement path is general rather than tuned to securities.

13.4 Status

The device and the store are ecosystem products with their own maturity, described here to show where the settlement layer sits rather than to claim they are finished. As section 17 states, no market — including this one — settles production traffic on CubeCore today.

14Lifecycle and corporate actions

An instrument's life does not end at issuance. Dividends are paid, shares split, commodities are redeemed and instruments mature. In a market with a settlement interval these events must be applied to positions that are partly in flight, which is why entitlement disputes exist.

On CubeCore, holders are known exactly at every block, so a corporate action is executed against the register itself: a distribution is a transaction to the holders of record, a split rewrites units, a redemption burns the instrument against delivery. Each is recorded on the same ledger as the trades that produced the positions.

15Integration

A venue or application integrates the layer through a settlement API: submit an instruction with both legs, receive the settled state. The same call that a market operator makes in production is available on a testnet with a faucet, against the same interface.

• REST and SDK access for settlement, balances and transfers.
• Contract templates in Solidity and Rust for issuers defining instrument rules.
• Webhooks delivering settlement and transfer events as they are written.
• A testnet with a faucet, exposing the identical API with no value at risk.

Because settlement is atomic, an integration has one failure mode to handle rather than several: the instruction settled, or it did not. There is no partially-settled state to unwind in application code.

16Custody, security and risk

Custody

Instruments are held by the accounts that own them. The layer does not take custody of participant assets in order to settle them, and a venue cannot move a member's position without an authorisation from that member.

Failure modes worth stating plainly

• Key loss is asset loss. Self-custody moves the failure from an institution to a key holder; it does not eliminate it.
• An instrument is only as good as its issuer. The layer guarantees that title moved, not that the issuer will honour what the title represents.
• Smart contract risk is real. Settlement logic is code, and code has defects; issuers should treat contract review as a prerequisite, not a formality.
• Chain liveness bounds market liveness. If the network cannot include transactions, settlement pauses — the design makes that visible rather than silently deferring it.

17Status and disclaimer

What is live

The layer settles production traffic today. ContiSX operates under its SEC licence on it, and the markets listed in section 8 settle through it. The network figures published on this site — total settled value, daily settlements, average settlement time and uptime — describe the operator's own measurements and are restated there rather than audited here.

What this document is not

This paper describes a settlement system. It is not an offer to sell or a solicitation to buy any instrument, it is not investment advice, and it does not describe the merits of any asset that settles on the layer. Statements about future capability are intentions, not commitments, and may change. Anyone relying on this document for a regulatory, investment or integration decision should verify the current behaviour of the system directly.