Everything upstream — authorization, clearing, netting — is bookkeeping until settlement discharges the obligation in central-bank money. This is the leaf where finality lives, where Herstatt risk was learned the hard way, and where SAMOS sits under South Africa.
Final transfer of value in the safest money there is.
Settlement is the act of transferring value to extinguish a payment obligation — the moment a debt between two banks actually moves and becomes final (irrevocable). Clearing calculates who owes whom; settlement pays it. Until settlement is final, an obligation can in principle fail.
The crucial detail: high-value and interbank settlement happens in central-bank money — balances banks hold at the central bank — because that is the only asset with no credit or liquidity risk attached. Commercial-bank money carries the risk of the bank failing; central-bank money does not. This is why the central bank sits at the base of the system.
Gross trades liquidity for safety; net trades safety for liquidity.
| Gross settlement (RTGS) | Net settlement (DNS) | |
|---|---|---|
| Mechanism | Each transaction settled individually, in real time | Obligations netted over a window, one figure settled per cycle |
| Liquidity needed | High — full value of each payment | Low — only the net position |
| Settlement risk | Minimal — immediate finality | Accumulates across the window until settled |
| Typical use | High-value / time-critical interbank payments | High-volume / low-value retail (cards, EFT) |
| Example | SAMOS (SA), TARGET2, Fedwire | Card and EFT clearing feeding a deferred net settlement |
The trade-off is liquidity versus risk. RTGS settles each payment immediately, so finality is instant but banks must fund the full value upfront. Deferred net settlement (DNS) needs far less liquidity but lets exposures build across the window. Most countries run both: RTGS for large/urgent, DNS for retail volume that nets down before touching the RTGS.
SAMOS is South Africa’s RTGS — now ISO 20022, and being renewed.
A Real-Time Gross Settlement system is the central bank’s engine for settling individual payments in central-bank money with immediate finality. South Africa’s is SAMOS — the South African Multiple Option Settlement system, operated by the SARB, live since 9 March 1998.
Net positions from BankservAfrica/PayInc clearing, and high-value interbank payments, settle across SAMOS in central-bank money. It is the bottom of the domestic stack.
SAMOS completed migration to ISO 20022 rich-data messaging in 2025, aligning SA’s settlement spine with the global standard.
The SARB launched an RTGS Renewal Programme (2021) to replace SAMOS with a next-generation system, part of the broader PEM modernisation.
TARGET2 (euro area), Fedwire Funds Service (US), CHAPS (UK) — every major currency has an RTGS as its finality layer.
Window length is a dial between liquidity cost and accumulated risk.
Net settlement runs in windows — scheduled cycles when accumulated net positions are settled across the RTGS. The window length shapes both liquidity demand and risk: shorter, more frequent windows reduce built-up exposure but demand liquidity more often; longer windows are liquidity-efficient but let risk accumulate.
Instant payment schemes compress this hard. PayShap (SA), FedNow, and SEPA Instant clear in seconds around the clock — but the interbank settlement behind them may still net into RTGS windows, with prefunding or settlement guarantees bridging the gap. The customer experiences instant; the settlement plumbing may still be deferred. Knowing the difference is essential for liquidity and risk management.
Herstatt is why finality and PvP matter — the gap between legs is where banks die.
Settlement risk is the risk that one party delivers its side of a transaction while the other fails before delivering theirs. The canonical case gave it a name: Herstatt risk.
Bankhaus Herstatt was shut down by German regulators mid-day, after it had received Deutsche Marks from counterparties but before it paid out the corresponding US dollars (different time zones). Counterparties lost the dollar leg entirely. The cross-currency timing gap became the textbook settlement-risk event.
Payment-versus-Payment (e.g. CLS for FX) and the spread of RTGS with intraday finality were direct responses — ensuring one leg cannot settle without the other, removing the open window.
Any time there is a gap between giving and receiving value — deferred net windows, cross-border, prefunded instant schemes — settlement risk is present. The discipline is to know exactly where your unfunded exposure sits and for how long.
Know your unfunded windows. Finality lives in central-bank money.
The question that matters operationally: at any given moment, who owes you value that is not yet final, and what happens if they fail before it settles? If you cannot answer that quickly, you are carrying settlement risk you have not priced. This is true for PSPs prefunding instant rails, for acquirers funding merchants ahead of settlement, and for anyone holding net positions across a window.
Every gap between clearing and final settlement is exposure. Know its size, duration, and counterparty — for cards, EFT, instant, and cross-border separately.
Settlement “in” a commercial bank is not final in the same sense — it carries that bank’s risk. For systemic exposure, finality means central-bank money (RTGS).
Underpricing settlement risk works until a counterparty fails inside your window — then it is a direct, often unrecoverable loss. Herstatt was not a rounding error; it reshaped global settlement. Treat the window as a live exposure, not a back-office detail.
For African and SA institutions: SAMOS gives strong domestic finality, but cross-border and correspondent flows still carry the classic timing gaps. Prefunded instant schemes (PayShap and peers) shift risk to the prefunding model — understand where the guarantee actually sits before you rely on “instant.”