RTGS vs deferred net settlement
Real-time gross settlement settles every payment individually and immediately; deferred net settlement accumulates payments and settles only the net positions at designated times. The trade is liquidity cost against credit risk.
| DIMENSION | RTGS | Deferred net settlement |
|---|---|---|
| Settlement unit | Each payment settles individually, at its gross (full) amount. | Many payments are accumulated and offset; only each participant's net position settles. |
| When settlement happens | Continuously, payment by payment, as soon as each is accepted and funded. | At designated settlement times — one or more cycles per day, depending on the system. |
| Finality | Immediate per payment: once settled, each transfer is final and irrevocable. | At cycle settlement: until the net positions settle, individual payments in the batch are not final. |
| Liquidity demand | High — participants need enough liquidity to fund payments at full value all day. Systems soften this with queuing, offsetting algorithms, and central bank intraday credit against collateral. | Lower — netting means a participant funds only its net obligation, a fraction of gross turnover. |
| Credit risk between participantsInternational risk standards for systemically important systems exist largely to keep this deferred-settlement exposure controlled. | Essentially none accumulates: obligations are extinguished as they arise. | Exposure builds up between cycles — if a participant fails before settlement, others face an unwind. Systems manage this with limits, collateral, prefunding, or guarantee arrangements. |
| Typical use | High-value and urgent payments: T2 in the euro area, CHAPS in the UK, Fedwire in the US. | High-volume retail payments where per-payment urgency is lower: for example SEPA batch clearing through STEP2, settling net positions in central bank money. |
| Cost per paymentActual pricing varies by system and participant tier; this row is directional, not a tariff. | Higher — the liquidity consumed and the infrastructure justify it for large or time-critical amounts. | Lower — netting and batching spread costs across huge volumes, which is why bulk retail rails use it. |
| What failure looks like | A payment that cannot be funded queues or is rejected; the rest of the system carries on. | A participant default before settlement triggers the system's default arrangements — recalculating positions or drawing on collateral — which is operationally rare but severe. |
| Where instant payments fitThis is why 'instant' describes the customer experience, not necessarily the settlement model underneath — always check how a given system actually settles. | Instant settlement services (such as TIPS) apply the gross, immediately-final model to retail payments, made affordable through pre-funded positions. | Some instant schemes instead clear in real time but settle net between cycles, protecting the gap with prefunding or guarantees. |
Sources for this comparison4
- Market practiceMarch 2003 edition
A glossary of terms used in payments and settlement systems ↗ — CPSS (now CPMI), Bank for International Settlements · definitions of RTGS and net settlement systems
Terminology has evolved since this edition; newer CPMI publications refine some definitions.
- Market practice
Principles for financial market infrastructures ↗ — CPMI and IOSCO (Bank for International Settlements) · credit and liquidity risk management principles
Published by the CPSS (now CPMI) and IOSCO; contains 24 principles plus responsibilities for authorities. This site uses it only for high-level concepts such as settlement finality.
- Scheme-specific rule
TARGET Services ↗ — European Central Bank · T2 RTGS service description
T2 replaced TARGET2 in March 2023. Detailed user functional specifications are published separately in the ECB's professional-use documents section.
- Simplified educational illustration
Payments Signal editorial teaching models — Payments Signal
What this simplifies: Real systems are hybrids: RTGS services use netting-like offsetting algorithms in their queues, and DNS systems use prefunding that resembles gross liquidity. The two columns present the pure models; named systems are examples, not complete descriptions.
Used wherever diagrams, scenarios, figures, or example values are didactic constructions rather than sourced facts; every such use carries a simplifications disclosure. All people, companies, banks, and list entries in examples are fictional.