Deferred Net Settlement vs Gross Settlement

Underneath every payment scheme is a settlement model that decides when and how banks actually exchange money with each other. The two dominant models are deferred net settlement (DNS) and real-time gross settlement (RTGS). They sound like plumbing details, but the choice drives how much liquidity a bank must hold, how much credit risk sits in the system, and how fast money finally becomes irrevocable. This post breaks down both models and the trade-offs that make each one useful.

Gross settlement: one payment at a time

In a gross settlement system, each payment is settled individually and immediately across accounts held at the central bank. If Bank A owes Bank B £1,000,000, that full amount moves from A's central-bank account to B's account as a discrete event. Nothing is batched or offset. Systems like the UK's CHAPS, the Eurozone's TARGET2 (now consolidated into T2), and the US Fedwire Funds Service are classic RTGS platforms.

The headline benefit is finality. Because each transfer settles in central-bank money the moment it is processed, there is virtually no settlement credit risk between the banks: the receiver has good funds instantly. That is why RTGS is the rail of choice for high-value and time-critical payments, where a failed settlement could cascade through the financial system.

The cost is liquidity. A bank must have the full amount available at the moment of each payment. On a busy day a large bank might push tens of billions through an RTGS system, and it needs enough intraday liquidity — balances plus central-bank credit lines — to cover the peaks. To ease this, RTGS operators run liquidity-saving mechanisms such as offsetting algorithms and payment queues that look for matching flows before consuming reserves.

Deferred net settlement: batch, net, and settle the difference

In a net settlement system, individual payments are exchanged and cleared throughout the day but not settled one by one. Instead the operator calculates each participant's net position — total incoming minus total outgoing — and settles only those net amounts at scheduled times. If Bank A sends B £1,000,000 across many small payments while B sends A £950,000, only the £50,000 net difference actually moves.

Netting dramatically reduces the value that must change hands. A scheme processing millions of retail payments might net down to a small fraction of gross value, so participants need far less liquidity. This is why most high-volume, low-value retail rails — legacy ACH batch systems, card networks, and many instant-payment schemes at the interbank layer — use deferred net settlement.

The catch: settlement risk in the window

The trade-off is timing risk. Between the moment a payment is cleared and the moment the net position settles, a participant has an obligation that is not yet backed by good funds. If a bank fails inside that window, its counterparties may not receive what they are owed. This is settlement risk (sometimes called Herstatt risk after the 1974 collapse of Bankhaus Herstatt).

Schemes manage this with safeguards: prefunded settlement accounts, net debit caps limiting how large a participant's obligation can grow, collateral pools, and loss-sharing arrangements so a survivor's default does not unwind the whole batch. Many modern instant-payment rails require participants to prefund a settlement balance, which removes most of the credit risk while keeping the liquidity efficiency of netting.

Instant payments blur the line

Real-time retail schemes create an interesting hybrid. The customer experience is instant and irrevocable — funds land in seconds, 24/7. But the interbank settlement often still happens on a deferred basis, several times a day or continuously against a prefunded balance. The scheme guarantees the payment to the receiver immediately, then reconciles the banks' positions on its own schedule. So a payment can be final to the end user while the underlying interbank settlement is still net and deferred behind the scenes.

How to choose

The decision usually comes down to value and volume:

• High value, low volume (large corporate transfers, bank-to-bank funding): gross settlement, because finality matters more than liquidity efficiency.
• Low value, high volume (retail payments, card transactions): net settlement, because netting saves enormous amounts of liquidity and the per-transaction risk is small.
• Instant retail at scale: prefunded net settlement, combining instant finality to the customer with liquidity-efficient interbank reconciliation.

Key takeaways

• Gross settlement moves each payment individually in central-bank money, giving immediate finality but demanding full intraday liquidity.
• Deferred net settlement offsets obligations and settles only the net, saving liquidity but introducing settlement risk during the netting window.
• Schemes mitigate net-settlement risk with prefunding, net debit caps, collateral, and loss-sharing rules.
• Instant-payment rails often pair customer-facing finality with deferred, prefunded interbank settlement underneath.
• Value and volume drive the choice: RTGS for large critical payments, netting for high-volume retail flows.