Hybrid Settlement System

A hybrid settlement system is a payment system that combines features of real-time gross settlement and net settlement to reduce the liquidity participants need while keeping settlement risk low. Payments are settled individually and with finality, as in a gross system, but the system also queues payments and offsets them against each other at frequent intervals, as in a netting system. The result is that a bank can often settle large outgoing payments using incoming ones rather than its own balance.

The term arose as designers of large-value systems tried to escape a trade-off. Pure deferred net settlement saves liquidity but leaves participants exposed to each other until the end-of-day settlement, while pure real-time gross settlement (RTGS) removes that exposure but needs much more liquidity. Today the line has blurred, because most RTGS systems include liquidity-saving mechanisms that give them a hybrid character.

Think of three friends who owe each other money after a trip. One option is for each to pay every debt in full, straight away, which needs everyone to have cash on hand. Another is to wait until the end of the week and pay only the differences, which needs less cash but means everyone is trusting the others in the meantime. A hybrid takes a middle path: every so often, the group checks who owes what, and whenever debts cancel out, those are cleared on the spot.

Each debt cleared this way is truly settled at that moment, so nobody is left holding an IOU for the rest of the week. But because debts are matched against each other, the friends need far less cash in their pockets than if they paid everything one by one.

Designs vary, but most hybrid systems share a few elements:

• Central queue. Payments that cannot settle immediately because the sender's balance is too low wait in a queue rather than being rejected.
• Offsetting algorithms. At set intervals or when triggered, the system searches the queue for payments that can settle together, either between two banks or across many.
• Simultaneous final settlement. Each payment in an offset group settles in full and becomes final at the same moment, so there is no later unwinding if a bank fails.
• Funded balances. Participants hold funded positions, often in central bank money or in prefunded accounts, and the system checks that each bank's net effect stays within its balance.

Unlike a classic net system, the hybrid does not build up unsettled positions that are only paid off at the end of the day. Unlike a plain RTGS system, it does not demand that each payment be covered by available funds the instant it is submitted.

CHIPS, the US dollar large-value system run by The Clearing House, is the best-known system designed as a hybrid. Since a redesign in 2001 it has released payments with real-time finality by applying them against prefunded balances, using algorithms that match and offset payments bilaterally and multilaterally throughout the day, with a closing process at the end of the day for any remaining payments. Many central bank RTGS systems moved in the same direction from the other side. The Eurosystem's T2 and its predecessor TARGET2, the Bank of Japan's BOJ-NET and the Bank of England's RTGS service for CHAPS have all added queuing and offsetting features, which is why the difference between RTGS and hybrid systems is now often treated as a matter of degree.

The main benefit is lower intraday liquidity needs. Because payments that offset each other settle together, banks need smaller balances or less intraday credit from the central bank. Hybrids also reduce the risk of gridlock, where every bank waits for incoming funds before sending its own payments.

There are costs. Queued payments are delayed, which matters for time-critical transfers, so systems usually let senders mark payments as urgent or set priorities. The algorithms add complexity that participants must understand, and operators must make sure their rules give clear settlement finality for payments settled by offsetting. Where balances are held in commercial bank money rather than central bank money, participants also face some exposure to the institution holding them.

Hybrid design also appears in large-value transfer systems outside the main reserve currencies. Some retail and instant payment arrangements settle in batches across the day, but those usually rely on multilateral netting with prefunding rather than a central queue.

• Committee on Payments and Market Infrastructures (CPMI)
• The Clearing House