A liquidity-saving mechanism (LSM) is a feature of a payment system, usually a real-time gross settlement system, that lets participants settle the same payments with less money on their settlement accounts. It does this mainly by holding payments in a queue and settling groups of them together when they offset one another, so that incoming payments help fund outgoing ones. Banks use LSMs to cut the cost of funding their payments during the day, and central banks build them to make payment flows smoother.
LSMs became common in the 2000s and 2010s as central banks looked for ways to reduce the large intraday liquidity needs of real-time gross settlement (RTGS). They are what turns a plain RTGS system into something closer to a hybrid settlement system. Unlike end-of-day netting, an LSM settles each payment in full and with finality at the moment it is released, so it saves liquidity without creating credit exposures between banks.
Liquidity-saving mechanisms in plain terms
Imagine two neighbours who each owe the other £100. If each must have £100 in hand before paying, both need cash, even though the debts cancel out. A sensible arrangement is to notice the match and settle both at once, with no cash changing hands. A liquidity-saving mechanism does this for banks, many times a day and across many banks at once.
It also breaks standoffs. If every bank waits for money to arrive before sending its own payments, nothing moves. With a shared waiting room for payments, the system can spot circles of payments that balance each other and release them together, a bit like a sorting office that ships parcels in matched pairs rather than holding everything until someone pays the postage.
Main techniques
- Central queuing. Payments that cannot settle immediately wait in a queue managed by the system rather than being rejected or held back in each bank's own systems.
- Bilateral offsetting. The system looks for payments between two banks that can settle simultaneously, so only the difference needs funding. This resembles bilateral netting, but each payment still settles gross.
- Multilateral offsetting. An algorithm searches across all queued payments for a set that can settle together given every bank's balance, applying the logic of multilateral netting while keeping gross settlement.
- Priorities. Senders can mark payments as urgent, normal or less urgent, so time-critical payments are not held up by the queue.
- Reservations and limits. Banks can reserve part of their balance for particular purposes, or set bilateral and multilateral sending limits so that they do not drain liquidity to a counterparty that is not sending payments back.
How it works in practice
A simple case shows the effect. Bank A has a payment of 50 million to send to Bank B, and Bank B has 40 million to send to Bank A. In plain RTGS, Bank A needs 50 million on its account and Bank B needs 40 million, or they must wait for each other. With bilateral offsetting both payments settle together: Bank A's balance falls by 10 million and Bank B's rises by 10 million. Only 10 million of liquidity is used instead of 90 million. Across a system with hundreds of participants, multilateral offsetting can save far more.
Many RTGS systems now include an LSM. The Eurosystem's TARGET2, launched in 2007, included queuing and offsetting algorithms, and its successor T2 keeps them. The Bank of England added a liquidity-saving mechanism to its RTGS service for CHAPS in 2013, and other central banks use similar features. Private systems such as CHIPS were designed around offsetting from the start.
Benefits and trade-offs
LSMs reduce the amount of collateral banks must pledge to the central bank for intraday credit, lower the risk of gridlock, and encourage banks to send payments earlier in the day, which spreads liquidity risk more evenly. They can also make an operational problem at one bank less damaging, because others can still settle among themselves.
The trade-offs are delay and complexity. A payment waiting for an offsetting partner settles later than it would if fully funded, so banks must manage priorities and deadlines carefully. Algorithms must be transparent enough for participants to predict the results of their liquidity management. Operators also need clear rules on settlement finality for payments released by offsetting, and on what happens to queued payments when a participant is suspended.
LSMs compared with deferred netting
An LSM and a deferred net settlement system both use offsetting to save liquidity, but they differ on risk. In deferred net settlement, payments are accepted throughout the day and only settled later, so participants carry exposures to one another until then. In an LSM, nothing is settled until it can be settled in full, and each released payment is final. That makes LSMs a way to keep the safety of gross settlement while recovering much of the efficiency of netting.