How Do Banks Settle Millions of Payments? Cancel Almost All of Them

If many parties owe each other many amounts, settling every obligation separately means moving an enormous total sum, all of which must be funded. An institution that collects every obligation and calculates what each party owes on balance reduces that to a single figure per party, and because most obligations offset each other the total actually moved is a small fraction of the gross. Reductions of ninety per cent and more are ordinary. That saves funding, reduces the number of transactions and shortens the time during which anybody is exposed, which is why the arrangement appeared as soon as the volume of payments made individual settlement impractical.

The origin is informal and the story is well documented:

• •Bank clerks in London walked the city exchanging cheques by hand
• •They began meeting at a coffee house to swap them in one place
• •That meeting became a formal daily session in dedicated premises
• •Only the net differences between banks were then settled
• •The same pattern appeared independently in other financial centres
• •Central banks eventually took over holding the settlement accounts

The more consequential function came later and is what the institutions are now mainly about. An organisation can insert itself into every trade as the counterparty to both sides, becoming the buyer to every seller and the seller to every buyer, so that neither party depends on the other any more. That means a participant who fails does not drag down whoever they traded with, since the institution stands behind the obligation. Achieving that requires collecting collateral from every participant, calculating it daily or more often as prices move, and maintaining a fund contributed by members to cover a failure. The risk is not eliminated, it is concentrated deliberately in one place that is regulated and capitalised accordingly.

The collateral system is the machinery that makes the guarantee credible and it operates continuously rather than at settlement. An initial amount is demanded when a position is opened, calculated to cover the loss that would be expected if the position had to be closed out over a day or two in difficult conditions. A further amount is demanded or returned every day, and sometimes several times a day, as prices move, so that unrealised losses are paid as they occur rather than accumulating. Members also contribute to a default fund that covers losses beyond the collateral of whoever failed. The sequence in which those resources are used is defined in advance, which is what makes a failure orderly.

After 2008 regulators pushed enormous volumes of derivative trading into these institutions, on the reasoning that opaque bilateral obligations had made the crisis impossible to contain. That has worked in the sense that positions are now visible and collateralised. It has also created a small number of institutions whose failure would be catastrophic, which is the concern repeatedly raised since. If prices move violently the institution demands more collateral from everybody at once, which forces participants to find cash in exactly the conditions where cash is hardest to find, and that can amplify a shock rather than absorb it. Whether the arrangement is safer overall is genuinely argued about, and the honest answer is that it has not yet been tested by a major failure.

Collecting everybody's obligations and offsetting them means only a small net residue has to move, frequently under a tenth of the gross. Stepping between both sides of every trade removes the dependence of each party on the other, funded by collateral collected continuously and by a members' fund. That concentrates risk deliberately in one regulated place, which is safer in ordinary conditions and untested in a severe one.