Settlement failures: what happens when a bank cannot settle

A settlement failure is the event in which a participant cannot meet its settlement obligation at the settlement moment, and it is the scenario that every payment system operator plans for most carefully. The failure may be a liquidity failure, where the bank has the money but cannot access it in time, or a solvency failure, where the bank does not have the money at all, or an operational failure, where the bank systems are down and it cannot send the instruction. The three causes have very different consequences, and the operator must be able to distinguish them quickly, because the response to a liquidity problem is to provide time or credit, while the response to a solvency problem is to exclude the bank and protect the other participants. Why settlement failures matter is that they are the point at which a single bank problem becomes a system problem. In a deferred net settlement system, if one bank fails to fund its net debit position, the settlement cannot complete as calculated, and the operator must either delay settlement for everyone or exclude the failing bank and recompute. Delaying punishes the healthy banks, because their customers do not get credited, while excluding changes every other bank net position and may turn a bank that expected a credit into a net debtor. The PFMI requires operators to have a default management procedure that is tested regularly, and the Bank of England and the European Central Bank both publish their procedures. The parties are the failing bank, the other participants, the system operator, and the central bank. The operator detects the failure when the funding deadline passes without the funds arriving, or when the central bank reports that the settlement instruction cannot be executed. The operator then follows the default procedure in the rulebook, which typically gives the failing bank a short grace period, then excludes it, recomputes the net positions, and settles the remainder. The central bank may also intervene, for example by providing emergency liquidity against collateral, or by suspending the failing bank from the system. The regulator is notified, and the incident is recorded as a reportable event. In practice the failure sequence runs like this. At 15:00 the settlement cycle closes and the net positions are computed. Bank A has a net debit of EUR 25 million and must fund it by 15:30. At 15:30 the funds have not arrived, so the operator calls Bank A treasury and gives a 30-minute grace period. At 16:00 the funds still have not arrived, so the operator declares a settlement failure, excludes Bank A payments from the cycle, and recomputes the net positions for the remaining banks. Bank B, which expected a EUR 30 million credit, now has a EUR 5 million debit, because EUR 35 million of the payments it expected to receive came from Bank A. Bank B must now fund EUR 5 million by 16:30, or it too will fail. The settlement completes at 17:00, two hours late, and every bank that was affected must explain the delay to its customers. The remaining EUR 55 million of Bank A's sent payments went to participants outside this four-bank cycle, so they are settled separately and are not part of the recomputation. A worked example shows why exclusion changes everything. Four banks in a cycle: A sends 120 and receives 95 (net debit 25), B sends 80 and receives 110 (net credit 30), C sends 60 and receives 55 (net debit 5), D sends 40 and receives 40 (net zero). If A fails and is excluded, the payments involving A are removed. B now sends 80 and receives 75 (because 35 of its receipts came from A), so its net position flips from a 30 credit to a 5 debit. C sends 60 and receives 55, but 20 of its receipts came from A, so it now sends 60 and receives 35, a 25 debit. D sends 40 and receives 40, but 10 of its receipts came from A, so it now sends 40 and receives 30, a 10 debit. The total net debit is now 40, and the operator must collect it from three banks that did not expect to pay. What happens after the failure is as important as the failure itself. The operator must investigate the cause, report to the regulator, and review whether the default procedure worked. The failing bank must explain why it could not fund, and if the cause was a liquidity problem, it must show how it will prevent a recurrence. If the cause was solvency, the bank may be suspended from the system and its payments may be returned. The other participants must also review their exposure: a bank that had a large net credit from the failing bank has learned that its expected receipts are not certain, and it may change its own liquidity management or its credit limits on that counterparty. The operational and accounting implications are severe. A settlement failure creates a backlog of unsettled payments, each of which must be tracked, and the reconciliation team must be able to show the status of every payment in the failed cycle. On the ledger, payments that were exchanged but not settled remain in the clearing account, and the bank must decide whether to credit its customers anyway, which creates exposure, or to hold the credit until settlement, which creates customer complaints. The treasury team must also manage the unexpected funding requirement, because a bank that expected a credit and now owes money must find liquidity at short notice. For a business analyst, settlement failure is a requirements and testing subject. You must specify the funding deadline, the grace period, the exclusion procedure, the recomputation logic, the notification requirements, and the reporting. You must also write test cases for each failure cause: a liquidity failure, an operational failure, and a solvency failure, and verify that the system behaves correctly in each. A common gap is failing to specify what happens to payments that were already credited to customers before the settlement failed, because that is the point at which the bank takes a real loss. Takeaway: a settlement failure is not a single bank problem, it rewrites every other bank position in the cycle, and the default procedure is the most important page in the rulebook. Further reading: the BIS Principles for Financial Market Infrastructures (PFMI) for default management; the Bank of England CHAPS default procedures; the ECB TARGET guideline on settlement failure. Last reviewed 7 November 2026.