Distinguish the techniques for following intraday flows from those for keeping watch on risk levels.
The idea
Monitoring tools fall into two families. Flow-following techniques record and project how money moves. They include real-time capture of every payment sent and received, the cumulative net position against its expected profile, balances at the central bank and correspondents, queues of unsettled payments, time-specific obligations, and throughput, meaning the share of outgoing payments settled by given times of day. Risk-level techniques compare consumption with capacity and tolerance. They include daily maximum usage, available liquidity at the start of the day, headroom or the usage-to-available ratio, limits and early-warning triggers, stress-test results, and credit lines extended to customers. Flow tools answer what is happening and when. Risk tools answer how close to the limit the bank is and what happens under stress. A low peak usage can hide late-day payment hoarding that only timing data reveals. Exam point: use the two together, never one alone.
Formulas
Intraday throughput
Throughput = settled by time ÷ total outgoing
- settled by time
- outgoing payments settled by the given time of day, cumulatively
- total outgoing
- all outgoing payments for the day
A flow measure of when payments are released. Use the cumulative amount settled, not just the latest interval.
Intraday liquidity utilisation
Utilisation = peak usage ÷ available liquidity
- peak usage
- the daily maximum intraday usage
- available liquidity
- intraday liquidity available at the start of the day
A risk-level measure; headroom is available liquidity minus peak usage. A low figure can hide late-day payment hoarding.
A practice question
Which of the following is a technique for following intraday flows rather than for monitoring risk levels?
- Setting a limit on daily maximum intraday liquidity usageWhy not: A limit is a risk-level control that compares usage with tolerance; it does not follow the flows themselves.
- Running a stress scenario in which a major counterparty stops sending paymentsWhy not: Stress testing is a risk-level tool that estimates the effect of adverse conditions on the position.
- Tracking the cumulative net position through the day against its expected profile using real-time payment data - correct
- Calculating the ratio of peak usage to available liquidity at the start of the dayWhy not: A usage-to-capacity ratio measures how close the bank is to its limit, which is a risk-level metric.
Flow techniques observe and project how payments move through the day. Limits, stress tests and utilisation ratios compare consumption with capacity and so belong to risk-level monitoring.
A harder one
A bank sent outgoing payments totalling USD 2,000 million today. Of these, USD 400 million had settled by 10:00, USD 1,100 million by 12:00 and USD 1,700 million by 14:00. The intraday throughput at 12:00, the share of total outgoing payments settled by that time, is:
- 20%Why not: Uses the 10:00 figure of 400 rather than the cumulative 1,100 settled by 12:00.
- 55% - correct
- 35%Why not: Counts only the payments settled between 10:00 and 12:00 (700) instead of everything settled by 12:00.
- 45%Why not: Reports the share still unsettled at 12:00 (900 of 2,000) rather than the share settled.
Throughput by 12:00 = cumulative payments settled / total outgoing payments = 1,100 / 2,000 = 55%. It is a flow measure showing when the bank releases payments.
The hardest tier
A bank has USD 380 million of available intraday liquidity. Its expected cumulative net position (receipts minus payments) has a trough of -310 at 12:00. At 11:15 the real-time flow tracker shows that a receipt of USD 100 million expected between 11:00 and 12:00 will not arrive until 16:00; everything else is on schedule. The projected peak usage and its implication are:
- USD 410 million, exceeding available liquidity by USD 30 million, so the bank should act now to mobilise collateral or defer non-time-specific payments - correct
- USD 310 million, leaving headroom of USD 70 million, so no action is neededWhy not: Ignores the delayed receipt that the flow tracker has flagged.
- USD 210 million, leaving headroom of USD 170 millionWhy not: Subtracts the delay from peak usage, treating a missing receipt as additional liquidity.
- USD 510 million, exceeding available liquidity by USD 130 millionWhy not: Adds the delayed receipt twice, once to the trough and again to a second outflow that does not exist.
The delayed receipt falls before the trough, so peak usage = 310 + 100 = 410 versus 380 available, a projected shortfall of 30. The flow tracker gives the early warning; the risk-level comparison tells how serious it is.
In the app: 6 practice questions on this outcome, spaced review that brings it back before you forget it.
Start free →