How a payment actually moves

Paying someone at another bank isn't one instant event — it's a sequence of messages and promises across ledgers that only becomes final at settlement.

FoundationsFinancial Infrastructure~13 min

Before this lesson

Lesson: Money is a ledger

When Alice and Bob bank at the same institution, paying Bob is trivial: one ledger, two entries, done. But most payments cross between banks — and now two independent ledgers have to end up consistent without either bank simply trusting the other. A payment is therefore not a single instant; it is a sequence.

The stages of a payment

Follow $100 from Alice (at Bank A) to Bob (at Bank B):

Instruction — Alice tells Bank A to pay Bob. Bank A debits Alice.
Clearing — Bank A passes the payment through a shared clearing system that routes it to Bank B and records that Bank A now owes Bank B $100.
Credit — Bank B credits Bob. Bob can see and often spend the money.
Settlement — Bank A actually transfers $100 to Bank B, usually in central-bank reserves. Only now is the payment truly final and irreversible.

Step through it in the widget below and watch the balances — and the interbank obligation — change at each stage.

Credited is not settled

The subtle, important point lives between steps 3 and 4. Bob has his money before the banks have settled between themselves. In that window, Bank A still owes Bank B. If Bank A failed at that exact moment, Bank B could be left out of pocket having already paid Bob. That exposure is settlement risk, and managing it is a central concern of every payment system — the reason banks hold reserves, post collateral, and set limits on each other.

Why there are so many "rails"

Different payment rails make different trade-offs on exactly this timing:

Real-time gross settlement (RTGS) settles each payment individually and immediately in central-bank money — no settlement risk, but it consumes a lot of reserves and is reserved for large or urgent payments.
Deferred net settlement (most retail rails, card networks, ACH) lets payments accumulate and settles only the net amounts later — cheap and efficient, but it carries risk during the delay.
Instant payment schemes push retail payments toward real-time credit while managing the settlement risk behind the scenes.

Cards, bank transfers, ACH, RTGS, instant rails — they differ mostly in this speed-versus-risk-versus-cost balance. The netting that makes deferred settlement so efficient deserves its own look: clearing and settlement, next.

Try it: how a payment moves

Alice

→

Bank A

→

Clearing /central bank

→

Bank B

→

Bob

Alice

$100

Bob

Bank A owes Bank B

Step 0: Before

Alice has $100 at Bank A. She wants to pay Bob $100, who banks at Bank B.

A payment isn’t one instant event — it’s a sequence of messages and promises that only becomes final at settlement. Real-time gross settlement (RTGS) settles each payment immediately to remove the risk in step 3; deferred net settlement waits and nets many payments together to save on reserves. That trade-off between speed, cost and risk is the heart of payment-system design.

Key takeaways

A cross-bank payment moves through instruction, clearing, crediting the payee, and finally settlement between the banks.
Being credited is not the same as settled — the payee can have funds before the banks square up, creating settlement risk.
Payment rails differ mainly in how fast and how finally they settle, and at what cost.

← All Financial Infrastructure lessons

Bring us the problem, not a perfect specification

Tell us what needs to change, who it affects and any important deadline. We will review the context and reply with useful next questions.

01Share contextDescribe the workflow, constraint or risk.
02Clarify togetherWe identify missing facts and useful options.
03Choose a startAgree a focused assessment or delivery step.

Start a conversation