The foreign-exchange market still settles about $1.4 trillion of obligations each day on a gross bilateral basis without risk mitigation. That exposure sits after the trade, which is why it receives less attention than prices even though a failure can involve the full principal amount.
The BIS 2025 Triennial Survey measured more than $14 trillion in average daily gross settlement obligations during April 2025. Payment-versus-payment systems handled $5.2 trillion, or 36%, eliminating principal risk by coordinating the two currency legs. A further $7.6 trillion used netting, intragroup settlement or timing controls that reduced but did not eliminate risk.
| Method | Daily amount | Risk treatment |
|---|---|---|
| Payment versus payment | $5.2tn | Eliminates principal risk |
| Netting/intragroup/controls | $7.6tn | Mitigates risk |
| Gross bilateral | $1.4tn | Fully exposed |
The main reasons for gross bilateral settlement included counterparties without access to a relevant PvP system, ineligible currency pairs and ineligible trade types. This is not simply a technology problem. Coverage and participation determine which obligations can use the safest method.
The BIS found settlement fails equal to only 0.01% of gross obligations in the survey month. That is reassuring operationally, but frequency and severity are different measures. If one party delivers a currency and the counterparty fails before returning the other, the amount at risk is the principal, not only the expected profit on the trade.
Shorter securities-settlement cycles add pressure because asset managers have less time to arrange the FX needed to fund cross-border purchases. Automation, matching and accurate settlement instructions become more important when the repair window shrinks.
Retail traders do not manage institutional settlement systems directly, but the issue explains why market resilience depends on infrastructure as well as liquidity. A deep order book cannot compensate for a breakdown in payment completion.
This September FX market outlook discusses current prices. It does not address settlement architecture, so the BIS study remains the relevant source.
Published June 2026. The methodology differs from some earlier surveys, so comparisons must follow the BIS caveats.