SWIFT Payment Statistics 2026
By Axis Intelligence Research
Co-author: Sarah Davis | Last updated: August 25, 2026 | License: CC BY 4.0
Swift carried 15.1 billion FIN messages in 2025, averaging 59.8 million per business day across more than 11,500 institutions in over 200 countries and territories. Traffic grew 12.1% year on year, the network’s second consecutive double-digit year, while the US dollar held 50.99% of global Swift payment value in July 2026.
Quick Answer
Swift moved 15,106 million FIN messages in 2025, up 12.1% on 2024, with payments accounting for 6,573 million messages (43.5% of traffic) and securities for 7,785 million (51.5%). The US dollar took 50.99% of global payment value in July 2026 and 59.58% once intra-eurozone payments are stripped out. According to Axis Intelligence Research, only 28.97% of Swift payment messages leave their region of origin, meaning the network’s dominant use is regional rather than intercontinental.
Key Findings
- According to Axis Intelligence Research, Swift’s Cross-Border Flow Intensity (CFI) reached 47.59% for the Americas region in 2025 against 17.32% for EMEA, a 2.75x gap in how often a region’s outbound payment traffic actually crosses a continental boundary.
- According to Axis Intelligence Research, the Speed Dispersion Ratio (SDR) for wholesale cross-border payments stood at 3.02x in the first quarter of 2025, with 73.1% of payments credited within one hour in North America against 24.2% in Africa.
- Swift reported 15,106 million FIN messages in 2025 at an average of 59.8 million per business day, a 12.1% increase on 2024, according to its Annual Review 2025.
- The US dollar accounted for 79.86% of trade finance value on Swift in July 2026 against 50.99% of global payment value, a gap of 28.87 percentage points that Axis Intelligence Research attributes to commodity invoicing concentration rather than settlement preference.
- According to Axis Intelligence Research, Swift connects roughly 6.29 times as many institutions as CIPS counts participants, comparing Swift’s 11,500 connected institutions against the 1,829 direct and indirect CIPS participants recorded in June 2026.
How Many Messages Does SWIFT Process Per Day?
Swift averaged 59.8 million FIN messages per day in 2025 and recorded a single-day peak above 68 million. Total annual volume reached 15,106 million messages, per the Swift Annual Review 2025.
That daily figure carries a methodological detail almost every secondary source gets wrong. Dividing 15,106 million by 59.8 million returns 252.6, not 365. According to Axis Intelligence Research, Swift’s headline daily average is computed over business days, not calendar days. A calendar-day reading of the same annual total yields 41.4 million messages per day. Anyone quoting 59.8 million as a seven-day figure is overstating daily throughput by 44.5%.
SWIFT FIN Traffic by Year, 2020 to 2025
| Year | Total FIN messages (millions) | Payments (millions) | Securities (millions) | Source |
|---|---|---|---|---|
| 2020 | 9,579 | 4,366 | 4,710 | Swift Annual Review 2025 |
| 2021 | 10,659 | 4,865 | 5,269 | Swift Annual Review 2025 |
| 2022 | 11,436 | 5,173 | 5,715 | Swift Annual Review 2025 |
| 2023 | 11,950 | 5,422 | 5,974 | Swift Annual Review 2025 |
| 2024 | 13,493 | 6,046 | 6,792 | Swift Annual Review 2025 |
| 2025 | 15,106 | 6,573 | 7,785 | Swift Annual Review 2025 |
Total FIN volume grew 57.7% between 2020 and 2025, a compound annual rate of 9.54%. Securities outpaced payments over the period, compounding at 10.57% against 8.53%.
SWIFT Traffic Mix by Market, 2025
| Market | Messages (millions) | Share of FIN traffic | Growth vs 2024 | Source |
|---|---|---|---|---|
| Securities | 7,785 | 51.5% | 14.7% | Swift Annual Review 2025 |
| Payments | 6,573 | 43.5% | 8.9% | Swift Annual Review 2025 |
| Treasury | 698 | 4.6% | Not disclosed | Swift Annual Review 2025 |
| Trade finance | 28 | 0.2% | Not disclosed | Swift Annual Review 2025 |
| System | 22 | 0.1% | Not disclosed | Swift Annual Review 2025 |
Sarah Davis: The payments share of Swift traffic fell 2.1 percentage points between 2020 and 2025, from 45.6% to 43.5%. Read that carefully, because the obvious inference is wrong. Payments volume rose 50.5% over the same window. What happened is that securities messaging grew faster, pulled by shortened settlement cycles and fixed income allocation. Swift is frequently described as the international wire network. On its own traffic mix it is a securities messaging network that also carries payments, and the securities side has been the growth engine for five straight years. That distinction matters for anyone pricing the franchise, because securities messaging revenue does not respond to the same pressures as the payment leg does. Our wire transfer statistics breakdown covers the retail end of that payment leg in more detail.
What Share of SWIFT Payments Is in US Dollars?
The US dollar held 50.99% of global Swift payment value in July 2026, ahead of the euro at 21.78% and sterling at 6.97%, according to the Swift Global Currency Tracker August 2026 edition. Excluding payments made within the eurozone, the dollar rises to 59.58% and the euro falls to 13.36%.
SWIFT Payment Currency Share, July 2026
| Rank | Currency | Global payments (% of value) | Excluding intra-eurozone (% of value) | Source |
|---|---|---|---|---|
| 1 | USD | 50.99 | 59.58 | Swift Global Currency Tracker, August 2026 |
| 2 | EUR | 21.78 | 13.36 | Swift Global Currency Tracker, August 2026 |
| 3 | GBP | 6.97 | 5.22 | Swift Global Currency Tracker, August 2026 |
| 4 | JPY | 3.35 | 4.87 | Swift Global Currency Tracker, August 2026 |
| 5 | CNY | 3.10 | 2.34 | Swift Global Currency Tracker, August 2026 |
| 6 | CAD | 2.76 | 2.60 | Swift Global Currency Tracker, August 2026 |
| 7 | HKD | 1.84 | 1.43 | Swift Global Currency Tracker, August 2026 |
| 8 | AUD | 1.58 | 1.88 | Swift Global Currency Tracker, August 2026 |
| 9 | SGD | 1.12 | 0.76 | Swift Global Currency Tracker, August 2026 |
| 10 | CHF | 0.91 | 1.80 | Swift Global Currency Tracker, August 2026 |
Note the ranking inversion between the two columns. The renminbi ranks fifth on the global measure and sixth once intra-eurozone traffic is removed, because the Canadian dollar overtakes it. According to Axis Intelligence Research, removing intra-eurozone payments adds 8.59 percentage points to the dollar’s share and subtracts 8.42 points from the euro’s, the single largest methodological swing in the dataset. Headlines citing one figure against the other, without naming which basis they used, are comparing different populations.
Currency Concentration on the SWIFT Network
Axis Intelligence Research computes a Herfindahl-Hirschman concentration reading from the ranked currency shares: 3,161 index points on the global basis and 3,803 excluding intra-eurozone payments. The formula and inputs appear in the Methodology section. Both readings sit far above the 2,500-point threshold that competition authorities treat as highly concentrated in product markets. The concentration is not a dollar story alone. The top three currencies take 79.74% of global payment value, leaving 20.26% for every other currency on the network combined.
SWIFT Trade Finance Currency Share, July 2026
| Rank | Currency | Share of trade finance value (%) | Source |
|---|---|---|---|
| 1 | USD | 79.86 | Swift Global Currency Tracker, August 2026 |
| 2 | CNY | 8.43 | Swift Global Currency Tracker, August 2026 |
| 3 | EUR | 6.24 | Swift Global Currency Tracker, August 2026 |
| 4 | JPY | 1.40 | Swift Global Currency Tracker, August 2026 |
| 5 | AED | 0.76 | Swift Global Currency Tracker, August 2026 |
| 6 | SAR | 0.75 | Swift Global Currency Tracker, August 2026 |
| 7 | THB | 0.65 | Swift Global Currency Tracker, August 2026 |
| 8 | VND | 0.56 | Swift Global Currency Tracker, August 2026 |
| 9 | IDR | 0.39 | Swift Global Currency Tracker, August 2026 |
| 10 | GBP | 0.13 | Swift Global Currency Tracker, August 2026 |
Sarah Davis: The interesting number here is the euro, not the dollar. The euro takes 21.78% of Swift payment value and 6.24% of trade finance value, a ratio of 0.29. The renminbi runs the opposite way: 3.10% of payments, 8.43% of trade finance, a ratio of 2.72. Trade finance instruments follow the invoice currency, and invoice currency follows the commodity. Europe pays for a great deal in euros and invoices comparatively little in them outside the bloc. China invoices in renminbi where it has pricing power and settles in dollars where it does not. If you want an early read on currency internationalisation, watch the trade finance column and ignore the payments column, because letters of credit commit to a currency months before the payment message does. Sterling at 0.13% of trade finance against 6.97% of payments is the same phenomenon in reverse, and it is a fair description of what the City actually sells.
Which Regions Send the Most SWIFT Payment Traffic?
EMEA originated 13,244 thousand payment messages per day in 2025, against 6,317 thousand for the Americas and 3,924 thousand for Asia-Pacific. Total outbound traffic across the three regions came to 23,485 thousand messages per day, which reconciles with total inbound traffic of 23,486 thousand, a one-unit rounding difference that confirms the regional table is a closed system.
SWIFT Regional Payment Traffic, 2025 Average Daily Volume
| Region | Sent (Kmsgs/day) | Received (Kmsgs/day) | Intra-regional (Kmsgs/day) | Share of outbound | Source |
|---|---|---|---|---|---|
| EMEA | 13,244 | 13,796 | 10,950 | 56.39% | Swift Annual Review 2025; share Axis-calculated |
| AME | 6,317 | 5,811 | 3,311 | 26.90% | Swift Annual Review 2025; share Axis-calculated |
| APAC | 3,924 | 3,879 | 2,420 | 16.71% | Swift Annual Review 2025; share Axis-calculated |
EMEA is the only net receiver of the three, taking in 552 thousand more messages per day than it sends. The Americas run a net outbound position of 506 thousand, Asia-Pacific a marginal 45 thousand.
The Axis Cross-Border Flow Intensity Metric (CFI)
CFI stands for Cross-Border Flow Intensity. It measures the share of a region’s outbound Swift payment messaging that actually crosses out of that region, separating genuine intercontinental flow from regional traffic that happens to travel on an international rail.
Formula
CFI = (Regional Sent - Regional Intra-regional) / Regional Sent x 100
Inputs: average daily FIN payment messaging volume by region, in thousands of messages, as published in the Swift Annual Review 2025.
Worked example, EMEA
CFI_EMEA = (13,244 - 10,950) / 13,244 x 100
= 2,294 / 13,244 x 100
= 17.32%
CFI Readings by Region, 2025
| Region | Sent (Kmsgs/day) | Cross-region (Kmsgs/day) | CFI | Source |
|---|---|---|---|---|
| AME | 6,317 | 3,006 | 47.59% | Axis Intelligence Research, from Swift Annual Review 2025 |
| APAC | 3,924 | 1,504 | 38.33% | Axis Intelligence Research, from Swift Annual Review 2025 |
| EMEA | 13,244 | 2,294 | 17.32% | Axis Intelligence Research, from Swift Annual Review 2025 |
| Network | 23,485 | 6,804 | 28.97% | Axis Intelligence Research, from Swift Annual Review 2025 |
Interpretation. A high CFI means the region uses Swift primarily to reach other continents. A low CFI means the region uses Swift primarily to move value among its own neighbours. According to Axis Intelligence Research, EMEA’s 17.32% reading is the clearest single measure of what European integration has done to payment geography: nearly 83% of EMEA outbound Swift payment traffic stays inside EMEA. The Americas, at 47.59%, use the network for close to half its outbound volume to reach beyond the region.
Scope. CFI is computed on message counts, not values, and Swift’s regional table is published on a different basis from its FIN market-share table. Dividing 2025 payments volume of 6,573 million by 252.6 business days yields 26.02 million messages per day against the regional table’s 23.485 million. Axis Intelligence Research does not bridge the two series and publishes both figures as reported rather than reconciling them silently. The regional table’s internal balance between sent and received traffic is what supports its use here.
Growth by Route, 2025
| Route | Growth vs 2024 | Source |
|---|---|---|
| EMEA to AME | 11.8% | Swift Annual Review 2025 |
| APAC to EMEA | 10.1% | Swift Annual Review 2025 |
| AME to EMEA | 9.9% | Swift Annual Review 2025 |
| AME to APAC | 9.2% | Swift Annual Review 2025 |
| APAC intra-regional | 8.9% | Swift Annual Review 2025 |
| AME intra-regional | 8.8% | Swift Annual Review 2025 |
| APAC to AME | 8.1% | Swift Annual Review 2025 |
| EMEA to APAC | 7.7% | Swift Annual Review 2025 |
| EMEA intra-regional | 4.5% | Swift Annual Review 2025 |
Sarah Davis: Every one of the six cross-region routes grew faster than EMEA’s intra-regional traffic. The slowest cross-region leg, EMEA to APAC at 7.7%, still beat EMEA’s domestic 4.5% by 3.2 points. That is the CFI reading in motion: European regional traffic is the mature, low-growth block, and it is large enough at 10,950 thousand messages a day to hold the whole EMEA growth rate down to 5.6%. Anyone modelling Swift volume growth off a global number is averaging a saturated European core with genuinely expanding transatlantic and transpacific corridors, and the average tells you nothing useful about either. The fastest route on the board, EMEA to AME at 11.8%, is the one that a European bank’s treasury desk would recognise as the dollar-funding leg.
How Fast Are SWIFT Cross-Border Payments in 2026?
Swift reports that 75% of payments reach the beneficiary bank within ten minutes and that 86% travel directly or through a single intermediary. The Financial Stability Board, using Swift transaction data, reports that just over 50% of wholesale cross-border payments were credited to the end customer within one hour in the first quarter of 2025, and above 90% within one business day.
Both figures are correct. They measure different things. Swift’s ten-minute statistic covers the in-flight leg, ending when the payment instruction arrives at the beneficiary bank. The FSB measure runs to the moment the beneficiary’s account is credited. Swift’s own research, published in September 2025, found that on average 80% of a cross-border transaction’s total elapsed time occurs after the payment arrives at the beneficiary bank.
Wholesale Cross-Border Payment Speed by Region, Q1 2025
| Region | Credited within one hour | Credited within one day | Source |
|---|---|---|---|
| North America | 73.1% | Almost all | FSB Consolidated progress report 2025 |
| Central and Latin America | 42.7% | Not disclosed | FSB Consolidated progress report 2025 |
| Middle East | 32.3% | 87.1% | FSB Consolidated progress report 2025 |
| Asia-Pacific | 25.6% | 84.6% | FSB Consolidated progress report 2025 |
| Africa | 24.2% | Not disclosed | FSB Consolidated progress report 2025 |
Wholesale is defined by the FSB as payments valued at USD 100,000 or above.
The Axis Speed Dispersion Ratio (SDR)
SDR stands for Speed Dispersion Ratio. It expresses how unevenly cross-border payment speed is distributed across the world by dividing the fastest region’s one-hour crediting share by the slowest region’s.
Formula
SDR = Fastest region one-hour crediting share / Slowest region one-hour crediting share
Worked example, Q1 2025
SDR = 73.1 / 24.2 = 3.02x
Interpretation. According to Axis Intelligence Research, the SDR reading of 3.02x for the first quarter of 2025, against an absolute spread of 48.9 percentage points, shows that the binding constraint on cross-border payment speed sits in domestic infrastructure and beneficiary-side processing rather than in the messaging network. The same Swift rails serve North America and Africa. The outcomes differ by a factor of three.
The G20 target is 75% of payments credited within one hour by the end of 2027. North America sits 1.9 percentage points short of it. Africa sits 50.8 points short. Africa did improve, moving from 20.4% in the first quarter of 2023 to 24.2% two years later, and the Middle East gained 6.6 points over the same window to reach 32.3%. The FSB’s own assessment is that satisfactory global improvement by 2027 is unlikely.
Sarah Davis: There is a temptation to read the regional spread as a story about correspondent banking withdrawal, and that is part of it, but the FSB is specific about the mechanism and it is duller than de-risking. Capital control processing, beneficiary verification, incomplete supporting documentation and manual sanctions review are what consume the hours. The IMF and FSB found that capital control checks alone can delay crediting by hours or days, and that this individualised review often causes greater delay than AML screening does. That has a direct commercial implication: a treasurer choosing a corridor is not choosing a network, they are choosing a jurisdiction’s document handling. Asia-Pacific credits 84.6% of wholesale payments within a day and only 25.6% within an hour, a 59-point gap that is almost entirely the beneficiary leg. Nothing about that gap gets fixed by faster messaging.
What Does It Cost to Send Money Internationally?
The global average cost of sending USD 200 in remittances was 6.36% in the third quarter of 2025, down from 6.49% in the first quarter, according to the World Bank’s Remittance Prices Worldwide database. Banks charged 14.99%.
Remittance Cost by Provider Type and Channel, Q3 2025
| Measure | Cost (% of USD 200 sent) | Source |
|---|---|---|
| Banks | 14.99 | World Bank Remittance Prices Worldwide, Issue 54 |
| Non-digital services | 7.30 | World Bank Remittance Prices Worldwide, Issue 54 |
| Global average | 6.36 | World Bank Remittance Prices Worldwide, Issue 54 |
| Money transfer operators | 4.72 | World Bank Remittance Prices Worldwide, Issue 54 |
| Digital services | 4.59 | World Bank Remittance Prices Worldwide, Issue 54 |
| SmaRT average (three cheapest per corridor) | 3.29 | World Bank Remittance Prices Worldwide, Issue 54 |
| UN SDG 10.c target for 2030 | 3.00 | World Bank Remittance Prices Worldwide, Issue 54 |
According to Axis Intelligence Research, banks charge 2.36 times the global average and 4.56 times the SmaRT benchmark that an informed consumer could obtain. The global average sits 3.36 percentage points above the 2030 SDG target, which is to say it is 112% above where it is meant to land in four years.
Remittance Cost by Receiving Region, Q3 2025
| Receiving region | Average cost of sending USD 200 (%) | Source |
|---|---|---|
| Sub-Saharan Africa | 8.46 | World Bank Remittance Prices Worldwide, Issue 54 |
| Europe and Central Asia | 6.80 | World Bank Remittance Prices Worldwide, Issue 54 |
| East Asia and Pacific | 5.83 | World Bank Remittance Prices Worldwide, Issue 54 |
| Latin America and Caribbean | 5.64 | World Bank Remittance Prices Worldwide, Issue 54 |
| South Asia | 5.30 | World Bank Remittance Prices Worldwide, Issue 54 |
| Middle East, North Africa, Afghanistan and Pakistan | 5.11 | World Bank Remittance Prices Worldwide, Issue 54 |
Sarah Davis: A 14.99% bank cost on a 200-dollar transfer is not a fee schedule, it is a business telling you it does not want the account. Strip it apart and most of it is the FX margin rather than the stated charge, which is why the headline price comparison that most banks publish clears their compliance review without ever showing the customer what they pay. The FSB’s receiver-side survey adds another 0.1% to 1.3% that never appears on the sender’s screen at all, plus an FX margin of 0.7% to 1.1% where the receiving institution handles conversion. The gap between 14.99% and the 3.29% SmaRT average is the clearest measure in payments of what happens when a product is priced on inertia rather than on cost. Domestic rails tell a different story entirely, and our ACH payments statistics pillar shows what per-transaction economics look like when volume and competition both work.
Is SWIFT Losing Ground to Alternative Payment Networks?
CIPS, China’s cross-border renminbi clearing system, counted 210 direct participants and 1,619 indirect participants as of June 2026, according to CIPS Co., Ltd.. Swift connects more than 11,500 institutions.
SWIFT and CIPS Network Reach Compared
| Measure | Swift | CIPS | Source |
|---|---|---|---|
| Connected institutions or participants | 11,500+ | 1,829 | Swift Annual Review 2025; CIPS, June 2026 |
| Countries and territories | 200+ | Not disclosed on this basis | Swift Annual Review 2025; CIPS, June 2026 |
| Currency scope | Multi-currency | RMB and HKD | Swift Annual Review 2025; CIPS, June 2026 |
| Function | Messaging | Clearing and settlement | Swift Annual Review 2025; CIPS, June 2026 |
According to Axis Intelligence Research, Swift’s institutional reach runs 6.29 times CIPS participant count. That ratio compares two different units and is offered as a scale marker, not an equivalence: Swift counts connected institutions on a messaging network, CIPS counts direct and indirect participants in a clearing system, and an institution can be both.
CIPS Indirect Participants by Region, June 2026
| Region | Indirect participants | Share | Source |
|---|---|---|---|
| Asia | 1,157 | 71.5% | CIPS Co., Ltd., June 2026 |
| Europe | 267 | 16.5% | CIPS Co., Ltd., June 2026 |
| Africa | 103 | 6.4% | CIPS Co., Ltd., June 2026 |
| North America | 34 | 2.1% | CIPS Co., Ltd., June 2026 |
| South America | 33 | 2.0% | CIPS Co., Ltd., June 2026 |
| Oceania | 25 | 1.5% | CIPS Co., Ltd., June 2026 |
Of 1,157 Asian indirect participants, 542 sit in the Chinese mainland. According to Axis Intelligence Research, 28.5% of CIPS indirect participants are located outside Asia.
On the sanctions side, the Council of the European Union states that its transaction ban now applies to more than 100 Russian banks plus institutions in Belarus and third countries. The EU upgraded its messaging-service ban to a full transaction ban in July 2025, and has prohibited EU entities from connecting to Russia’s SPFS messaging system since June 2024.
Sarah Davis: The de-dollarisation trade gets discussed as though a share shift is underway, and the tracker does not support that reading yet. The renminbi held 3.10% of global Swift payment value in July 2026. What the CIPS participant map does show is a build-out phase: 267 European indirect participants and 103 African ones is not a rounding error, and Africa nearly doubled its count inside a year. A clearing system with 1,829 participants and a single-currency scope is not a Swift substitute, it is an RMB corridor with growing optionality attached. The question worth tracking is not whether CIPS displaces Swift, it is whether a bank that has already built the connection starts routing dollar-adjacent Asian trade through it when a sanctions decision makes the correspondent route expensive. Optionality gets used when the alternative gets priced.
How Complete Is the ISO 20022 Migration?
The coexistence period between MT and ISO 20022 messages for cross-border payments and reporting ended on 22 November 2025. By that point 97% of financial institution to financial institution payment instructions on Swift were already in ISO 20022 format, per the Swift Annual Review 2025.
The CPMI’s harmonised ISO 20022 data requirements carry a separate end-2027 implementation deadline, and the FATF’s revised Recommendation 16 on payment transparency is expected to be fully implemented by end-2030. Swift told the FSB that payment speed improves once both counterparties have completed migration.
Network Resilience Indicators, 2025
| Indicator | Value | Source |
|---|---|---|
| FIN availability | 99.998% | Swift Annual Review 2025 |
| SwiftNet availability | 99.998% | Swift Annual Review 2025 |
| Business continuity exercises completed | 580+ | Swift Annual Review 2025 |
| Customer Security Programme attestation coverage of traffic | 99%+ | Swift Annual Review 2025 |
| Active payment routes | 40,000+ | Swift Annual Review 2025 |
| Reachable accounts | 4 billion+ | Swift Annual Review 2025 |
| Market infrastructures connected | 235+ | Swift Annual Review 2025 |
| Digital wallets integrated | 1.7 billion | Swift Annual Review 2025 |
Sarah Davis: Ending coexistence at 97% adoption is the part of the ISO 20022 story worth noting, because migrations of this size usually finish at a long tail of holdouts and a deadline extension. The value is not in the format, it is in what structured data does to straight-through processing rates and to the false-positive rate on sanctions screening. Both of those feed straight into the beneficiary leg where the delays actually live. Whether that shows up in the 2026 KPI round is the thing to watch, and it is the one place where a standards project could plausibly move the speed numbers that five years of policy work have not.
Methodology
Collection. Every figure in this dataset was retrieved from a primary source document during the production session on 25 August 2026. Six source documents were used: the Swift Annual Review 2025; the Swift Global Currency Tracker, August 2026 edition, reporting July 2026 data; the Financial Stability Board’s G20 Roadmap for Enhancing Cross-border Payments Consolidated progress report for 2025, which uses Swift transaction data for its wholesale KPIs; the World Bank’s Remittance Prices Worldwide Issue 54; the Council of the European Union’s sanctions questions and answers page; and the CIPS Worldwide Participants disclosure. No figure in this article derives from secondary aggregation.
Swift data basis. The Global Currency Tracker reflects statistics computed from MT 103, MT 202, pacs.008 and pacs.009 messages for payments, MT 300 for foreign exchange, and MT 400 and MT 700 for trade finance. Swift states that because financial institutions have multiple means of exchanging transaction information, the Tracker does not represent complete market or industry statistics, and that shares can move for reasons unrelated to underlying currency usage, such as changes in message-type adoption. Currency shares are value-based. FIN figures in the Annual Review include InterAct payment messages resulting from the FIN to ISO 20022 migration, with historical figures adjusted on the same basis.
Formula: Cross-Border Flow Intensity (CFI).
CFI = (Regional Sent - Regional Intra-regional) / Regional Sent x 100
Inputs are the 2025 average daily FIN payment messaging volumes by region published in the Swift Annual Review 2025, expressed in thousands of messages per day. The network-level reading uses the summed totals: (23,485 – 16,681) / 23,485 x 100 = 28.97%.
Formula: Speed Dispersion Ratio (SDR).
SDR = Fastest region one-hour crediting share / Slowest region one-hour crediting share
Inputs are the FSB wholesale speed KPIs for the first quarter of 2025. Regions are the FSB’s own wholesale groupings, which differ from the regional groupings used for retail and remittance KPIs.
Formula: currency concentration reading.
HHI = Sum of (currency percentage share)^2 across ranked currencies
The global reading of 3,161 points covers the 20 ranked currencies representing 98.26% of payment value. The ex-eurozone reading of 3,803 points covers the 20 ranked currencies representing 98.64% of value. Currencies outside the published ranking are excluded rather than estimated.
Arithmetic. All derived figures were computed in Python from the source values and are reproducible from the accompanying CSV. Every calculated row carries a method note identifying its formula.
Disclosed source discrepancy. Swift’s regional payments table and its FIN market-share table are published on different bases. Total regional outbound traffic of 23,485 thousand messages per day does not reconcile with the 26,020 thousand implied by dividing 2025 payments volume of 6,573 million by 252.6 business days. Axis Intelligence Research publishes both series as reported and does not interpolate between them. The regional series balances internally, with outbound of 23,485 against inbound of 23,486, and that internal consistency is the basis for its use in the CFI calculation.
What this reading does not capture. Swift is a messaging network, not a settlement system, and message counts are not transaction values. Payments settled bilaterally, through regional systems, on correspondent books or over closed networks do not appear. The FSB wholesale KPIs cover payments of USD 100,000 or above and are drawn from a single quarter.
About This Dataset
This dataset contains 255 observations covering Swift network traffic, currency share, regional payment flows, cross-border payment speed and cost, message standards adoption, and comparative network reach, spanning 2020 to 2026 with global and regional coverage.
Licence: CC BY 4.0
Citation line: Axis Intelligence Research, SWIFT Payment Statistics 2026, 2026
How to Cite This Research
APA Axis Intelligence Research. (2026). SWIFT payment statistics 2026: Network volume, currency share, speed and cost data. Axis Intelligence. https://axis-intelligence.com/swift-payment-statistics/
MLA Axis Intelligence Research. “SWIFT Payment Statistics 2026: Network Volume, Currency Share, Speed and Cost Data.” Axis Intelligence, 2026, axis-intelligence.com/swift-payment-statistics/.
Chicago Axis Intelligence Research. “SWIFT Payment Statistics 2026: Network Volume, Currency Share, Speed and Cost Data.” Axis Intelligence. 2026. https://axis-intelligence.com/swift-payment-statistics/.
Frequently Asked Questions
Why do SWIFT currency share figures differ between the global and international measures?
The two series count different populations. The global payments measure includes payments made between institutions inside the eurozone; the international measure strips them out. Because intra-eurozone traffic is euro-denominated by construction, removing it lifts the dollar’s share by 8.59 percentage points and cuts the euro’s by 8.42. Any comparison across articles or across time needs to confirm which basis is in use, because the same month can produce a dollar share of either 50.99% or 59.58%.
Does SWIFT actually move money, or only messages?
Only messages. Swift transmits standardised payment instructions between institutions; the value moves across correspondent accounts, domestic real-time gross settlement systems and clearing houses. This is why Swift’s 40,000 active payment routes and 4 billion reachable accounts are reach measures rather than settlement measures, and why Swift traffic counts cannot be converted into transaction values.
Why does SWIFT report 75% of payments arriving in ten minutes when the FSB reports just over 50% credited within an hour?
The measurement windows differ. Swift’s figure ends when the instruction reaches the beneficiary bank, covering the in-flight leg. The FSB figure ends when the beneficiary’s account is credited, adding the beneficiary leg. Swift’s own September 2025 research found 80% of total elapsed transaction time occurs after arrival at the beneficiary bank, which accounts for most of the difference between the two.
What counts as a wholesale cross-border payment in the G20 monitoring framework?
The FSB defines wholesale cross-border payments as those valued at USD 100,000 or above. Retail is below that threshold excluding remittances, with separate reference amounts per use case: USD 20,000 for B2B, USD 5,000 for B2P, USD 1,000 for P2P and USD 100 for P2B. Speed KPIs exclude weekends and public holidays in the bank’s country of location.
Has the end of MT and ISO 20022 coexistence changed payment processing?
The coexistence period for cross-border payments and reporting closed on 22 November 2025 with 97% of institution-to-institution payment instructions already in ISO 20022. Swift has told the FSB that speed improves once both counterparties complete migration. A separate CPMI deadline for harmonised data requirements runs to end-2027, so the data-quality benefit is still arriving.
Why are bank remittance costs so much higher than money transfer operator costs?
Banks averaged 14.99% on a USD 200 transfer in the third quarter of 2025 against 4.72% for money transfer operators, a gap that reflects cost structure rather than a single fee. Bank pricing carries correspondent chain costs, FX margin taken at the conversion leg, and compliance overhead allocated across low-value transfers that were never designed for the rail. The FSB also documents receiver-side costs of 0.1% to 1.3% that do not appear in sender-side pricing at all.
How many Russian banks have been removed from SWIFT?
The Council of the European Union states that its transaction ban applies to more than 100 Russian banks, along with certain institutions in Belarus and non-EU countries. The EU converted its specialised financial messaging ban into a full transaction ban in July 2025, meaning no EU operator may transact with listed banks directly or indirectly. EU entities have also been barred from connecting to Russia’s SPFS messaging system since June 2024.
Is CIPS a replacement for SWIFT?
Not on current scope. CIPS clears and settles renminbi and Hong Kong dollars; Swift carries messages in any currency. CIPS reported 210 direct and 1,619 indirect participants in June 2026 against Swift’s 11,500-plus connected institutions, a reach gap of roughly 6.29 times on those unit definitions. The two also sit at different layers of the stack, which is why a participant in one is frequently a participant in the other.
Which region is slowest for receiving cross-border payments and why?
Africa recorded the lowest share of wholesale cross-border payments credited within one hour in the first quarter of 2025, at 24.2%, improving from 20.4% two years earlier. The FSB attributes regional lag to capital control processing, beneficiary and transaction-purpose verification, incomplete supporting documentation and manual processes at the beneficiary bank rather than to messaging network performance.
Is SWIFT traffic growth driven by payments or securities?
Securities. Securities messaging grew 14.7% in 2025 against 8.9% for payments, and compounded at 10.57% against 8.53% over the 2020 to 2025 period. Securities reached 51.5% of FIN traffic in 2025 while payments fell to 43.5%, down 2.1 percentage points from 2020 even though payments volume itself rose 50.5%.
Related research:
Cross-border payments statistics | Wire transfer statistics | ACH payments statistics | Same Day ACH statistics | ACH transfer limits | ACH return rate statistics