Wholesale settlement network comparison matrix — Fnality / Partior / JPM Kinexys / mBridge / Agorá / Mariana

Wiki route

This entry sits under fintech index as the six-way side-by-side comparison matrix for the most-cited wholesale settlement networks. It complements the per-network deep dives at Fnality fnPS, Partior, JPM Onyx / Kinexys, mBridge, and BIS Project Agorá. For the governance pattern see Multi-Megabank Consortium Governance; for the central-banking-unbundling context see Unbundling of central-banking functions: the 5 layers; for the interoperability story with stablecoins see Circle USDC and JPMorgan JPMD.

[!info] TL;DR Six wholesale-settlement networks anchor the 2026-05 public dataset: Fnality fnPS (BoE-licensed central-bank-reserve token, GBP live 2024-Q2, 20+ shareholder banks), Partior (4-bank Singapore-MAS-licensed consortium, USD/SGD/EUR/GBP/JPY live, $1.5T cumulative), JPM Kinexys (single-bank-owned $5B/day, $1.5T cumulative, JPM Coin + Liink + Onyx Digital Assets), mBridge (BIS multi-CBDC bridge — PBoC/HKMA/BoT/CBUAE/SAMA after BIS exit 2025-10), Project Agorá (7 G10 central banks + 40 commercial banks, design-phase since 2024-04), and Project Mariana (BIS multi-CBDC FX swap PoC with BdF/MAS/SNB, completed 2023). Read across the matrix the structural pattern: settlement asset (central-bank reserve vs commercial-bank deposit vs synthetic) is the binding regulatory variable, consortium membership shape (single-bank vs multi-bank vs central-bank-only) dictates governance speed, and only Fnality has 100% central-bank-reserve backing with full BoE approval as a systemic payment system.

Why a six-way comparison

Single-network narratives (“Fnality is the consortium model,” “Kinexys is the single-bank model”) capture different parts of the story but obscure the binding structural variable: what settlement asset backs each token. By placing all six networks side-by-side on settlement asset / consortium membership / currencies live / daily volume / regulatory status / central-bank backing / interoperability with stablecoins, the regulatory-arbitrage and competitive-positioning patterns become visible. The matrix also clarifies why Fnality is the only currently live system with full RTGS-equivalent finality (central-bank reserves), why Kinexys leads in volume despite single-bank constraint (decision speed without multi-shareholder coordination), and why mBridge survived the BIS withdrawal (the operational machinery was already running and the political coordination problem the BIS posed was already solved). For the broader institutional-stablecoin-deposit-token contrast see Institutional Market Stablecoins = Only Bank-Issued Deposit Tokens Are Structurally Viable.

Matrix A · Settlement asset and finality

Network	Settlement asset	Finality type	Equivalent to
Fnality fnPS	Central-bank reserves (omnibus account at BoE / Fed / ECB / BoJ; 1:1 reserve-backed tokens)	RTGS final settlement (legally final)	Tokenised central-bank reserves
Partior	Commercial-bank deposits (4 main shareholder bank balance sheets interlocking)	Near-RTGS (atomic PvP on commercial-deposit basis)	Multi-bank commercial-deposit network
JPM Coin (Kinexys)	JPM commercial-bank deposits (single bank balance sheet)	Internal RTGS within JPM customer base	Single-bank tokenised deposit
mBridge	Wholesale CBDCs of 5+ participating central banks (PBoC / HKMA / BoT / CBUAE / SAMA)	RTGS final settlement (atomic PvP)	Multi-currency wholesale CBDC bridge
Project Agorá	Mixed: wholesale CBDCs (M0 central-bank money) + tokenised commercial bank deposits (M1) on shared platform	Design phase; finality model not yet fully specified	Hybrid M0 + M1 interoperability layer
Project Mariana	Synthetic — wholesale CBDC + automated market maker for cross-currency swap	PoC completed 2023; not in production	FX swap experiment using wCBDC + AMM

The central distinction is M0 vs M1 vs synthetic: Fnality and mBridge use M0 central-bank money (tokenised reserves / wholesale CBDC); Partior and Kinexys use M1 commercial-bank deposits; Agorá explicitly experiments with both layers simultaneously; Mariana uses synthetic wCBDC + AMM as a research design. This is the single most consequential dimension because it determines (a) legal finality, (b) regulatory status, (c) whether the system can serve as final settlement vs intermediate settlement, and (d) whether the system competes with or complements existing RTGS rails.

Matrix B · Consortium membership

Network	Membership shape	Member count	Governance model
Fnality fnPS	Multi-bank consortium + infrastructure	20+ shareholders: Lloyds, Santander, UBS, BNY Mellon, Barclays, Goldman, MUFG, ING, Sumitomo Mitsui, State Street, Nasdaq Ventures, DTCC, Euroclear	Multi-shareholder governance; slow decision speed (£ fnPS took 9 years from USC concept to launch)
Partior	4 main shareholders + 30+ network members	JPMorgan + DBS + Temasek + Standard Chartered (2024)	Singapore-anchored; faster than Fnality but slower than Kinexys
JPM Kinexys	Single bank	JPM only	Fastest decision speed; 180+ institutional clients via Liink
mBridge	Central-bank consortium	5 active central banks post-BIS exit: PBoC + HKMA + BoT + CBUAE + SAMA	Multi-central-bank with PBoC operational lead post-2025-10 BIS exit
Project Agorá	G10 central banks + global commercial banks	7 G10 central banks (FRBNY, BoE, BdF, BoJ, BoK, Banxico, SNB) + 40+ commercial banks	BIS-coordinated; design phase
Project Mariana	3 central banks (PoC)	BdF + MAS + SNB	Time-limited research PoC (2022-2023)

Decision-speed scaling: Kinexys (1 bank) > Partior (4 banks) > mBridge (5 central banks) > Fnality (20+ banks) > Agorá (47+ institutions). The Fnality 9-year-to-launch vs Kinexys’ 4-year-to-$1.5T comparison is the canonical case for “multi-shareholder governance is slow.” But Fnality has central-bank-reserve backing that Kinexys structurally cannot match (a commercial bank cannot tokenise other banks’ reserves), making the comparison about scope of legitimacy rather than pure speed.

Matrix C · Currencies live (2026-05)

Network	Currencies live	Currencies in pipeline
Fnality fnPS	GBP (£ fnPS live 2024-Q2; first commercial tx Lloyds × Santander × UBS)	USD / EUR / JPY in regulator negotiation (Fed / ECB / BoJ); 2026-2027 expected
Partior	USD / SGD / EUR / GBP / JPY (5 live)	INR / IDR planned 2026-Q1
JPM Kinexys	USD / EUR / GBP (3 live); SGD planned 2026-Q2	SGD via partnership; long-term JPY
mBridge	CNY / HKD / THB / AED / SAR (5 currencies tied to participating central banks)	Brazil DREX (BCB) Q2 2026 H2 interoperability test
Project Agorá	Test currencies for design phase; no production currencies live	n/a (Phase 2 design ongoing)
Project Mariana	EUR / SGD / CHF (PoC)	n/a (PoC completed 2023)

Currency coverage scoreboard: Partior has the most production-live currencies (5), Fnality has the highest regulatory weight per currency (only one live, but it is BoE-approved central-bank-reserve-backed), mBridge has the largest non-USD coverage, Kinexys has the deepest single-currency penetration (USD). Agorá and Mariana are research/design programmes without production currency coverage.

Matrix D · Daily volume and cumulative scale

Network	Daily volume (2026-Q1)	Cumulative throughput	Equivalent to
Fnality fnPS	Growing post-launch; not yet at meaningful steady state	<£10B cumulative	Early-stage; first commercial year
Partior	Variable; sources suggest single-digit-billion per day	~$1.5T cumulative (since 2021-08 launch)	Comparable to Kinexys cumulative
JPM Kinexys	$5B+ per day average (~30-40% of USDC daily)	$1.5T+ cumulative (since 2024-11 rebrand)	Largest single-network volume
mBridge	Small but growing; specific 2026 daily volume not publicly disclosed	Estimated <$100B cumulative	Limited but politically important
Project Agorá	n/a (design phase)	n/a	n/a
Project Mariana	n/a (PoC completed)	n/a	n/a

Kinexys leads in volume despite single-bank constraint — because JPM can deploy without coordinating with competitor banks. Partior is comparable on cumulative but represents 4-bank shared activity. Fnality is early-stage with central-bank-reserve backing as its long-term moat. mBridge has lower volume but structurally different reach: it covers non-USD cross-border commodity / oil settlement that no other network has access to.

Matrix E · Regulatory status and central-bank backing

Network	Lead regulator	Central-bank approval status	Regulatory category
Fnality fnPS	Bank of England	BoE 2023-11 statement confirms £ fnPS as systemically important payment system	Licensed systemic payment system (first globally)
Partior	MAS Singapore	MAS-licensed wholesale payment network operator	Singapore-anchored licensed PSP
JPM Kinexys	OCC + Fed (via JPM bank charter)	JPM Coin = on-balance-sheet commercial deposit; legally part of JPM banking license	Tokenised commercial deposit within G-SIB charter
mBridge	PBoC + HKMA + participating central banks post-BIS exit	Central-bank-operated platform; no separate regulatory category	Multi-central-bank inter-CB platform
Project Agorá	BIS coordination + 7 G10 central banks	Research and design phase; no licensing yet	BIS Innovation Hub design phase
Project Mariana	BdF + MAS + SNB	Completed PoC	Research-only

The regulatory category split is binary: Fnality is the only system that is both production-live AND holds a dedicated systemic-payment-system license. Kinexys is technically a feature of JPM’s banking license rather than a separate licensed system. Partior is licensed as a PSP. mBridge operates as a multi-central-bank inter-CB platform without commercial regulatory framework. Agorá and Mariana are explicitly research / design.

Matrix F · Technology stack

Network	DLT platform	Consensus	Cross-chain interop
Fnality fnPS	Enterprise Ethereum (Quorum → migrated to Hyperledger Besu)	Permissioned BFT	SWIFT bridging (2024-Q4 partnership)
Partior	Hyperledger Fabric derivative	Permissioned BFT	Bilateral bridges with member banks’ internal systems
JPM Kinexys	Quorum (JPM-developed Ethereum fork; donated to ConsenSys)	Permissioned BFT	JPMD extension to public chains: Base (2025-11) → Canton (2026-01)
mBridge	Custom permissioned DLT	HotStuff BFT	None planned beyond participating CB networks
Project Agorá	Not finalised; design phase exploring multiple platforms	Not publicly specified in the cited design materials	Designed to be platform-agnostic
Project Mariana	Distributed ledger with AMM module	PoC ledger	n/a (research)

The technology stack pattern: all production systems use permissioned BFT consensus (no public-chain consensus); Quorum / Besu / Fabric dominate; Canton emerges in 2026 as JPM’s chosen privacy-and-composability target for JPMD; HotStuff BFT is unique to mBridge. The platform-agnosticism of Agorá is a deliberate design choice — BIS coordination across 7 G10 central banks cannot mandate a single stack.

Matrix G · Cross-border PvP / DvP capability

Network	PvP across currencies	DvP for securities	Atomic FX swap
Fnality fnPS	Cross-currency PvP via fnPS GBP × fnPS USD (planned 2026-2027 when $ fnPS live)	DvP modules in development	Designed as Herstatt-risk-killer
Partior	24×7 atomic FX swap (live 2024-Q3)	DvP for tokenised securities	Direct competitor to Fnality cross-currency PvP
JPM Kinexys	Cross-currency within Kinexys via Onyx Digital Assets	DvP for intraday repo + tokenised treasuries ($700B+ cumulative)	Within JPM customer base
mBridge	PvP atomic settlement between participating CBDCs	Securities DvP not primary scope	Across mBridge participants
Project Agorá	Cross-border PvP + DvP are explicit design goals for Phase 2	DvP design phase	Cross-currency atomic swap is design goal
Project Mariana	Cross-currency wCBDC swap via AMM (PoC result)	n/a	Mariana’s central PoC finding

Herstatt risk elimination is the explicit goal of all six networks — but only Partior (live), Fnality (regulator-approved, $-fnPS pending), and mBridge (live PvP across 5 central banks) can claim production capability. Kinexys’s cross-currency capability is internal-only; Agorá and Mariana are design / research. Project Mariana’s 2023 PoC validated that wCBDC + AMM could clear cross-currency FX atomically — a foundational result that influenced both Agorá design and mBridge architecture.

Matrix H · Interoperability with stablecoins / tokenized assets

Network	Stablecoin interop	Tokenised MMF / RWA interop
Fnality fnPS	Not direct; fnPS is final settlement asset and SC could redeem into fnPS via member bank	BUIDL as IM collateral in CME/DTCC derivative scenarios (2026-Q1 test)
Partior	Stablecoin-bridge concept (USDC) discussed but not live	Tokenised asset DvP design
JPM Kinexys / JPMD	JPMD is JPM’s commercial deposit tokenised onto Base + Canton public chains — direct SC-adjacent	BUIDL IM collateral + Onyx Digital Assets repo
mBridge	Not in scope; mBridge is wCBDC-only	Not in scope
Project Agorá	Explicit goal: stablecoin → bank deposit → wCBDC three-layer pass-through	Tokenised securities + RWA interop is design goal
Project Mariana	n/a	n/a

This is the dimension where Agorá’s strategic positioning becomes clearest: it explicitly aims to be the interoperability layer between stablecoins (private), tokenised deposits (commercial-bank M1), and wholesale CBDC (central-bank M0). If the three-circle MRA framework crystallises (US GENIUS + EU MiCA + HK / SG / JP cross-recognition), Agorá would become the natural settlement anchor where compliant stablecoins ultimately redeem. Fnality and Kinexys both have stablecoin-adjacent strategies but are not architected as multi-layer interoperability fabrics.

Matrix I · Strategic position relative to SWIFT

Network	Replaces SWIFT messaging?	Replaces SWIFT correspondent banking?	Coexists with SWIFT?
Fnality fnPS	No (uses SWIFT messaging via 2024-Q4 partnership)	Partial — settlement leg moves on Fnality	Yes — SWIFT × Fnality public collaboration
Partior	Partial (replaces correspondent-banking 2-3-day settlement)	Yes for member-bank corridors	Coexists for non-member corridors
JPM Kinexys / Liink	Liink replaces SWIFT MT messaging (1100+ banks)	Partial for JPM customer base	Direct competitor to SWIFT in some segments
mBridge	Yes — explicitly designed to bypass SWIFT for non-USD cross-border CBDC	Yes	Politically intended to coexist outside SWIFT
Project Agorá	Designed to interoperate with SWIFT messaging	Replaces correspondent-banking leg	Coexists — SWIFT is participant
Project Mariana	n/a	n/a	n/a

The SWIFT-relationship pattern: Fnality + Agorá explicitly coexist with SWIFT (Agorá includes SWIFT as participant; Fnality has a public collaboration); Partior + Kinexys partially replace SWIFT in specific corridors; mBridge explicitly bypasses SWIFT for non-USD CBDC corridors. The 2024-Q4 SWIFT × Fnality collaboration is the clearest signal that incumbent message-rail and tokenised-settlement-asset are not zero-sum — SWIFT provides cross-border messaging, Fnality provides on-ledger settlement, and the two layers connect.

Why these dimensions

The nine matrix axes (settlement asset, consortium membership, currencies live, daily volume, regulatory status, central-bank backing, interoperability with stablecoins, technology stack, SWIFT relationship) were chosen because they map to the gating variables in wholesale settlement adoption:

Settlement asset — determines legal finality and regulatory classification (M0 vs M1 vs synthetic).
Consortium membership — determines decision speed and governance complexity (single-bank vs multi-bank vs central-bank-only).
Currencies live — measures realised scope (5 currencies vs 1).
Daily volume — measures realised adoption (the only metric that matters post-launch).
Regulatory status — determines whether the system can serve as final settlement (Fnality’s BoE license is unique).
Central-bank backing — determines legitimacy ceiling.
Stablecoin interoperability — determines whether the system fits into the post-GENIUS three-circle MRA framework.
Technology stack — determines operational risk and cross-chain reach.
SWIFT relationship — determines competitive positioning vs the incumbent.

These nine dimensions together explain why Fnality is the only system that can claim “tokenised central-bank money” as a product category, why Kinexys leads in raw throughput despite single-bank constraint, why Partior occupies the Asian-corridor commercial-deposit middle ground, why mBridge survived BIS withdrawal (operational machinery already running), and why Agorá is positioned as the future interoperability layer rather than a competitor to any of the live systems. The matrix view exposes these structural patterns side-by-side.

Reading the matrix

Fnality is “tokenised central-bank money”; Kinexys is “tokenised commercial-bank deposit”; mBridge is “wholesale CBDC bridge”; Partior is “multi-bank deposit consortium”. These are four structurally different products that get casually grouped as “wholesale settlement” but have completely different regulatory classifications and finality properties.
Fnality’s 9-year-to-launch reflects multi-shareholder governance friction. Kinexys’s 4-year-to-$1.5T reflects single-bank decision speed. The architectural choice (consortium vs single-bank) drives the speed differential — but Fnality has central-bank-reserve backing that Kinexys structurally cannot match.
BIS withdrawal from mBridge (2025-10) was a political decision, not a technical one. The operational platform continued under PBoC operational lead; if anything, BIS withdrawal clarified the political alignment rather than disrupting operations. Agorá launch (2024-04) was BIS’s pivot toward a Western-led parallel.
Project Mariana’s 2023 PoC validated that wCBDC + AMM could clear cross-currency FX atomically — a foundational result that influenced both Agorá design and mBridge architecture. Many of Agorá’s Phase 2 design goals trace back to Mariana findings.
Stablecoin interoperability is the next-decade boundary condition. The three-circle MRA framework (GENIUS / MiCA / EPI) plus Agorá’s M0-M1-stablecoin three-layer model implies an architecture where compliant stablecoins ultimately redeem into wholesale CBDC via tokenised bank deposits. Fnality and Kinexys are positioned for this architecture; mBridge is not (wCBDC-only by design).
SWIFT × Fnality is the canonical “incumbent + tokenised” partnership (2024-Q4). It suggests SWIFT’s strategic positioning is to provide cross-border messaging on top of tokenised settlement assets rather than to replace its own correspondent-banking model directly.
The “consortium of 20” vs “single bank” vs “five central banks” governance comparison is the canonical case for the multi-megabank consortium governance tradeoff: speed vs scope of legitimacy.

Sources

Fnality International public materials: https://www.fnality.org/
Bank of England — statement on Fnality (2023-11): https://www.bankofengland.co.uk/news/2023/november/boe-statement-on-fnality
Partior public materials: https://www.partior.com/
MAS — Partior launch announcement (2021): https://www.mas.gov.sg/news/media-releases/2021/partior-launch
JPMorgan Kinexys public site: https://www.jpmorgan.com/kinexys
JPMorgan Onyx legacy page: https://www.jpmorgan.com/onyx
BIS Innovation Hub CBDC topic landing: https://www.bis.org/about/bisih/topics/cbdc.htm
BIS Project Agorá page: https://www.bis.org/about/bisih/topics/cbdc/agora.htm
BIS Working Paper No. 1116 — “The next-generation monetary system — a blueprint” (2024): https://www.bis.org/publ/work1116.htm
BIS Project Mariana — published research report (2023)
SWIFT × Fnality collaboration announcement (2024-Q4): https://www.swift.com/news-events/news/swift-fnality-collaboration