Multi-Network Messaging Switch
For decades, the operating assumption for cross-currency and high-value payments was simple: one network, one connection, one queue. SWIFTNet carried it all.
That world is changing fast. Major RTGS operators are moving toward parallel network channels -- the Eurosystem's TARGET2 has long offered access via both SWIFT and SIA-Colt, the Bank of England has stated its RTGS renewal aims at a "message network-agnostic" design, and the HKMA has now gone further by mandating prescriptive usage quotas across two channels (see below). Instant payment schemes such as FedNow, The Clearing House RTP, UK Faster Payments, Australia's NPP and SEPA Instant Credit Transfer already run on their own rails. ISO 20022 is the connective tissue across all of them.
The result: institutions that used to run a single SWIFT pipe now need to operate two, three or more network channels in parallel -- often carrying the same ISO 20022 message standard, addressed to the same counterparties, settling in the same currencies, but going out over different wires under different rulebooks.
Bolting a second connector onto a legacy SWIFT stack creates duplicated archives, fractured audit trails, inconsistent validation, and routing logic spread across scripts. The right answer is a network-agnostic switch at the centre of the payment stack -- one engine that knows how to talk to every network and decides, per message, which one to use.
Network Abstraction
Configure each network as an independent channel -- SWIFTNet, domestic IP networks such as ICLNet or CIPS, instant payment schemes, internal rails. Back-office systems do not need to know which one will be used.
Policy-Based Routing
Visual rule builder routes per message: by amount, currency, BIC, counterparty, value date, business unit, time of day, or any payload field. Rules are configuration, not code.
ISO 20022 Native
Same pacs.008 / pacs.009 / camt.* model across every channel. Apply network-specific usage guidelines (CBPR+, HVPS+, regional variants) on the way out without rewriting the business payload.
Quota Enforcement
Distribute traffic across channels to meet regulator-mandated usage ratios -- monthly volume splits, minimum days per channel, per-CHATS quotas. The Hub keeps running counters and steers routing to stay in band.
Contingency Switching
Calendar-driven "single-channel days" exercise each rail end-to-end. Operator-triggered or automatic failover moves part or all of the traffic to the alternate network within seconds.
Format Translation
The same engine that powers our ISO 20022 migration solution converts MT↔MX on the fly when one network speaks MT and another speaks MX, or when usage guidelines diverge.
Unified Audit & Archive
Every message -- regardless of which network it took -- is stored in a single full-text searchable archive with a single audit trail. Reconcile against any monthly network-operator report from one place.
Unified Validation
Apply the right rulebook per network at the point of routing: CBPR+ for SWIFTNet cross-border, HVPS+ for T2/CHAPS/Fedwire, scheme-specific rules for instant payments. Reject before sending, not after.
Operational Visibility
Real-time dashboards show traffic split, queue depth, and per-network status. Operators see at a glance whether the institution is on-track against monthly quotas and contingency requirements.
The mandate
On 20 May 2026 the Hong Kong Monetary Authority issued the Real Time Gross Settlement Dual Network Channel Usage Guideline (Ref B5/12C), conveyed to CHATS members via HKICL Circular 2026/154.
From the DNC service launch (expected end-2026), every Authorised Institution must be able to connect to HKD, USD, EUR and RMB CHATS via ICLNet -- HKICL's proprietary IP network -- in addition to the existing SWIFTNet channel. Both networks operate in parallel.
Each AI must maintain dual connectivity at all times and be able to switch part or all of its RTGS traffic between channels swiftly, without causing material delay or disruption.
What you have to prove monthly (from 1 April 2027)
- 20%–80% of outgoing Inter-bank Fund Transfers via ICLNet, per CHATS the AI participates in.
- Receive payments via ICLNet on 5 to 15 working days per month, per CHATS.
- ≥2 working days/month solely on ICLNet and ≥2 working days/month solely on SWIFTNet for contingency.
- Reconcile against the monthly HKICL DNC monitoring report.
From the May 2026 announcement, CHATS members have about seven months to be ready for parallel operation at DNC launch and roughly ten months to be inside the usage quotas. Institutions evaluating their target architecture should look at the network layer first.
Discuss your DNC readiness plan →Cross-border + instant
Route low-value, time-sensitive flows to instant schemes (SEPA Instant, FedNow, The Clearing House RTP, UK Faster Payments, NPP) and reserve SWIFTNet for higher-value or non-supported currencies -- per-payment, automatically.
CIPS + SWIFTNet for RMB
CIPS direct participants can choose between CIPS's own messaging and SWIFTNet to reach the system. The Hub lets you split CNY traffic between the two based on counterparty reachability and cost, without changing back-office integration.
Architecture-ready for new mandates
The HKMA RTGS DNC is among the most prescriptive dual-network mandates issued by a major regulator to date. The Hub gives institutions an architecture that is in place when -- or if -- their regulator follows a similar template.
Internal vs. external
Route on-us payments through an internal book-transfer rail and keep external traffic on the regulated networks, with the same workflow and audit on both sides.
Cost-aware routing
Use the cheapest reachable network per corridor at every moment. Policy rules can take pricing tables, value date, and currency into account.
Migration windows
Run new and legacy networks in parallel during cutovers, with a controlled volume ramp on the target network. The Hub keeps a single view across both.
One integration, many networks
Back-office systems integrate once with the Hub. New networks become a configuration exercise, not a project.
Regulatory readiness
Quota, contingency and audit requirements are first-class capabilities -- not custom scripts grafted onto a single-rail platform.
Lower operational risk
One archive, one audit trail, one validation engine, one team. Network switches stop being a release event.