The thesis is that Singapore, Hong Kong, the UAE, and increasingly Japan and Korea are developing institutional digital-asset infrastructure as an interconnected cluster. Stablecoin licensing, wholesale CBDC, regulated venues, custody, and tokenised-asset issuance sit in adjacent jurisdictions with overlapping bank participants and shared regulatory grammar. The US, despite hosting the largest single-bank tokenisation programme (Kinexys) and the deepest stablecoin issuer set, runs without a wholesale CBDC, without Federal Reserve account access for non-bank issuers, and across a fragmented map of state and federal supervisors. For an institutional allocator routing cross-border flow through Asia, the cluster proposition matters operationally: the same banks (DBS, Standard Chartered, HSBC, MUFG, SMBC) appear repeatedly across MAS Project Guardian, HKMA Project Ensemble, and the BIS-led Project Agorá, with the regulators themselves coordinating more visibly than the EU, US, or UK supervisors do.
What the cluster actually is
Three observations make the cluster claim concrete rather than rhetorical. First, regulator-to-regulator coordination is dense. MAS, HKMA, and the Bank of Thailand have run named bilateral wholesale-payment work; HKMA's Project Ensemble sandbox includes overseas counterparties; MAS's Project Guardian partner roster includes Japanese and European banks alongside the Singapore three. The cluster is not a closed regional bloc, but the standing infrastructure (sandboxes, common workstreams, shared partner sets) is denser inside APAC than across the Atlantic.
Second, the bank participant set overlaps. Standard Chartered, DBS, HSBC, MUFG, SMBC, OCBC, and UOB recur across the major APAC tokenisation venues. The same legal entities show up in the Singapore guardian workstreams, the Hong Kong ensemble sandbox, the Japan Progmat consortium, and the BIS-coordinated wholesale CBDC trials. By contrast, US bank participation in tokenisation is concentrated at JPMorgan with Citi, BNY, State Street, and Goldman as significant but more siloed entrants. The cluster effect is partly a function of fewer regional banks generating more participation density.
Third, regulated-venue density is higher. Hong Kong has SFC-licensed virtual-asset trading platforms, an HKMA-licensed stablecoin perimeter under Cap. 656, an active wholesale CBDC sandbox, and at least three live tokenised-bond programmes. Singapore has the MAS SCS framework, the DTSP regime, the Project Guardian rails, and DDEx as a bank-operated regulated venue. The US has no wholesale CBDC, no live deposit-token interbank rail comparable to Partior, and a state-by-state money-transmission regime that any platform without an OCC trust charter has to navigate. The federal infrastructure gap is real.
What the US has that the cluster does not
The cluster claim should not be overstated. JPMorgan alone runs more deposit-token volume on Kinexys than every APAC bank combined; the Kinexys Digital Payments rail has cleared trillions in cumulative notional with $5B+ daily flow, and JPMD on Base is the only public-chain native deposit token in production from a GSIB. BlackRock's BUIDL, anchored by Securitize in the US perimeter, is the single largest tokenised MMF (money-market fund) in the world. Apollo's ACRED on Plume, Janus Henderson's JAAA on Centrifuge, Franklin Templeton's BENJI: the named US institutional issuance set is deeper than any APAC bank's.
The cluster question is not who issues the most tokenised product. It is whether a corporate treasurer running a multi-currency cross-border programme (SGD, HKD, JPY, USD with MAS, HKMA, FSA, and OCC counterparties) finds the integrated APAC stack easier to operate against than the US-anchored alternative. As of 2026, the answer is increasingly yes for SGD-HKD-JPY corridors, and increasingly no for USD payments and USD-denominated tokenised securities, where US infrastructure dominates.
Where the cluster could break
Three failure modes for the thesis. First, regulator divergence. The cluster works because MAS, HKMA, and the Japanese FSA largely agree on the substance of stablecoin licensing, custody segregation, and bank-led tokenisation. If the Singapore DTSP regime expands extraterritorially in ways HKMA does not match, or the Hong Kong Stablecoins Ordinance develops a renminbi corridor that diverges from MAS's G10-only SCS perimeter, the regulatory grammar shifts and the cluster loses coherence.
Second, US infrastructure consolidation. If the US gets a second-bank entrant on tokenised deposits at scale (State Street, BNY, or Citi) within 12 months, and that entrant chooses public-chain deployment, the structural advantage the cluster currently has on multi-bank tokenised cash erodes. The arbitrage window the GENIUS Act has opened (interest on tokenised deposits, no interest on stablecoins) is a US-specific accelerant.
Third, China posture. The cluster does not include the mainland, and the offshore renminbi corridor is the most contested cross-border lane. If People's Bank of China extends mBridge-adjacent rails into HKMA or MAS-licensed venues with mainland banks as full participants, the cluster's geometry changes substantially. If it does not, Hong Kong becomes the gateway by default.
Operating implications
For an institutional tokenisation operator deciding where to hub, the cluster lens reframes the choice. Singapore plus Hong Kong with optional Tokyo or Seoul gives access to multiple regulated venues, a deeper bank counterparty set, and explicit cross-border infrastructure (Project Agorá, mBridge-graduated successors, Partior). A US-only hub gives access to the deepest USD liquidity and the largest tokenised-MMF venues but requires bilateral relationships across each state and federal supervisor.
The agentic-commerce layer is downstream of the same question. Whichever cluster lands its regulatory grammar first for AI-agent-controlled wallets holding tokenised money sets the structural default. Singapore and Hong Kong are publicly engaged on the question; the US is not.
Open questions
- Whether any APAC platform vendor reaches global-tier scale in tokenisation infrastructure. As of 2026 the institutional-provider tier is dominated by US firms (Securitize, Zerohash, Digital Asset); the cluster has the regulatory and bank density without a comparable platform provider.
- Whether a Korean or Japanese bank joins the multi-bank tokenised-deposit settlement layer, extending Partior beyond its current SGD-USD axis.
- Whether the US infrastructure gap closes through OCC trust-charter holders (Anchorage, Circle, Ripple, Paxos) rather than through GSIBs, in which case the cluster's relative advantage narrows differently.