What the Ordinance covers
The Ordinance applies to issuers of fiat-referenced stablecoins, defined as cryptographically secured digital representations of value that purport to maintain a stable value by reference to one or more fiat currencies. The core perimeter is "fiat-referenced" rather than "single-fiat-referenced." That choice matters: it is broader than the MAS SCS scope, which deliberately limits to single G10 currencies, and broader than the GENIUS Act framework in the US, which restricts permitted payment stablecoin issuers to USD-denominated instruments. Hong Kong has carved space for designs that reference HKD, offshore CNH, and combinations or derivatives thereof, although the operational rules tighten as the design strays further from a pure single-currency reserve.
What the Ordinance does not cover is also load-bearing. Tokenised deposits issued by licensed banks sit outside the Ordinance, supervised under existing banking law and through Project Ensemble for any cross-bank infrastructure. Tokenised money-market funds and tokenised SFC-authorised funds sit under SFC HK supervision and are explicitly fund interests, not stablecoins. Algorithmic constructions are foreclosed in practice by the reserve and redemption rules; the Ordinance does not formally ban them, but no design that fails to hold high-quality liquid asset reserves at par could meet the licensing bar.
Licensing perimeter and applicant population
A fiat-referenced stablecoin issuer operating in or from Hong Kong must hold an HKMA licence. The licence is its own perimeter, distinct from a banking licence, a stored-value-facility licence, or any of the existing payment licences. Banks can apply, but the licence itself is non-bank in shape. That design choice mirrors MiCA's separation between credit-institution and electronic-money-institution routes for EMT issuance, although the Hong Kong frame is less prescriptive about the institutional form the licensee must take.
The applicant queue, as visible in public communications through late April 2026, splits into three contender clusters:
- Bank-led contenders: licensed bank groups that have signalled intent to issue an HKD or offshore-CNH-referenced stablecoin out of a Hong Kong subsidiary or branch. The commercial logic is integration with existing tokenised deposit and trade-finance flows, where the stablecoin functions as a non-bank-money complement to bank money on the same operator's distribution rails.
- Payment-firm-led contenders: groups already holding stored-value-facility or money-services licences in Hong Kong who see the Ordinance as a route to a higher-grade settlement instrument than the existing SVF perimeter allows. This cluster is where the cross-border remittance use case sits most cleanly.
- Technology-firm-led contenders: large platform groups, often with parent companies headquartered in mainland China or Southeast Asia, applying via Hong Kong subsidiaries. Their interest is partly the offshore-RMB corridor and partly positioning a regulated stablecoin alongside their existing wallet and merchant-acquiring infrastructure.
HKMA awarded the first two issuer licences on 10 April 2026 to The Hongkong and Shanghai Banking Corporation Limited (the wholly-owned HK subsidiary of HSBC Holdings plc) and Anchorpoint Financial Limited (HSBC HK statement on the licence award). The two licences were drawn from a queue reported at 36 applicants by industry coverage. HSBC has signalled an HKD-denominated stablecoin launch in H2 2026, integrated into the bank's existing PayMe and HSBC HK App distribution rails, with three named initial use cases: peer-to-peer payments via PayMe, peer-to-merchant payments via PayMe, and tokenised investment subscriptions via the HSBC HK App. Anchorpoint is structured as a joint venture between Standard Chartered (Hong Kong), HKT, and Animoca, with the licence awarded to the JV vehicle rather than to the parent banks individually. Beyond these two named licensees, HKMA has not published a complete register, and operators tracking the regime should treat any further "licensee count" reported in secondary press as provisional until HKMA publishes its own register.
The technology-firm-led cluster is the thinnest of the three in practice, and the reason is traceable to a specific event rather than a lack of interest. HKMA named JINGDONG Coinlink Technology (Hong Kong) Limited and RD InnoTech Limited among the participants in its first stablecoin issuer sandbox cohort in July 2024, alongside the entities that later combined into the Anchorpoint joint venture (Wu Blockchain, 10 Apr 2026). Ant International had separately said it intended to apply once the formal application window opened. By October 2025, both Ant Group and JD.com had paused their Hong Kong stablecoin plans after Chinese regulators, the People's Bank of China and the Cyberspace Administration of China, reportedly instructed them to hold off; PBoC officials were cited as concerned that private-sector currency-like issuance from major platform groups could compete with the digital yuan and dilute the state's monopoly on money issuance (Fintech News Hong Kong, 21 Oct 2025). JD.com has denied exiting the space entirely; Jingdong Coinlink's sandbox participation and its stablecoin ambitions more broadly remain publicly live, but neither the sandbox cohort's status nor a fresh application from either group has been confirmed in HKMA's own communications since the pause was reported. RD InnoTech's application status is likewise not confirmed in HKMA-published material. Editorial note for review: whether Beijing's stance reads as a durable structural constraint on mainland-parented applicants specifically, versus a pause pending clearer guidance, is an interpretive call worth making deliberately rather than by default; flagged here rather than asserted in the prose above.
Reserve composition rules
Reserves must be composed of high-quality liquid assets at least equal to the par value of the stablecoin in circulation. The HQLA definition tracks the Basel framing closely: cash, short-dated sovereign debt, central-bank reserves, and similar instruments, with haircuts applied to less liquid components. The aggregate reserve must be at least equal to par at all times, which means an issuer cannot run a structurally undercollateralised float and rely on top-up flows to close the gap.
Segregation is mandatory. Reserves must be held separately from the issuer's general assets, in custody arrangements designed to insulate them from the issuer's insolvency. The practical effect is that a holder's claim on the reserve survives an issuer-level default, which is the foundational difference between a regulated stablecoin and a typical e-money or SVF instrument under the older perimeter.
Reserve disclosure is monthly at minimum, in line with the floor that has emerged across MiCA, GENIUS, and the major APAC regimes. Some Hong Kong applicants have signalled intent to publish more granular on-chain reserve attestations alongside the regulatory filings, which is a credible read of where institutional treasury counterparties have pushed the issuer market over the past two years (see Stablecoin types section 2.1.1 on attestation cadence).
The interest-payment question is settled the same way it is under MiCA and GENIUS: licensed issuers cannot pay interest to holders tied to the holding period of the stablecoin. The deliberate design choice across all three regimes is to keep regulated stablecoins from competing with bank deposits or money-market funds on yield, and to keep the perimeter narrow enough that the instrument behaves like programmable cash rather than a near-deposit substitute.
Redemption right at par
Holders have a statutory right to redeem at par, in the reference currency, within a defined period. The redemption obligation runs to direct holders, not to every bearer of the on-chain instrument. In practice this means a Hong Kong-licensed issuer can structure direct mint and redeem through a permissioned set of distributors and counterparties, while the on-chain transferability is broader. The shape is similar to the USDC distribution model, where a small set of direct-redemption counterparties carry the issuer's primary obligation and the secondary on-chain market clears against those counterparties.
The redemption right is what distinguishes the Hong Kong regime from the MAS SCS framework most cleanly. Both regimes require redemption at par, but the MAS framework is structurally bank-grade in its capital and reserve obligations and channels issuance through Major Payment Institution licensees, while the Hong Kong route is a standalone licence built specifically for the instrument. An operator picking between the two on jurisdiction often picks Hong Kong if the design references offshore RMB or anything other than a single G10 currency, and Singapore if the design is pure SGD or pure USD with a regional anchor.
The offshore-RMB angle
The offshore-renminbi piece is the part that distinguishes Hong Kong's framework most sharply from any non-APAC regime. Offshore CNH has cleared through Hong Kong for over a decade as the principal offshore-RMB centre, and the Stablecoins Ordinance perimeter accommodates a CNH-referenced design without requiring it to fit inside the onshore RMB framework, which is closed to private stablecoin issuance.
The design space splits into two routes. The first is a CNH-referenced stablecoin issued by a Hong Kong-licensed entity, with reserves held in offshore-CNH instruments. The second is an HKD-referenced stablecoin used as a settlement asset in flows that ultimately price in offshore RMB, with the FX leg handled by the holder rather than embedded in the instrument. The first route is more interesting for cross-border trade finance and payment-versus-payment use cases; the second is closer to how the HKD stablecoin contenders are most likely to ship in their first generation.
What the Ordinance does not do is connect to onshore RMB. Any cross-border use case touching a mainland counterparty has to clear through the existing settlement infrastructure, which is unchanged by the Ordinance. The Ordinance is a Hong Kong instrument with offshore-RMB optionality, not a bridge into the onshore market.
Comparison with MAS SCS, MiCA, and GENIUS
The four major regimes diverge sharply on scope, issuer form, and reserve rigour. The cleanest way to read them side by side:
Scope. MAS SCS limits to single G10 currencies. MiCA Title IV (EMT) limits to single fiat currencies. GENIUS limits in practice to USD payment stablecoins. Hong Kong allows fiat-referenced designs broadly, including HKD, offshore CNH, and combinations in principle.
Issuer form. MAS SCS requires Major Payment Institution licensing with bank-grade overlay. MiCA EMT requires credit-institution or electronic-money-institution authorisation. GENIUS requires permitted payment stablecoin issuer status, available to bank subsidiaries, federally qualified non-banks, or state-qualified issuers below the $10bn threshold. Hong Kong requires a standalone HKMA stablecoin licence, with banks eligible to apply but the licence itself non-bank in shape.
Reserve rigour. All four regimes require segregated, high-quality liquid asset reserves at par. MAS SCS layers bank-grade capital requirements on top. MiCA EMT requires reserves sufficient to meet redemption obligations and tightens further for significant EMTs. GENIUS specifies cash, short-dated Treasuries, repo, and similar narrow instruments. Hong Kong specifies HQLA at least equal to par with monthly attestation.
Interest. All four prohibit interest payments to holders tied to holding period.
Redemption. All four require redemption at par in the reference currency. MAS SCS and MiCA EMT specify the timeline more tightly; Hong Kong and GENIUS leave more operational flexibility.
For a product team, the four regimes are converging on the substantive shape (HQLA at par, segregated, redeemable, no interest), and diverging on the licensing route and the scope. Picking a jurisdiction is mostly a question of which licensing route fits the operator and whether the design needs the broader fiat-reference scope that Hong Kong offers.
Open questions
- Beyond the 10 April 2026 HSBC and Anchorpoint awards, HKMA has not published a complete register of additional issued or pending licences. The remaining 34-applicant queue's composition is partially reported in trade press but not in regulator-published form.
- The reserve-composition disclosure cadence and per-issuer attestation format for the HSBC and Anchorpoint launches. Both licensees have committed to the segregated-HQLA-at-par requirement, but specific attestation publishing arrangements and the ledger or platform-level disclosure for circulation data are not in current public coverage.
- The treatment of cross-licensee interoperability inside Project Ensemble is not yet resolved. Specifically, whether the HSBC HKD stablecoin (when it launches in H2 2026) can settle alongside tokenised deposits inside the EnsembleTX tiered-ledger architecture, or whether the stablecoin layer sits parallel to EnsembleTX rather than inside it. HSBC is a named EnsembleTX participant on the tokenised-deposit side, which makes the stablecoin-vs-tokenised-deposit substitution question inside HSBC's product range structurally interesting.
- The CNH-referenced stablecoin question is more politically loaded than the HKD route, and the public material is light on how HKMA intends to supervise a CNH instrument relative to the People's Bank of China's posture on offshore RMB. Anchorpoint's HKT and Animoca affiliations could plausibly support a CNH track, but neither the JV nor HKMA has named CNH as in-scope for the initial Anchorpoint product.
- Whether the second wave of licence awards under the Ordinance includes additional bank-led applicants (Bank of China HK, ICBC HK) or additional payment-firm-led applicants. The technology-firm-led cluster's near-term prospects look thinner following Ant Group and JD.com's reported October 2025 pause; whether either resumes a Hong Kong application, and whether JD Coinlink's continued sandbox participation converts into a formal licence application independent of its parent's mainland posture, are both unresolved in HKMA-published material as of this writing.