---
title: "Cross-Border Stablecoin Feasibility Over Time"
subtitle: "A Trigger-Conditioned Forward Map of Pending Regulatory Transitions"
author: "Yunjie Fan · Independent Researcher, Digital Asset Regulation"
date: "June 2026 (Version 0.1.0)"
documentclass: article
fontsize: 10pt
geometry: margin=1in
mainfont: "Noto Serif CJK SC"
colorlinks: true
linkcolorblue: true
toc: true
toc-depth: 3
---

> **Revision v0.1.0 (June 2026; register integration updated August 2026).** First pre-release working-paper draft. The original regulatory reading is a June 2026 snapshot and is not silently restated as current. It defers to the *Multi-Jurisdiction Stablecoin Compliance Matrix* (v0.9.7) for version-stamped jurisdictional facts, to the *Cross-Border Stablecoin Architecture* working paper (v0.2.8) for the eight-constraint vocabulary and six interaction sets, and to the *Cross-Border Digital-Finance Corridor Atlas* (v0.2.5) for the directed-edge model. The forward events, signal table, dated transitions and explicit freshness warnings are integrated with the Cross-Border Stablecoin Register (CBSR, v0.11.0). Every forward-looking item carries a conditional status flag; the paper conditions feasibility and does not predict outcomes. **On references:** the positioning in §1.4 engages adjacent methodological traditions; the formal citations in the References section carry their DOI or stable identifier.

> **Corrections applied at register v0.10.0 (July 2026).** This draft has been corrected against the register's own data on four points that the pre-v0.10.0 corridor artifact had wrong, all recorded in `docs/PAPER_ERRATA.md` and traceable to `analysis/signal_table.json`. **E1 / F-US-1 / E8:** the United States is a *dated* regime-in-transition destination, not a live Category II one — the GENIUS Act §18 inbound gate is enacted but not commenced, effective on the earlier of 18 January 2027 or final rules plus one hundred and twenty days (§20), so there are **two** dated horizons and the United States one is the *earlier*. The eight token-holding origins into the United States therefore read Category T. **NF-2:** the §4.2 corridor-sensitivity ordering is restated with three regime-in-transition destinations, ordered by the earliness and the kind of their dates. **E2 / F-EU:** the MiCA Article 58(3) means-of-exchange ceiling is conjunctive ("and"), not disjunctive. **NF-1:** the class rule is destination-first — a destination prohibition or pre-regime absence dominates an origin-side issuance gap — which is what the Corridor Atlas v0.2.5 §3.4/§4.1 states and what `tools/class_rule.py:class_of` now applies to all one hundred and thirty-two directed edges. Each correction is enforced by a build gate rather than asserted: `run_invariants.py` DC5 re-derives every class from the published rule on every build, and DC7 proves no class rests on a market fact.


**WORKING PAPER**

*Disclosure.* Independent research. No compensation (direct, contingent, or otherwise) has been received for this paper, the underlying research, or any companion work in the portfolio to which it belongs. No advisory, consulting, employment, or other compensated relationship exists with any party having a financial interest in the regulatory outcomes discussed. Public-source synthesis throughout. This paper describes no non-public commercial transaction or arrangement.

ORCID 0009-0005-6762-084X · SSRN Author ID 11463068

## Abstract

The companion works in this portfolio characterize cross-border stablecoin feasibility as it stands today: the Compliance Matrix as a twelve-jurisdiction node table, the Architecture working paper as an eight-constraint composition framework, and the Corridor Atlas as a one-hundred-thirty-two-edge directed map. Each is a description of a single moment. This paper adds the missing dimension: feasibility is dated, and the public regulatory record already contains the changes in law that will move it. Its central instrument is the *trigger*, a change in law such as a commencement or an enactment, never a market event, and its central method is conditioning rather than forecasting: for each pending change the paper states which directed edges change feasibility class, which records the change rests on, and under what conditional status the change is contingent. Two contributions follow. The first is the conditioning method itself, applied to a deliberately sparse current instance that will densify as enactments land. The second is a typology of triggers by the kind of certainty each carries, separating the one kind that moves a dated horizon from the kinds that do not, and including a category a class-centric reading easily misses: the *intra-regime gating change* that moves accessibility within a live regime, with the European Union's 1 July 2026 MiCA transitional expiry as its worked example.

From the register of changes in law visible on the public record as of 30 June 2026, set out in full in §3, the paper derives a corridor-sensitivity ordering. The forward map's class-level motion is concentrated at the one remaining pre-regime origin, South Korea, and at the three regime-in-transition destinations: the United States, whose GENIUS Act §18 inbound gate is enacted but not commenced and resolves to a channelled Category II on an outer cap of 18 January 2027; the United Kingdom, resolving to a clean Category I on a gazetted 25 October 2027; and Taiwan, whose Virtual Asset Service Act passed at third reading on the dateline of this draft and which resolves to a channelled Category II on an as-yet-ungazetted commencement. The prohibition destination, Mainland China, stays insensitive to any positive trigger. A second finding the register's structure makes legible is a cross-jurisdictional one: independent regimes converging on a single functional line between prohibited holder-yield and permitted activity-linked rewards, documented most fully for the United States and appearing as a sibling restriction elsewhere.

A final and more interpretive section, resting on softer evidence as it says plainly, reads a structure the static map cannot express: the cross-border stablecoin stack is divided along one geopolitical fracture at three layers at once. The settlement substrate is split between rival bloc experiments rather than shared, with only Project Agorá led by the Bank for International Settlements, Project Ensemble led by the Hong Kong Monetary Authority, and mBridge run since the Bank for International Settlements' 2024 withdrawal by a People's-Republic-of-China-and-Gulf group; the regulatory layer is open, channelled, or closed at each boundary by the availability of a recognition determination; and the issuance layer is fragmented per instrument. The paper advocates no outcome: the transitions it maps are analytical, and their operational viability depends on developments enumerated against the constraint framework, not asserted.

*Keywords:* stablecoins; cross-border payments; payment corridors; digital asset regulation; regulatory transitions; monetary sovereignty; wholesale settlement; Project Agorá; mBridge; Project Ensemble; comparative financial regulation. *JEL Classification:* G28; K22; K23; F33; G15; E42; F38; O38.

## Methodology Note

This paper is the product of independent research and draws exclusively on public sources: enacted statutes, published or proposed regulations, agency releases and orders, central-bank publications, public corporate disclosures, and published academic and practitioner literature. Where a forward development is referenced, a conditional regulatory-status flag is preserved at the relevant point to make the dependency explicit. The paper proposes no architectural arrangement specific to any party and describes no non-public transaction. The author welcomes correction; readers identifying citation errors, factual misstatements, or instances in which a forward-looking claim has not been properly qualified are invited to flag the issue via the SSRN abstract page or the register's issue tracker.

A single discipline governs the entire paper and is stated once, here, because everything below inherits it. The unit that this paper allows to move feasibility is a *change in law*: a statute that commences, a bill that is enacted, an instrument that takes effect. A product launching, an issuer registering, a banking rail going live, a licence cohort being announced: these are facts about the market, not propositions of law, and they are never treated here as the thing that flips a corridor. This is not a stylistic preference. It is the same orthogonal-axes distinction the register enforces between a legal claim and an operational one, carried into the time dimension. When the paper says an edge "flips at" a date or "would flip if" an enactment occurred, the date or the enactment is always a `tier1_legal` event, and the transition it triggers is always a transition in the binding status of an instrument, not in the state of the market.

## §1. Introduction

### 1.1 The static map and its limit

The three companion works in this portfolio answer, between them, the question of what can be operated today across the twelve jurisdictions they cover. The Compliance Matrix answers it at the level of the node: for each jurisdiction and each of fifteen dimensions, what the binding instrument requires, permits, or prohibits. The Architecture working paper answers it at the level of composition: which pairs of constraints, jointly binding, exclude architectures that either constraint alone would admit, and through which of six interaction sets that joint binding operates. The Corridor Atlas answers it at the level of the directed edge: for value moving from an origin to a destination, what authorization or recognition the destination's inbound gate requires, which administrator gates it, and into which of six composite feasibility classes the edge falls. Each answer is precise, sourced, and, this being the limit, instantaneous. It is a photograph of a regulatory landscape that is, on the public record itself, visibly in motion.

The motion is not speculative. As of late June 2026 the United Kingdom has *made* but not *commenced* its comprehensive regime, with a gazetted commencement day in the future. The United States has enacted the issuer layer of its framework and reported, but not passed, the market-structure layer that would complete it. Taiwan and South Korea have each fixed a regulatory posture and carried a comprehensive bill to a known procedural stage short of enactment. Singapore has finalized a framework as policy and announced that it is ready to begin the implementing legislation. The European Banking Authority has an in-force cap whose operative methodology it has not yet finalized. Hong Kong has issued a first licence cohort and not yet a second. In each case the change is not a forecast about what some actor might do; it is a development already on the public record, at a known stage, whose resolution will move one or more cells of the static map from one value to another. A description of feasibility that cannot represent these pending movements is, for an operator planning across a multi-year horizon or a supervisor anticipating the supervisory perimeter of a regime not yet operative, incomplete in a way that matters.

### 1.2 The contribution: conditioning, not forecasting

This paper adds the time dimension to the static map. The discipline it works under is conditioning: for each pending change in law, it states precisely which directed edges change feasibility class when the change occurs, which records the change rests on, and, preserved at every step, the conditional status under which the change is itself contingent. It does not forecast which bills will pass, when contingent enactments will occur, or how an open methodology will be finalized; that is left to the reader. A reader learns not "the United States will complete its framework in 2026" but "if the CLARITY Act is enacted substantially as reported, the following United States cells change binding status and the following intermediary-layer overlays become operative; as of late June 2026 that enactment is contingent, on the Senate Legislative Calendar but not passed, with the practical floor window running to the August recess." The conditional structure, the *if-then* with its dependency made explicit, is the contribution.

This conditioning rests on machinery the register already provides and which this paper renders into prose. The register's `compose(origin, destination, as_of)` function applies every scheduled change in law effective by a given date and re-runs the Atlas's own feasibility algorithm, so that the same edge can read one class at today's date and another once a regime is operative; its event calendar holds the dated and contingent changes in law, each annotated with the records it moves and the signal-field transitions it causes; and its computed layer derives feasibility from a per-jurisdiction signal table grounded only in propositions of law. The paper does not introduce a new model. It reads the forward structure the register encodes, ranges the pending triggers by the kind of certainty they carry, and draws out the corridor-level and infrastructure-level consequences a tabular engine does not narrate.

### 1.3 What this paper does, and what it does not do

The paper's spine is the forward map, and it is what the title names: a *trigger register* (§§2–4) built under the conditioning discipline, comprising the definition of a trigger, the pending and recently-effective changes in law visible on the public record, the feasibility-class transitions each would cause, and the corridor-sensitivity ordering that locates where on the directed map the pending movement is concentrated. On that spine the register makes two further readings legible, offered as secondary to the forward map, not as co-equal contributions. The first is a *cross-jurisdictional convergence* (§4.5): read across jurisdictions instead of along edges, the register surfaces independent regimes settling on a common functional line around stablecoin yield. The second is a *structural analysis the static map cannot express* (§5): the open–closed boundary in the wholesale settlement layer, where the directed map's Digital-Financial archetype and its infrastructure-overlap attribute become the entry point to a reading of how the cross-border stablecoin stack is splitting along geopolitical bloc lines, into rival settlement substrates beneath a competitive issuance layer, and of where the regulatory layer is standardizing toward interoperability versus remaining walled. Both secondary readings rest on evidence the paper grades explicitly and, in §5's case, openly weaker than the forward map's, and neither is asserted as a headline result.

It also observes four boundaries, stated in full in §6: it advocates no regulatory outcome, predicts the resolution of no contingent development, describes no non-public commercial arrangement, and asserts none of its derived transitions as authoritative, the computed layer beneath them being a labelled preview whose disagreements with the authored analysis are surfaced as findings, in the manner the companion methodology paper sets out.

One framing should be made explicit at the outset, because it determines how the paper should be read and resolves a potential ambiguity about what is being claimed. The contribution is the *method*, the discipline of conditioning feasibility on changes in law and the typology of triggers it organizes (§3), together with its *current, deliberately sparse instantiation*; it does not claim a rich, already-populated forward map. By the paper's own discipline, **two** pending transitions presently carry future dates, and the earlier of them is the United States, not the United Kingdom. The United States GENIUS Act framework is enacted but not commenced: §20 makes its operative provisions effective on the earlier of eighteen months after enactment (18 January 2027) or one hundred and twenty days after the primary federal payment stablecoin regulators issue final regulations, an *outer cap* that may fall sooner, so its §18 inbound comparability gate is not yet available. The later is the United Kingdom systemic-stablecoin regime, with commencement *gazetted* for 25 October 2027. Every other trigger is contingent or enacted-but-undated, and moves no horizon until a date is fixed. [Corrected in register v0.10.0; see errata E1 / F-US-1 and backlog VB-06. The earlier draft recorded a single dated horizon because it coded the §18 gate as an in-force Category II gate that no jurisdiction had yet passed through — letting the *market* observation that no comparability determination has been granted do a *proposition of law's* work, which is exactly the claim-class error the companion methodology paper exists to prevent.] That sparseness is not a defect to be hidden but a direct consequence of taking the conditioning discipline seriously: a forward map honest enough to refuse invented dates will, at any given moment, be thin, and will *densify* only as the pending enactments land and their transitions become datable. The durable contribution is therefore the method and its typology, with the present register of triggers as the method's first, sparse application, one the same machinery will thicken over successive releases without redesign. The contribution is the method, not the map.

### 1.4 Relation to adjacent methods (positioning)

The method developed here sits near several established traditions without being any of them, and locating it among them both clarifies the contribution and forestalls a natural misreading. To a *regulatory and policy* readership the nearest neighbour is **horizon-scanning**, the structured anticipation of regulatory developments that supervisors and standard-setters practise. This paper differs in that it does not survey the horizon impressionistically but conditions a specific, machine-derived feasibility model on a closed set of developments, each tied to a record and to a binding-status transition. To a *strategy and foresight* readership the nearest neighbour is **scenario analysis**; the difference is that scenario analysis typically constructs a small number of narrative futures, whereas this paper refuses to assign likelihoods or to build narratives and instead exposes, for each pending change in law taken one at a time, exactly which directed edges it would move. This is the distinction that most needs making, because without it "event-conditioned, register-derived forward feasibility" reads as scenario analysis under another name; with it, the difference is precise: no narratives, no probabilities, one if-then per change in law, each grounded in a record.

To an *empirical-finance* readership the method rhymes with the **event-study** tradition, which measures the effect of a discrete event on a quantity of interest. The analogy is real but the object differs: an event study estimates a market reaction to an event that *has* occurred, whereas this paper states a structural reclassification that *would* follow a change in law that has *not* occurred, and the "event" is always a change in law (a `tier1_legal` commencement or enactment), never a market move. And to a readership in **legislative forecasting and prediction markets**, the paper is adjacent in that it *reports* prediction-market pricing for one contingent enactment (§3.2) as a fact about the public record, but it is emphatically not itself a forecasting exercise: it conditions on the enactment as an if-branch and takes no position on whether or when the market's implied probability will be realized. The prediction-market figure is data about the world's current expectation; the paper's own claims are the conditional consequences, not the expectation.

Located among these, the contribution is narrow and specific: an *event-conditioned* forward analysis whose events are changes in law only, whose conditioning is derived from a machine-readable register, not from expert narrative, whose discipline is to refuse probabilities and invented dates, and whose output is a typed set of if-then transitions over a directed feasibility graph. That particular combination, law-only triggers, register-derived, conditioning-not-forecasting, organized by an explicit trigger typology, is what the adjacent traditions, each strong in its own object, do not jointly provide. The same located-contribution serves both readerships this paper addresses: to *law-and-technology and financial-regulation* readers it offers a disciplined forward reading of a fast-moving regulatory frontier; to *regulatory-foresight and computational* readers it offers a reproducible, register-grounded alternative to narrative anticipation.

## §2. Method: forward feasibility as conditioning

### 2.1 The trigger, defined

A *trigger* is a change in law that moves at least one per-jurisdiction signal on which a corridor's feasibility class depends. The signals are those the register's composition engine reads, namely a jurisdiction's regime status (pre-regime, in transition, live, or prohibition), the type of its inbound gate (open, channelled, comparability-gated, or closed), whether it has an exportable, comprehensively authorizable token, and whether it imposes an egress or export override, each annotated with the node record that justifies it. A trigger is therefore not any regulatory news; it is specifically an event that changes one of those signal values, and through it the feasibility class of every edge the signal touches.

Three properties follow from the definition and are enforced wherever the paper applies it. A trigger is always a `tier1_legal` event: it is a commencement, an enactment, or an instrument taking effect, and never a market launch, a registration, or a cohort announcement. A trigger is distinguished from a change in *feasibility class*: some changes in law refine a rule without moving any signal (Japan's now-in-force trust-type reserve relaxation alters no inbound gate, because Japan is already live), and such an event belongs in the calendar as a control case, precisely to show that the calendar distinguishes the two. And a trigger is either *scheduled* or *contingent*: a scheduled trigger has an effective date and is applied by the engine whenever the as-of date reaches it; a contingent trigger depends on an enactment that has not happened, carries no firm date, and is *never* applied by date, surfaced only as a pending transition that would change the verdict if the enactment occurred. This last property is the structural guarantee that the paper conditions rather than forecasts: a contingent trigger moves no horizon until a date is fixed.

### 2.2 From a trigger to an edge transition

A directed edge's feasibility class is read at the destination's inbound gate, with a drag contributed by the origin where the origin imposes an export or egress restriction. A trigger therefore propagates to edges in one of two ways. A destination-side trigger, such as a regime commencing or an inbound gate opening or being created, changes the class of every edge *terminating* in that jurisdiction. An origin-side trigger, a pre-regime origin acquiring for the first time an exportable authorizable token, changes the class of every edge *originating* in that jurisdiction, typically lifting those edges out of the composition-unresolved class into which the absence of an exportable token had placed them. A small number of triggers do both at once for a single jurisdiction, because a jurisdiction that moves from pre-regime to live simultaneously acquires an exportable token (changing its outbound edges) and an operative inbound gate (changing its inbound edges).

The paper records each trigger's propagation in those terms: the signal it moves, the direction (inbound, outbound, or both) in which edges change, and the before-and-after classes. Where a trigger's target signal value is itself provisional pending the enacted text, for instance whether a pre-regime jurisdiction's eventual inbound gate will be open or channelled, the paper marks the target as provisional and does not resolve the post-trigger class beyond what the draft text supports.

### 2.3 The feasibility classes, in brief

For reference, the six composite classes the Corridor Atlas defines, which the transitions below move between, are: Category I (dual authorization available: each end separately authorizable, no equivalence step required in the direction of flow); Category II (equivalence, recognition, or channel determination required, where entry turns on a determination not generically granted, such as the United States comparability gate, the Japanese recognition-and-registration channel, the United Arab Emirates onshore usage-channel registration, or the Brazilian counterparty-conditional foreign-exchange channel); Category III (composition problem unresolved, where the origin lacks an exportable authorizable token or the two regimes interact unresolvably, reducing the corridor to partnership or coordination options); regime-in-transition at destination (a comprehensive regime adopted but not yet operative); blocked at destination (an issuance prohibition); and pre-regime at destination (no authorizable issuance pathway yet). An origin imposing an export or egress restriction overlays an origin-override flag on any of these.

## §3. The trigger register

This part sets out the changes in law visible on the public record as of 30 June 2026, together with the control cases, in descending order of the certainty attached to each change. Each entry states the trigger, its binding-status transition, the records it rests on, the edge transitions it would cause, and its conditional flag. The transitions are conditional throughout: a scheduled trigger's transition is applied at its date and not before; a contingent trigger's transition is surfaced as an if-enacted branch and applied by no date.

The ordering is not merely a list but a *typology of regulatory triggers by the kind of certainty each carries*, and the typology is itself a contribution of the paper, a second one alongside the conditioning method, and one that should travel to any forward analysis of a regulated domain in transition. Six kinds recur, and the subsections below are organized by them. A **fully-scheduled** trigger (§3.1, the United Kingdom) has a gazetted commencement date and is applied by that date; it is the only kind that moves a horizon. An **enacted-but-not-commenced** trigger (§3.3, Taiwan) has passed into law but awaits subsidiary legislation with no gazetted commencement date, so it has partly fired, since enactment is real, yet operationalizes no signal until commencement. It is the regime-in-transition category, and Taiwan became its newest instance on the dateline of this draft. A **contingent, no-date** trigger (§3.2 and §3.4, the United States and South Korea) depends on an enactment that has not occurred, carries no date, and is surfaced only as an if-enacted branch; it is the most common kind, and the reason an honest forward map is sparse. A **contingent change that is not a class change** (§3.5, Singapore) is a pending enactment whose effect is to harden a binding status without moving any feasibility class, because the baseline that determines the class is already in force. A **parameterization of an in-force rule** (§3.7, the European Banking Authority) is the operationalization of a constraint that already binds, which sharpens the operative content of a class without creating or moving one. An **intra-regime gating change** (§3.10, the European Union MiCA transitional expiry) is a dated change in law that alters accessibility *within* a live regime, the eligibility to operate or to be admitted, without moving any feasibility class; it is the category a class-centric model most easily overlooks, and §3.10 and §6 treat it explicitly. An **open compositional question** (§3.8, the second Hong Kong cohort) is information, not a change in law, whose resolution feeds another jurisdiction's analysis. And a **resolved-tightening, open-relaxation** posture (§3.9, the People's Republic of China) is the asymmetric case in which a boundary's recent movement is settled in one direction while the developments that could move it back remain open. Two further entries earn their place because they are *not* ordinary class-flipping triggers and would distort a naive forward map if treated as such: the Japanese reserve relaxation (§3.6), a **dated-empty-effect** change that is dated and now in force with an explicitly empty feasibility effect, paired on the same date with an **inbound-recognition** change (§3.6) that opens an inbound equivalence channel handled by conditioning rather than by asserting a class change; and the People's-Republic-of-China posture (§3.9), where the tightening is in force and only a relaxation, bearing on the typology of responses and not on the blocked class, remains available. A forward analysis lacking this typology would tend to treat every regulatory development as a pending class flip; the typology's work is to separate the one kind that moves a dated horizon from the kinds that do not, and to keep in view the intra-regime gating changes that move accessibility without moving class.

### 3.1 Scheduled: the United Kingdom regime commences (25 October 2027)

The United Kingdom is the one pending transition with a gazetted date. The Financial Services and Markets Act 2000 (Cryptoassets) Regulations 2026, SI 2026/102, were made on 4 February 2026, with preparatory provisions from 25 February 2026 and *full commencement set for 25 October 2027*, at which point the Bank of England systemic-stablecoin regime and the Financial Conduct Authority conduct regime become operative. Until that day the United Kingdom is, in the directed model, a regime-in-transition destination: the rule exists on the statute book, but an inbound corridor cannot be completed as a fully authorized pathway because the regime is not yet operative.

The transition is therefore clean and dated. Every edge *terminating* in the United Kingdom whose origin holds an exportable authorizable token reads Category T (regime adopted but not yet operative) at any as-of date before 25 October 2027 and Category I at any as-of date on or after it; the register's `compose(as_of)` flips the United States→United Kingdom edge from T to I precisely at that day. At the same instant, the regime-in-transition caveat that the register attaches to United Kingdom pairs resolves in full: the eight caveated pairs collapse to zero. The binding status of the underlying United Kingdom cells moves from `made_not_commenced` to `in_force_enacted`, which is why those cells, though their official text has already been read, are held below the citable-as-current-law tier until the commencement date, since the instrument must be in force, not merely located, to be cited as binding law.

One subtlety must be preserved, because it distinguishes a transition that resolves by date from a modelling difference that does not. After 25 October 2027 the *directed* edges into the United Kingdom are cleanly Category I, but two *undirected* pairs (European Union–United Kingdom and United Kingdom–United States) remain flagged as findings in the register's computed layer even past 2027. That residual is structural, not temporal: the Atlas scores a live-to-live relationship in the relevant configuration as a hybrid class, while the Architecture paper's pairwise analysis authored the same pairs as cleanly bridgeable. The commencement date retires the regime-in-transition artifact; it does not retire the structural difference, which is a genuine modelling finding the register exists to surface, not to paper over. Edges whose United Kingdom-bound origin has no exportable token, namely Mainland China, Taiwan, and South Korea, do not participate in this transition at all; the origin drag holds them in Category III regardless of the destination's commencement.

*Conditional flag.* The date is gazetted (`precision: day`); the transition is scheduled and applied by date. What remains genuinely open is interpretive rather than temporal: the three-limb territorial perimeter of the regime, on which the Financial Conduct Authority's summer 2026 policy statement is expected. That openness affects which operations fall inside the perimeter, not the date on which the perimeter becomes operative.

### 3.2 Contingent: the United States completes its market-structure layer

The United States entry has two components that the earlier draft ran together, and separating them is the substance of errata F-US-1 / E8. The *first* is a **scheduled-with-cap** trigger, not a contingent one, and it belongs with §3.1 rather than here: the GENIUS Act is enacted (18 July 2025) but **not commenced**. Its §20 sets the effective date at the earlier of eighteen months after enactment (18 January 2027) or one hundred and twenty days after the primary federal payment stablecoin regulators issue final regulations; at the 30 June 2026 snapshot only *proposed* rules had issued (the OCC notice published 2 March 2026, comments closed 1 May 2026; the FDIC and the joint FinCEN/OFAC notices of 10 April 2026; the Treasury §4(c) notice of April 2026). The §18 foreign-issuer comparability gate is therefore **not yet available**, and the eight edges from token-holding origins into the United States read Category T today, resolving to a channelled Category II on that outer cap. Saying instead that "no comparability determination has been granted to date" states a *market* observation where a *proposition of law* is required — the gate is not in force — and it is precisely the claim-class conflation the companion methodology paper's first axis exists to prevent. The *second* component, which is genuinely contingent and is what the rest of this section analyses, is the market-structure layer.

The United States has enacted the issuer layer of its framework and reported, but not passed, the layer that would complete it. The GENIUS Act governs the issuer level, including the §4(a)(11) prohibition on issuer-paid yield to holders "solely in connection with the holding, use, or retention" of a payment stablecoin, which bites from the §20 effective date. The Digital Asset Market Clarity Act (H.R. 3633) is the pending market-structure layer: the Senate Banking Committee advanced its substitute on a fifteen-to-nine vote on 14 May 2026, the bill was reported to the Senate on 1 June 2026 and placed on the Senate Legislative Calendar as Calendar No. 423, and it still requires Senate floor passage, reconciliation with the Senate Agriculture and House versions, and Presidential signature.

The trigger is the enactment of CLARITY, and its propagation is to the United States intermediary and routing layer rather than to a wholesale change in the United States inbound gate. On enactment, the United States bank-and-non-bank-routing cell moves from `pending_proposal` to `in_force_enacted`, and the §404 intermediary-yield prohibition and the market-structure overlay (carried today as flagged-pending overlays on the United States yield, securities, and issuer cells) become operative. The most consequential downstream effect is on the boundary of the three-layer routing architecture the Architecture working paper develops: the §404 prohibition extends the issuer-level yield ban to covered intermediaries, and the question of whether a routing arrangement that converts a non-yielding payment stablecoin into shares of a yield-bearing registered investment company falls inside or outside that prohibition turns on the operative, functional test the architecture sets out, namely the direction of the conversion, disclosure of the distinct ownership position, timing independence, yield attribution, and audit trail. The Senate Banking substitute carries the Tillis–Alsobrooks compromise on stablecoin yield, under which yield for merely holding a balance is prohibited while stablecoin rewards and activity-linked incentives are allowed; this supports, but does not yet enact, the position that user-initiated, activity-linked routing falls outside the holding prohibition.

*Conditional flag.* No firm enactment date. As of late June 2026, 2026 passage is roughly even (prediction-market pricing near forty-eight percent in late June, down from about seventy-four percent a month earlier), with the practical window requiring Senate floor passage before the August recess. The trigger is modelled as contingent and moves no as-of horizon; the joint Notice of Proposed Rulemaking that §404 (as reported) directs the Securities and Exchange Commission, the Commodity Futures Trading Commission, and the Treasury to issue within twelve months of enactment is a second-order contingency on top of the first, and its content, which will set the operative boundary of the routing test, is itself an open question (§3.7).

### 3.3 Enacted, not commenced: Taiwan's regime awaits subsidiary legislation

Taiwan moved, on the very day of this draft's dateline, from a pre-regime destination to a regime-in-transition one, and the change is instructive enough to record precisely. Until 30 June 2026 only an anti-money-laundering registration regime was in force, under the amended Money Laundering Control Act and the associated registration regulations operative since 30 November 2024, while the dedicated Virtual Asset Service Act that would create stablecoin-issuance licensing remained a bill that had completed its article-by-article first review on 3 June 2026. On 30 June 2026 the Legislative Yuan passed that bill at third reading, enacting the Act. Enactment is not commencement. The Financial Services Commission has indicated that the regime depends on nine items of subsidiary legislation, which it expects to complete at the earliest in the first quarter of 2027, with operators able to re-apply for licences within one year of the law taking effect, that is, by the end of the first quarter of 2028. Taiwan now occupies the posture the United Kingdom occupies in §3.1: an instrument passed and on the statute book but not yet operative, with the operationalization of any signal awaiting commencement, not enactment.

The structural consequence has the same form as the United Kingdom's but a different content, and the difference is itself new information. Enactment has occurred yet operationalizes no signal: Taiwan's regime status does not move to live, no exportable authorizable token yet exists, and the inbound gate is not yet operative, exactly as the analogous United Kingdom cells remain held until their commencement date. What separates Taiwan from the United Kingdom is the *kind* of inbound gate the enacted text describes. The Act conditions issuance and the inbound treatment of offshore stablecoins on Financial Services Commission approval, with a transition of roughly twenty-one months for operators. When Taiwan commences, its inbound gate will therefore be approval-gated, which places inbound edges in Category II, a channelled and determination-dependent interface, rather than the clean Category I the United Kingdom's inbound edges reach on commencement. Taiwan is thus a regime-in-transition destination whose eventual gate type is channelled, not open. The contrast with the United Kingdom is a substantive finding, not a mere analogy: two jurisdictions in the same lifecycle stage will, on commencement, resolve to different feasibility classes because their enacted inbound gates differ in kind.

On the origin side nothing changes, and the typology stays self-consistent. Taiwan still has no exportable authorizable token, because enactment alone does not create one; that awaits commencement and the actual issuance of licences. Every edge originating in Taiwan therefore remains in Category III under the origin drag, exactly as §3.1 already records for origins without an exportable token, and that earlier statement needs no revision. It was correct before enactment and remains correct after it, which is a small confirmation that the origin-versus-destination treatment is carved at the right joint.

This entry is also the clearest illustration in the paper of why a forward map must be a dated, register-backed living document. The typology defines a regime-in-transition destination *a priori*, as a structural category; Taiwan filled that category in real time, on the afternoon of the snapshot date, when a paper written that morning still read "remains a bill." A static description published on 30 June 2026 was already stale by that evening on this one corridor. Only a register with dated, versioned snapshots can absorb that kind of movement without silently going out of date, which is the methodological point §6 returns to.

*Conditional flag.* Enactment occurred at third reading on 30 June 2026; commencement is pending nine items of subsidiary legislation, with the Financial Services Commission's policy timetable pointing to the first quarter of 2027 and operator re-application by the end of the first quarter of 2028. No commencement date is yet gazetted, so the transition is applied by no date. The post-commencement inbound class is provisionally Category II (channelled) on the strength of the enacted approval requirement, pending the subsidiary legislation that will fix its operative content.

### 3.4 Contingent: South Korea enacts an issuance regime

South Korea is pre-regime in the same structural sense as Taiwan, at an earlier procedural stage and with an unresolved design dispute at its centre. The Virtual Asset User Protection Act has been in force since 19 July 2024, but it is a user-protection and market-abuse statute, not an issuance-licensing framework, and domestic won-stablecoin issuance is effectively prohibited. The Digital Asset Basic Act that would create an issuance regime remained an un-enacted bill as of late June 2026, in the National Assembly Political Affairs Committee subcommittee, with passage pushed past the June local elections and committee reconstitution into the second half of 2026. The defining open question is one of monetary sovereignty and issuer eligibility jointly: the Bank of Korea seeks to require bank-led consortia holding at least fifty-one percent, together with a binding multi-agency consent body, while the Financial Services Commission favours a more fintech-inclusive model, and the leading draft would classify cross-border won stablecoins as a foreign-exchange means of payment.

The trigger is enactment and commencement of the Act, and South Korea is now the register's one remaining pre-regime origin. On enactment and commencement it would move from pre-regime to live, acquire an exportable token, and gain an inbound gate, so that inbound edges leave the pre-regime class and outbound edges leave Category III. This is the both-directions propagation Taiwan exhibited until its own enactment on 30 June 2026 moved it into the enacted-but-not-commenced category; South Korea now holds that pre-regime, both-directions position alone. It carries one element of its own that Taiwan's enacted text handles differently: a central-bank-approval and bank-consortium sub-mechanism within issuer eligibility that becomes operative only on enactment, which the Architecture paper records as a named conditional sub-mechanism because it could recur in other central-bank-led jurisdictions.

*Conditional flag.* No firm enactment date, and procedurally less advanced than Taiwan (still in subcommittee). The fifty-one-percent issuer-eligibility dispute is unresolved and bears directly on what the enacted inbound gate and issuer pathway will permit. The trigger is contingent and applied by no date.

### 3.5 Contingent, but not a class change: Singapore's implementing legislation

Singapore illustrates a category the register is careful to keep distinct: a change in law that is genuinely pending but does *not* change the jurisdiction's feasibility class, because the baseline that determines the class is already in force. The Payment Services Act 2019 digital-payment-token baseline is operative. The Single-Currency Stablecoin framework, which restricts the "MAS-regulated" label to Singapore-dollar or G10-pegged tokens issued from Singapore-incorporated entities and forbids such issuers from lending or staking, was finalized as policy in August 2023, and the Monetary Authority of Singapore announced at the November 2025 Singapore FinTech Festival that it was ready to begin drafting the implementing legislation, with the framework expected to take effect in mid-2026.

The trigger is the enactment of that implementing legislation. Its propagation is to *binding status*, not to the inbound gate: the eight Single-Currency-Stablecoin-specific requirement cells move from `finalized_policy_pending` (and a transitional status) to `in_force_enacted`, and the MAS-regulated-stablecoin label moves from finalized policy to operative. Crucially, it does not move Singapore's feasibility class, because the directed model already reads Singapore through its in-force baseline; downgrading the Single-Currency-Stablecoin-specific cells to draft did not flip Singapore into the pre-regime set, precisely because issuer eligibility rests on the in-force Payment Services Act regime. The trigger therefore belongs in the register as the second control case alongside Japan: a real, dated-by-policy change in law whose effect on feasibility class is none.

*Conditional flag.* No firm enactment date; the Monetary Authority's guidance points to a mid-2026 effective date and drafting was announced ready as of November 2025. The trigger is contingent and applied by no date; its effect, when it lands, is to harden the binding status of the Single-Currency-Stablecoin cells, not to change a corridor class.

### 3.6 Two same-day changes: Japan's reserve relaxation (control case) and its foreign-stablecoin recognition

Japan supplies both the clearest demonstration that the trigger register distinguishes a change in law from a change in feasibility *and*, on the same date, an example of a change in law that does bear on feasibility, which makes it a more interesting entry than a pure control case. Japan operates a live Electronic Payment Instruments regime under the Payment Services Act, in force since 1 June 2023, with a closed issuer trichotomy of bank, registered funds-transfer provider, and trust company or trust bank, and a channelled inbound route through a registered Electronic Payment Instruments Exchange Service Provider, reserves held in Japan. Act No. 66 of 2025, promulgated on 13 June 2025, took effect on 1 June 2026 and is now in force; it relaxed the backing rule for trust-type issuers, permitting up to fifty percent of issuance value in low-risk short-term assets rather than wholly in demand deposits.

The reserve relaxation is the control case. It is a `tier1_legal` change, now in force, recorded with an explicitly empty feasibility effect: Japan is already live, the relaxation refines reserve composition for one issuer mode, and it changes no inbound gate, so no edge changes class. The event sits in the calendar precisely to make the distinction visible, because the register tracks changes in law and a change in law that moves no signal is recorded as moving nothing. An analysis that read "a stablecoin rule changed in Japan" as "a Japanese corridor changed" would be wrong, and the control case keeps that error out of the forward map.

What the earlier draft missed is that 1 June 2026 carried a *second* change with a different character. Alongside the reserve relaxation, the framework that took effect set out the requirements for foreign-issued stablecoins to circulate within Japan, and a Cabinet Office Ordinance of the same date recognised foreign trust-type stablecoins judged equivalent to the Japanese regime as electronic payment instruments. That is an inbound-side development, not a domestic reserve adjustment, and the claim in the earlier draft that the Japanese changes "change no inbound gate" was therefore too strong. The honest statement is narrower: the reserve relaxation changes no inbound gate, but the same-day equivalence recognition opens a recognition pathway for qualifying foreign stablecoins, and its effect on the class of inbound-to-Japan edges depends on the recognition criteria and how they are applied. The paper handles it by conditioning rather than by asserting a class change: a recognised foreign trust-type stablecoin would reach Japan through the recognition-and-registration channel the directed model already scores as Category II, and the operative reach of that channel turns on the equivalence determination, which is the open variable. This entry also gives §5.3 a dated, concrete instance of recognition functioning as the regulatory layer's interoperability mechanism, to which that section refers.

*Conditional flag.* Both changes are in force as of 1 June 2026. The reserve relaxation has no feasibility-class effect (control case). The equivalence recognition is an inbound recognition pathway whose effect on inbound-to-Japan edge class depends on the recognition criteria; it is handled by conditioning, and no class change is asserted on its account beyond the existing Category II channelled treatment.

### 3.7 Open parameterization: the European Banking Authority operationalizes the Article 58(3) cap

Not every pending development is a binary enactment. Some are the operationalization of an in-force rule, where the rule is settled but the parameters that determine its bite are not. The European Union's Article 58(3) means-of-exchange cap, which extends the Article 22, 23, and 24(3) thresholds *mutatis mutandis* to electronic-money tokens denominated in a currency that is not an official currency of a Member State, at one million daily transactions **and** two hundred million euros in daily transaction volume measured on a quarterly average — both thresholds, the statutory conjunction of Article 23(1) applied *mutatis mutandis* (errata E2; an "or" cap would bite earlier than the statute does), and applies only to use as a means of exchange within a single currency area, is in force. What is not finalized is the European Banking Authority's operative methodology and the Member State competent-authority practice that will determine how the cap is measured and enforced.

This development does not change a feasibility *class* in the directed model so much as it sharpens the operative content of a class that already carries the cap. It bears most directly on the cross-currency, cross-border relationship analyzed through Interaction Set D (reserve composition jointly binding with monetary sovereignty): the cap is the operative regulatory response to the demand a United-States-dollar-referenced token at scale would place on reserve assets and monetary policy within the euro area, and its operationalization determines the scale at which such a token can be used as a means of exchange there even after any authorization question is resolved. The register carries it as an open question, not a dated trigger, because there is no commencement to apply by date; what will change is the precision with which the existing cap can be modelled, not the existence of the constraint.

*Conditional flag.* The European Banking Authority has not finalized the relevant methodology guidance; the thresholds are fixed by the instrument, but their operationalization is pending. Modelled as an open parameterization, not a dated transition.

### 3.8 Open compositional question: the second Hong Kong licence cohort

Hong Kong has a live regime and a single observation of how it is being administered; the second observation is pending and will resolve a compositional question rather than change a rule. The Stablecoins Ordinance (Cap. 656) entered force on 1 August 2025, and on 10 April 2026 the Hong Kong Monetary Authority granted the first two issuer licences, to The Hongkong and Shanghai Banking Corporation Limited and to Anchorpoint Financial Limited (a joint venture of Standard Chartered Bank (Hong Kong), HKT, and Animoca Brands), with no entity connected to the People's Republic of China in the first batch. The composition of the second cohort, including its currency coverage, whether it includes any People's-Republic-of-China-connected applicant, and whether it admits United-States-dollar-referenced products, has not been announced.

The pending observation is not a change in Hong Kong's rule; it is information that bears on the feasibility of United-States–Hong-Kong and European-Union–Hong-Kong pairs (through the currency and product coverage of the cohort) and, more consequentially, on the stability of the People's Republic of China boundary analyzed in §3.9. A second cohort that included a People's-Republic-of-China-connected entity would itself be a relaxation signal on that boundary; the register carries the cohort composition as an open question whose resolution feeds the boundary-stability question rather than as a trigger that flips a class on its own.

*Conditional flag.* The Hong Kong Monetary Authority has not announced the second cohort. Carried as an open question; not a dated trigger.

### 3.9 Resolved tightening, open relaxation: the People's Republic of China boundary

The People's Republic of China boundary is the one item in the register whose recent movement is *resolved*, a tightening in writing, while the developments that could move it again are *open*. The boundary tightened in explicit written law in February 2026: the eight-ministry Notice 银发〔2026〕42号, in force on 6 February 2026, repealed the 2021 Notice 银发〔2021〕237号 and, for the first time in a joint multi-agency instrument, put in writing a prohibition on issuing renminbi-pegged stablecoins offshore without approval, an extraterritorial bar on entities of the People's Republic of China and their controlled offshore affiliates issuing virtual currencies abroad, and a restriction on onshore real-world-asset tokenization (paired the same day with a China Securities Regulatory Commission guidance, 公告〔2026〕1号, on offshore issuance of asset-backed-security tokens by onshore originators). Every edge terminating in the People's Republic of China is, and remains, blocked at destination; the renminbi has no exportable authorizable private token, so every edge originating there falls into Category III.

The boundary's relevance to the forward map is that its *relaxation* paths are open while its tightening is settled. The Architecture paper's three-pattern typology of responses to the cross-border boundary, comprising direct subsidiary licensing (currently constrained by the joint operation of issuer eligibility and cross-border data sovereignty, Interaction Set A), partnership distribution (operationally implemented), and a constraint-derived separated-entity option whose viability depends on developments not yet resolved, is configured against the boundary as currently observed. None of the three developments that would relax it has occurred: the People's Republic of China has not modified its cross-border data framework to facilitate supervisor-to-supervisor information-sharing with the Hong Kong Monetary Authority; the Hong Kong Monetary Authority has not announced a second cohort including a People's-Republic-of-China-connected entity (the §3.8 question); and the People's Republic of China has not publicly authorized state-owned bank Hong Kong affiliates to apply or operate under Hong Kong licences. The boundary's own framing as a dynamic assessment leaves domestic issuance not currently permitted, pending observation of Hong Kong, rather than permanently foreclosed.

*Conditional flag.* The tightening has occurred and is in force (42号). None of the three relaxation developments has occurred as of publication; the typology patterns are configured against the boundary as currently observed, and a relaxation would alter Pattern A viability without, on its own, changing the destination-blocked class of inbound edges.

### 3.10 Intra-regime gating change: the European Union's MiCA transitional expiry (1 July 2026)

The directed model's feasibility classes capture whether a corridor can be authorized at all; by construction they do not capture a change that alters who may operate *within* an already-live regime without moving the class. The European Union supplies the clearest dated instance of that kind of change, and the earlier draft's silence on it was a transmission gap rather than a research gap, because the register carries the event, as a structured intra-regime gating item in its event calendar, while this forward map omitted it.

The fact is dated and hard, though it is not a single date. The transitional grandfathering window under MiCA Article 143(3), which allowed crypto-asset service providers operating under prior national regimes to continue while seeking MiCA authorization, has 1 July 2026 as its outer limit, the longest window any Member State could grant; the European Securities and Markets Authority confirmed on 17 April 2026 that there would be no extension and that Member States may only shorten the window, never lengthen it. The window is deliberately not uniform: Member States could reduce or disapply it, and most did. Germany and Ireland set twelve months and closed at the end of 2025; Finland, Latvia, Lithuania, Hungary, the Netherlands, Poland, and Slovenia set six months and closed in mid-2025; Sweden closed at nine months in late 2025; and France, Malta, and Luxembourg run the full window to the 1 July 2026 cap. What binds a given cross-border corridor is therefore the earliest deadline among the Member States it actually touches, not the outer cap, and several such deadlines lapsed as early as the middle of 2025. A grandfathered provider was in any case not a MiCA-authorized provider and could not passport, so its cross-border reach during the window already depended on each host State still applying its own transitional regime. On the token side, as of late June 2026, of the largest stablecoins only Circle's USDC and EURC hold MiCA authorization as electronic-money tokens, while USDT, which did not seek authorization, has been delisted from European Union trading venues.

This is not a change in the European Union's feasibility *class*. MiCA is live before and after 1 July 2026, so the directed model reads the European Union as a live regime throughout, and a class-centric measure registers nothing. What changes, on a staggered schedule that completes at the 1 July 2026 outer cap rather than on a single day, is the texture of accessibility *inside* the live regime: the national side-doors close one Member State at a time, MiCA authorization with a passport becomes the single inbound pathway for serving European Union clients, and the set of admissible tokens narrows to those actually authorized. By the paper's own consistency this development has at least the standing of the Singapore and Japan control cases, which earn their place as real, dated changes that move no class. It carries more weight than a control case, because the control cases harden a binding status, whereas this event changes which actors and which instruments can reach the market inside an unchanged class.

The methodological point is the one §6 develops. A sensitivity model centred on class flips systematically under-weights changes that alter accessibility within a live regime: passport eligibility, the set of entities permitted to serve a market, and the set of tokens admissible to it. Such changes can matter more to actual cross-border flow than some of the class flips the paper does track, and the European Union cliff is the example. The paper handles the gap in two ways at once. It records the cliff here, as the first instance of the intra-regime gating category named in the typology; and it states, in §6, the scope boundary explicitly, so that the European Union's low position on the class-sensitivity ordering reads as "no class flip pending," not as "nothing happening in the European Union." The register carries the underlying event in its own event calendar, as a structured intra-regime gating item (effective_date at the 1 July 2026 outer cap, with the staggered national windows recorded beneath it and an empty feasibility-class effect), with the related second-order European Union development, the expected secondary authorization for the custody and transfer of electronic-money tokens, recorded alongside it.

*Conditional flag.* The outer cap is fixed and dated (1 July 2026), the longest window any Member State could grant, and the European Securities and Markets Authority has confirmed no extension; beneath that cap the national windows are staggered and most have already closed, so the date that binds a given corridor is the earliest among the Member States it touches. The change is not contingent. Its effect is an intra-regime gating change, not a feasibility-class transition; the paper records it as such and treats the class-centric sensitivity measure's blind spot to this category as a stated scope limitation (§6), not as an omission.

### 3.11 The trigger register, in one table

The following table collects the eleven entries. "Direction" records whether the trigger changes inbound edges (edges terminating in the jurisdiction), outbound edges (edges originating there), or both; "Class effect" records whether the trigger moves a feasibility class or only a binding status, or acts within a class; "Certainty" records the kind of conditioning the trigger carries.

| Trigger | Jurisdiction | Kind | Direction | Class effect | Certainty |
|---|---|---|---|---|---|
| Regime commences (SI 2026/102) | UK | Commencement | Inbound | T → I (8 caveated pairs → 0) | **Scheduled** — 2027-10-25 |
| Virtual Asset Service Act enacted | TW | Enacted, not commenced | Both (on commencement) | Pre-regime → live; inbound → **Cat. II (channelled)** | **Enacted 2026-06-30**; commencement pending subsidiary legislation, no date gazetted |
| CLARITY Act enacted (H.R. 3633) | US | Enactment | Intermediary/routing layer | Binding status + §404 overlays operative | Contingent — Calendar No. 423 |
| Digital Asset Basic Act enacted | KR | Enactment + commencement | Both | Pre-regime → live; out of Cat. III | Contingent — in subcommittee |
| SCS implementing legislation enacted | SG | Enactment | — | Binding status only (no class change) | Contingent — expected mid-2026 |
| Act No. 66 of 2025 reserve relaxation | JP | Dated-empty-effect | — | **None** (control case) | **In force 2026-06-01** |
| Foreign trust-type stablecoin recognition | JP | Inbound recognition | Inbound | Recognition pathway; class by conditioning (Cat. II) | **In force 2026-06-01** |
| Article 58(3) cap operationalized | EU | Parameterization | — | Sharpens existing class | Open question — no commencement |
| MiCA Art. 143(3) transitional expiry | EU | **Intra-regime gating** | Inbound (accessibility within live regime) | **No class change**; closes national side-doors, passport-only | **Outer cap 2026-07-01**; staggered national windows, most closed earlier (ESMA: no extension) |
| Second licence cohort | HK | Administration | Inbound (feeds CN boundary) | Information, not a rule change | Open question — not announced |
| Boundary stability | CN | Posture | Both (typology viability) | Tightening resolved; relaxation open | Mixed — 42号 in force |

## §4. Corridor sensitivity and the macro readings it supports

Section 4 does two things in sequence. It first establishes the sensitivity measure and the ordering it produces (§§4.1–4.2), reading the trigger register edge by edge to locate where on the directed map the pending movement is concentrated. It then draws the macro readings that measure makes possible: a compact where-to-watch ordering (§4.3), the same forward map re-sorted into the per-jurisdiction view a supervisor needs (§4.4), and a cross-jurisdictional convergence the register surfaces when it is read across jurisdictions instead of along edges (§4.5). The first pair is the sensitivity analysis proper; the readings that follow are what the register makes legible once the measure is in hand.

### 4.1 What sensitivity measures

Reading the trigger register edge by edge yields a structural observation the static map cannot: the pending movement is not evenly distributed across the directed space. A corridor's *sensitivity* is the number of its directed edges that change feasibility class across the pending triggers, holding the contingent triggers as if-enacted branches. A high-sensitivity jurisdiction is one whose pending trigger, on resolution, reclassifies many edges; a low-sensitivity jurisdiction is one whose edges are stable across the pending set, either because it is already live and its pending change is a control case, or because its destination class is insensitive to any positive trigger.

It is worth separating what is nearly obvious here from what is not, because the obvious part is only a sanity check and the analytical content lies elsewhere. The headline that a pre-regime jurisdiction is highly sensitive and a prohibition destination is insensitive is close to a tautology: a jurisdiction with no issuance regime has the most to gain from acquiring one, and an inbound edge blocked by a prohibition has nothing a *positive* trigger can move. That the measure ranks them so is a check that it behaves sensibly, not a finding. The non-obvious content is two facts about the *topology* of the directed corridor graph. The first is an edge-count asymmetry: a both-directions trigger at a pre-regime origin reclassifies edges on *both* the inbound and the outbound side of that jurisdiction at once, so the number of edges it moves is on the order of twice that of a trigger that opens only one side. The volatility concentrates at the pre-regime origins not only because they have "more to gain" but because, in a directed graph where every jurisdiction connects to every other, a both-directions change touches roughly twice the edge count of a one-directional one. The second is a second-order, in-degree effect: a single destination-side commencement reclassifies *every* inbound edge from a token-holding origin simultaneously, so the leverage of one change in law scales with the jurisdiction's in-degree in the corridor graph. The United Kingdom commencement is "clean" not only because it is dated but because one event resolves an entire fan-in of edges at once. Sensitivity, properly read, is therefore a statement about graph structure, about whether a trigger is one- or both-directional and the in- and out-degree of the jurisdiction it touches, and the pre-regime/prohibition headline is the sanity check that the structural measure is working.

### 4.2 The ordering

The ordering has three anchors. First, the one remaining pre-regime origin, South Korea, is the single highest-sensitivity jurisdiction, because it is the subject of a both-directions trigger: enactment would change every edge terminating in South Korea, lifting it out of the pre-regime class, and every edge originating there, lifting it out of Category III. A single contingent enactment in South Korea therefore reclassifies the largest number of edges of any pending trigger, which is why, after Taiwan's enactment, the forward map's remaining both-directions volatility concentrates on this one origin. Second, the **three** regime-in-transition destinations — the United States, the United Kingdom and Taiwan — are the next most sensitive, and the comparison among them is itself informative, because they differ on both axes that matter to a forward plan: the *kind* of certainty their date carries, and the *class* they resolve to. The United States comes first, because its horizon is the earliest: its eight inbound edges from token-holding origins sit at Category T and resolve to a channelled Category II on the earlier of 18 January 2027 or final rules plus one hundred and twenty days — a `scheduled_with_cap` trigger, not a gazetted date, so the horizon is an upper bound that may fall sooner. The United Kingdom is the highest-*certainty* scheduled jurisdiction: its commencement reclassifies its eight inbound edges from Category T to a clean Category I on a known day, 25 October 2027, and resolves the eight-pair caveat in full, the cleanest worked transition in the register because its date is gazetted rather than capped. Taiwan sits in the same lifecycle stage but resolves differently: it is enacted rather than scheduled, with no gazetted commencement date, and its inbound edges will resolve to Category II, a channelled interface, rather than Category I, because its enacted gate is approval-based. The two regime-in-transition destinations therefore share a stage but differ on both axes that matter to a forward plan, the certainty of timing and the class they resolve to. Third, the prohibition destination, Mainland China, is insensitive: no positive trigger can reach an inbound edge blocked by an issuance prohibition, so every edge terminating in the People's Republic of China is stable across the entire pending set, and the only available movement is a relaxation of the boundary that bears on the typology of responses, Pattern A viability, and not on the destination-blocked class itself.

The remaining jurisdictions occupy the low-to-middle range for distinct reasons. The United States is middle-sensitivity in an unusual way: its pending trigger changes the intermediary and routing layer and the binding status of specific cells without reclassifying its inbound gate wholesale, so its sensitivity concentrates in the operative content of the routing architecture, not in a large count of reclassified edges. Singapore and Japan are low-sensitivity by construction, each having a change in law that the register records as a control case with no feasibility-class effect, because the baseline that determines the class is already in force. The European Union is the case the next paragraph and §6 single out, because it is low-sensitivity in the class dimension and active in two others. Its Article 58(3) parameterization sharpens the operative content of an existing cap without moving a class, and its 1 July 2026 MiCA transitional expiry is an intra-regime gating change that closes national side-doors and narrows the set of admissible tokens, again without moving a class. A class-centric sensitivity reading therefore places the European Union low, and that placement must be read as "no class flip pending," not as "nothing happening," because the intra-regime gating change is real, dated, and consequential for who can actually serve the market. Hong Kong is stable in its own class and relevant chiefly as the source of information, the second cohort, that feeds the People's-Republic-of-China boundary question.

### 4.3 The macro reading

The sensitivity ordering produces a compact "where to watch" map for anyone tracking the forward feasibility of the cross-border space. The one contingent pre-regime enactment, South Korea, is the event that would reclassify the most edges and is therefore the highest-information pending trigger, notwithstanding that it carries no date. The two regime-in-transition destinations are the next to watch, on different clocks: the United Kingdom on a gazetted date in 2027, resolving to a clean Category I, and Taiwan on an as-yet-ungazetted commencement following its 30 June 2026 enactment, resolving to a channelled Category II. The United States enactment is the event whose effect is deep rather than broad, concentrated in the routing-architecture boundary and the intermediary-yield question, and whose resolution is roughly even on the current public record. The European Union is the case to watch not for a class flip but for accessibility changes inside a live regime: the 1 July 2026 MiCA transitional expiry and the operationalization of the Article 58(3) cap both change who and what can reach the market without moving a class, the kind of change a class-centric reading must be reminded to track. The prohibition axis, Mainland China, is the axis on which to watch for relaxation signals rather than reclassification, with the second Hong Kong cohort the most proximate such signal. And the control cases, Japan and Singapore, are the events to record as changes in law that are not changes in feasibility, so that the forward map is not moved by them. This is the macro contribution of the trigger register: not a forecast of which events will occur, but a structured account of which corridors each pending change in law would move and by how much, together with a reminder of the changes that move accessibility without moving class, so that attention can be allocated to where the directed map is actually in motion.

### 4.4 A per-jurisdiction reading, for the supervisor

The register organizes the forward map by trigger, which suits an analyst asking "what is moving across the whole space." A supervisor asks a different question, organized by jurisdiction rather than by event: "which pending developments will change the corridors into and out of *my* jurisdiction, and which counterpart jurisdictions am I most exposed to." The same underlying data answers both, because every trigger in §3 already names the jurisdiction it touches, the direction in which its edges change, and the counterpart ends affected, and the register exposes a per-jurisdiction projection that re-sorts the trigger register into exactly that view. For a given jurisdiction the projection lists its own pending events, the directed edges into and out of it that those events would reclassify, and the counterpart jurisdictions on the other end of the most sensitive of those edges. A European Union supervisor, to take the jurisdiction this paper most needs to serve well, would read from that projection not "the European Union is low-sensitivity" but the operationally useful statement that no inbound or outbound corridor changes feasibility class in the pending window, while the 1 July 2026 transitional expiry changes which domestic and third-country actors may serve European Union clients and which tokens they may offer. Presenting the forward map in this per-jurisdiction form is the single change that would make it directly usable by the supervisory readership, and it requires no new data, only the re-sort the register already supports.

### 4.5 A cross-jurisdictional convergence the register makes legible: the yield line

Read across jurisdictions rather than edge by edge, the register surfaces a second macro finding the corridor view does not: independent legislative systems are converging on the same functional boundary around stablecoin yield, between yield paid for merely holding the instrument, which is prohibited, and activity-linked rewards, which are permitted. The convergence is documented most fully for the United States, where the GENIUS Act prohibits issuer-paid yield to holders "solely in connection with the holding, use, or retention" of a payment stablecoin (§3.2) and the Senate Banking substitute for the pending market-structure layer carries the Tillis–Alsobrooks compromise, under which yield for holding a balance is prohibited while stablecoin rewards and activity-linked incentives are allowed. Singapore expresses a sibling form of the same principle: its Single-Currency Stablecoin framework forbids the issuer from lending or staking the reserve (§3.5), holding the token to a pure means of payment, not a yield product. The claim is made at the register's own discipline, which is what keeps it a finding, not an impression. The yield treatment is documented at citable depth for the United States and recorded as a sibling restriction for Singapore, while the corresponding cells for the other live and pending regimes in the set are carried in the verification backlog, not asserted here as drawing the same line. What makes the convergence checkable is that yield is one of the typed dimensions every jurisdiction record carries, so the same red line can be confirmed or refuted jurisdiction by jurisdiction against the primary instrument as those cells are verified, and the functional test that decides whether a given arrangement falls inside or outside it, the conversion-direction, ownership-disclosure, timing-independence, yield-attribution, and audit-trail test the Architecture working paper sets out, is the same one across them. A convergence asserted this way, documented where the record supports it and flagged as backlog where it does not, is the kind of cross-jurisdictional observation a register supports and a flat compliance table cannot.

## §5. The open–closed boundary in wholesale settlement infrastructure

### 5.1 Why the static map cannot express this

The directed map carries an infrastructure attribute on every edge, recording whether the origin and destination sit on a shared wholesale settlement experiment, and it distinguishes a Digital-Financial archetype, the broader channel in which a wholesale settlement infrastructure and the data-sovereignty and capital-flow rules jointly determine what can move and how it is recorded. But the map reads that attribute edge by edge; it does not step back to the layer at which the attribute is decided. This part does. It reads the same infrastructure attribute as the entry point to a structural question about how the cross-border stablecoin stack is organized: which layers of that stack are consolidating into *shared, open* infrastructure, which remain *competitive, proprietary, closed*, and where the regulatory layer is standardizing toward interoperability versus remaining walled.

The question is not a commercial one about any actor's product. It is an architectural reading of the public record: the wholesale settlement experiments are central-bank-coordinated, multi-jurisdiction efforts described in the public central-bank literature, and the regulatory interoperability mechanisms of recognition, equivalence, and comparability are public legal constructs. The open–closed boundary in this part is the boundary between layers of a system, read off the public record, not the boundary between an open and a closed codebase of any firm.

One discipline must be stated before the analysis proceeds, because this paper's companion methodology insists on it and this paper is bound by it: the two halves of §5 rest on evidentiary bases of *different grade*, and the difference is declared, not smoothed over. The regulatory-interface reading (§5.3), which maps the feasibility classes onto open, channelled, and closed regulatory interfaces, re-anchors entirely to the `tier1_legal` instruments of §3. It is a re-description of propositions of law about in-force and pending instruments, and it carries their evidentiary grade. The settlement-substrate reading (§§5.2 and 5.4) rests on two softer bases: the public central-bank project literature, which is reliable as to the existence and design of the experiments but forward-looking as to their interaction with the private issuance layer; and the register's `infrastructure_overlap` attribute, which the register *itself* classifies as a Tier-2, operational enrichment layer. That tier is explicitly weaker than the propositions of law on which §3 rests, and the register tracks it as a per-edge enrichment backlog, not as confirmed legal fact. The reader is therefore asked to apply, inside this paper, the same `claim_class` discipline the register applies to its own data: the substrate reading is operational-grade context, not a proposition of law, and it is offered as a structural lens, not a citable legal conclusion. This is the documentary analog of the citable-purity gate the methodology paper sets out. Just as that gate forbids a market fact from entering the instrument a lawyer is shown, and the signal-provenance gate forbids a derived legal conclusion from resting on an operational fact, this section marks where its own reading crosses from the legal tier into the operational-enrichment tier, so that no reader mistakes the softer half for the harder one. Crucially, *no feasibility-class conclusion in §§3–4 rests on this softer layer*. The load-bearing forward map, meaning the triggers, the binding-status transitions, and the corridor-sensitivity ordering, is derived entirely from `tier1_legal` instruments, and the settlement-substrate reading that follows is descriptive context layered on top of it, not a load-bearing input to it.

### 5.2 The bifurcation, corrected: a settlement substrate split along bloc lines, beneath a competitive issuance layer

The cross-border stablecoin stack has two layers with different organizing principles, and the wholesale settlement substrate beneath the issuance layer is not the single shared, open rail an earlier reading suggested. The public record shows the substrate itself split along geopolitical bloc lines, open within a bloc and closed across blocs, and getting both the attribution and the structure right is what turns this section from a soft analogy into a checkable claim.

The attribution must be exact. Of the wholesale settlement experiments, only Project Agorá is led by the Bank for International Settlements, together with seven central banks of the G7-aligned bloc and more than forty private financial institutions; it is a tokenized-correspondent design that keeps value moving through the existing dollar-centred correspondent structure. Project Ensemble is led by the Hong Kong Monetary Authority, was launched in March 2024, and is not a Bank for International Settlements project. And mBridge is no longer a Bank for International Settlements project at all: the Bank for International Settlements withdrew on 31 October 2024, and the platform is now operated by the People's Bank of China, the Hong Kong Monetary Authority, and the central banks of Thailand, the United Arab Emirates, and Saudi Arabia, with the large majority of settlement volume reported in digital yuan and no Western central bank among its operators. The earlier draft attributed all three to the Bank for International Settlements, which was wrong, and the correction is not cosmetic, because the true operators of each experiment are what reveal the structure.

That structure is a substrate split along bloc lines. Project Agorá is the G7-aligned bloc's answer and preserves dollar routing; mBridge is the answer of a People's-Republic-of-China-and-Gulf bloc, designed to settle across monetary areas while bypassing the dollar leg. The two are competing, bloc-sponsored answers to the same problem, their memberships do not overlap, and they are not mutually interoperable; the earlier era of a single multilateral central-bank interoperability effort has given way to rival bloc rails. Project Ensemble, led by Hong Kong, sits in the structurally interesting bridging position between the two blocs. The settlement substrate is therefore not a single open layer but two bloc-internal layers, each open to its own members and closed across the divide, and a corridor's place on that substrate depends on which bloc, if any, both of its ends belong to.

This is what the directed map's infrastructure-overlap attribute can record edge by edge, and recording it that way is what makes the bifurcation a checkable structural finding rather than a narrative. For each directed edge the attribute can mark whether both ends sit on the same bloc-internal substrate, on opposite blocs, or off the shared experiments entirely, which yields a concrete edge-level classification a reader can verify against the public membership of each experiment. The claim is then not the soft assertion that settlement is consolidating but the specific one that the substrate is bloc-partitioned and that the partition is visible at the level of individual corridors.

The reframing strengthens §5 by unifying its layers under a single structure. The issuance layer is fragmented per instrument and per national authorization, as the Architecture paper establishes; the settlement substrate is partitioned by geopolitical bloc, as just corrected; and the regulatory layer, in §5.3, is open or closed at each boundary by the availability of a recognition determination, with those determinations themselves tending to follow bloc lines. The three layers are projections of one fracture. The regulatory recognition boundary (§5.3), the settlement-substrate boundary (this section), and the People's-Republic-of-China issuance boundary (§3.9) are the same geopolitical division seen at three layers of the stack, and that unified claim, resting on the public membership of the settlement experiments and the public legal record of recognition, is both more accurate and more useful than the single-open-substrate picture it replaces.

### 5.3 Recognition and equivalence as the interoperability protocol of the regulatory layer

There is a second open–closed boundary, in the regulatory layer itself, and it is the one most directly connected to the feasibility classes this paper has been moving. The regulatory layer "interoperates" through recognition, equivalence, and comparability determinations, the legal constructs by which one jurisdiction admits another's authorized instrument, and these function as the interoperability protocol of the regulatory layer in the same way a shared settlement experiment functions as the interoperability protocol of the infrastructure layer. Where such a determination exists or is available, the regulatory layer is "open" at that boundary: the corridor is a Category I dual-authorization relationship, or a Category II relationship operable through a recognition channel. Where no such determination exists, the regulatory layer is "closed" at that boundary, and the corridor falls into the composition-unresolved, blocked, or pre-regime classes.

This maps the feasibility classes onto the open–closed reading directly. A Category I edge is an open regulatory interface: each end separately authorizable, the routing architecture portable. A Category II edge is a *channelled* interface, open but only through a specific determination not generically granted, such as the United States comparability gate, the Japanese recognition-and-registration channel, the United Arab Emirates onshore usage-channel registration, or the Brazilian counterparty-conditional foreign-exchange channel. A Category III edge, a blocked edge, and a pre-regime edge are, in this reading, *closed* regulatory interfaces, for three different reasons: an unresolved composition, an issuance prohibition, and the absence of any authorizable pathway. The asymmetry the directed model insists on, that inbound treatment is systematically tighter than outbound, so that the edge from the United States to the European Union (Category I) need not match the edge from the European Union to the United States (Category II, the comparability gate), is in the open–closed reading simply the observation that a regulatory interface can be open in one direction and channelled in the other.

Recognition as the interoperability protocol is not only an abstraction; it has a dated instance on the public record. On 1 June 2026 Japan, by Cabinet Office Ordinance, recognised foreign trust-type stablecoins judged equivalent to its own regime as electronic payment instruments (§3.6). That is exactly this protocol in operation: one jurisdiction admitting another's authorized instrument through an equivalence determination, opening a channelled, Category II interface whose operative reach is set by the recognition criteria. The Japanese recognition is the worked example that grounds the otherwise abstract claim that recognition and equivalence are the regulatory layer's interoperability mechanism, and it dates that mechanism to a specific instrument on a specific day.

### 5.4 Where the boundary is moving

The two open–closed boundaries, the horizontal one between the settlement substrate and the issuance layer and the boundary within the regulatory layer between open and closed interfaces, are themselves moving, and the trigger register of §3 is, read through this lens, a record of where the regulatory interfaces are opening or being created. A pre-regime enactment (South Korea now, and Taiwan on its commencement) *creates* a regulatory interface where none existed, opening a previously-closed boundary for the first time. A regime commencement (the United Kingdom) *opens* an interface that was adopted but not yet operative. A market-structure enactment (the United States) sharpens the operative content of an interface that is open at the issuer layer but whose intermediary-and-routing interface is still being defined. A boundary's tightening (the People's Republic of China) holds an interface closed and, in writing, narrows the conditions under which it might open. And the open parameterization (the European Union Article 58(3) cap) sets the *width* of an interface that is open but capped: how much can flow through it as a means of exchange within the currency area.

The settlement substrate, by contrast, is the layer on which the open questions about infrastructure implications bear, and the substrate is partitioned by bloc, as §5.2 corrects. How the rival bloc rails will develop, whether either will admit the private issuance layer, and whether any bridge across the divide emerges are forward-looking and unresolved, and the Corridor Atlas develops the infrastructure-overlap attribute precisely so that this boundary can be tracked edge by edge as the experiments mature. The structural prediction this part does *not* make is which way the boundary will move. The structural observation it does make is that the cross-border stablecoin stack is divided along one geopolitical fracture at three layers at once: the feasibility classes are the regulatory layer's open-or-closed state, the pending triggers are the events that open, channel, widen, or hold closed the regulatory interfaces, and beneath them the settlement substrate is split between bloc-internal rails on which the deeper and slower contest is being decided. Running through the layered reading is a single organizing axis worth naming: regimes structurally distinguish foreign-currency from domestic-currency stablecoins, an axis visible at the regulatory layer in the Article 58(3) cap's explicit condition on tokens denominated in a currency that is not an official currency of a Member State (§3.7) and at the substrate layer in the split between a dollar-routing rail, Project Agorá, and a dollar-bypass rail, mBridge (§5.2). That monetary-sovereignty axis is, like the rest of the substrate reading, an interpretive structural observation resting on the softer evidence §5.1 flags, not a class-level conclusion, and it attributes no motive to any particular legislature beyond what the cited text states.

*Conditional flag.* The infrastructure-level implications of the wholesale settlement experiments are forward-looking; the public central-bank record describes the experiments but their interaction with the private issuance layer at scale is unresolved. As stated in §5.1, the settlement-substrate half of this section rests on the central-bank project literature and on the register's `infrastructure_overlap` attribute, a Tier-2, operational enrichment layer the register classifies as an explicitly weaker evidence tier than the `tier1_legal` instruments behind §3 and tracks as a per-edge enrichment backlog. The substrate reading is therefore offered as operational-grade structural context, not as a proposition of law, and the same `claim_class` discipline the register applies to its own data is asked of the reader here. The regulatory-interface reading (§5.3), by contrast, rests on the in-force and pending instruments analyzed in §3, with their conditional flags preserved and their evidentiary grade carried; and no feasibility-class conclusion in §§3–4 depends on the softer substrate layer.

## §6. What this paper is not, and its limitations

The paper is a conditioning, not a forecast. Every contingent trigger in §3 is surfaced as an if-enacted branch and applied by no date; the paper takes no position on whether or when any contingent enactment will occur, and the prediction-market and procedural-stage figures it cites are reported as the state of the public record, not as the author's estimate. Of the changes that carry a date, the United Kingdom commencement is the one applied by a future gazetted date; the Japanese reserve relaxation and recognition of 1 June 2026 are in force and recorded as such, as is the European Union MiCA transitional expiry, whose national windows are mostly closed beneath a 1 July 2026 outer cap; and Taiwan is enacted but carries no gazetted commencement date, so its transition too is applied by no date. The open parameterizations and open compositional questions (the Article 58(3) operationalization, the second Hong Kong cohort, the People's-Republic-of-China relaxation paths) are carried as open questions, not dated transitions, because there is no commencement to apply by date.

One scope boundary must be stated explicitly, because the European Union entry of §3.10 exposes it. This paper's sensitivity model is centred on the feasibility *class*: it tracks changes in law that move a corridor between classes, and it ranks jurisdictions by how many edges their pending changes reclassify. That focus systematically under-weights a different and sometimes larger kind of change, the intra-regime gating change, which alters accessibility within a live regime without moving any class, governing which entities may be authorized or passported to serve a market and which instruments are admissible to it. The European Union's 1 July 2026 transitional expiry is exactly such a change, and its effect on real cross-border flow may exceed that of some class flips the paper does track, because it determines who can lawfully reach the largest and most active regulated market in the set. The paper does not claim to measure these intra-regime changes on the same footing as class flips; it names the category, records the European Union instance, and states the boundary, so that a low position on the class-sensitivity ordering is read as the absence of a pending class flip and not as the absence of regulatory movement. Extending the sensitivity model to weigh intra-regime gating changes alongside class flips is identified as future work, not claimed here.

The paper rests on the public regulatory record and defers to the companion works for the underlying facts: the Compliance Matrix for the version-stamped jurisdictional provisions, the Architecture paper for the constraint vocabulary and the interaction sets, and the Corridor Atlas for the directed-edge model and the infrastructure attribute. Its forward transitions inherit the register's standing limitations: the computed layer beneath them is a labelled preview, not asserted authoritative, and its disagreements with the authored analysis (the European-Union–United-Kingdom and United-Kingdom–United-States structural residual after 2027) are surfaced as findings rather than resolved; the feasibility derivations rest on a hand-curated signal table whose backing cells include a declared backlog of records confirmed at regime-status level but not yet against the official text; and the precise post-commencement class of the inbound edges created by a pre-regime enactment is provisional pending the enacted text. None of these is concealed; each is the honest boundary of a forward map built on a record that is itself in motion.

Finally, the paper advocates no outcome. The transitions it maps are analytical, and the open–closed structure it reads is descriptive of the system's architecture as currently in force and pending. Whether any of the pending interfaces *should* open, whether the wholesale settlement substrate *should* interoperate with the private issuance layer in any particular way, and whether any feasibility class *should* be moved by regulators are questions outside its scope, addressed, where the author addresses them at all, in separately-labelled advocacy work and never in this descriptive map.

## §7. Conclusion

The companion works in this portfolio map cross-border stablecoin feasibility as it stands; this paper maps how it moves. Its instrument is the trigger, a change in law and never a market event, and its method is conditioning: for each change in law visible on the public record as of 30 June 2026 it states which directed edges change feasibility class, which records the change rests on, and under what conditional status the change is contingent. Two contributions follow. The first is the conditioning method itself, applied to a deliberately sparse current instance that will densify as enactments land. The second is a typology of triggers by the kind of certainty each carries, which separates the one kind that moves a dated horizon from the kinds that do not, and which now includes a category a class-centric reading easily misses, the intra-regime gating change that moves accessibility within a live regime, with the European Union's 1 July 2026 transitional expiry as its worked example. The corridor-sensitivity ordering finds the forward map's class-level motion concentrated at the one remaining pre-regime origin, South Korea (a both-directions trigger over twenty-two edges), and the three regime-in-transition destinations: the United States resolving to a channelled Category II on an outer cap of 18 January 2027, the United Kingdom resolving to a clean Category I on a gazetted 25 October 2027, and Taiwan resolving to a channelled Category II on an as-yet-ungazetted commencement after its enactment on the dateline of this draft, with the prohibition destination insensitive to any positive trigger. The more interpretive reading it layers on top, on softer evidence, finds a stack divided along one geopolitical fracture at three layers at once: a settlement substrate split between rival bloc experiments, with only Project Agorá led by the Bank for International Settlements, mBridge run since that institution's 2024 withdrawal by a People's-Republic-of-China-and-Gulf group, and Project Ensemble bridging from Hong Kong; a regulatory layer open, channelled, or closed at each boundary by the availability of a recognition determination, dated by Japan's 1 June 2026 equivalence recognition; and an issuance layer fragmented per instrument. The map is forward-looking and the record is in motion, as Taiwan's same-day enactment shows; the paper's discipline is to make the if-then explicit, to keep the map a dated and register-backed living document, and to leave the if to the reader.

## References

> **A note on this bibliography.** Consistent with the verification discipline of the companion methodology paper and of the register this paper rests on, the four formal citations below have been verified against their primary records (the publishers' own pages: the American Economic Association, JSTOR, Elsevier's *Futures*, and *MIT Sloan Management Review*), and each carries its DOI or stable identifier and confirmed volume, issue, and pages. Two of them, Habegger and Schoemaker, are named as representative primary sources for the horizon-scanning and scenario-analysis traditions this paper positions against; a different canonical source may be substituted to fit a particular venue. The author's portfolio is self-published with persistent identifiers. The regulatory facts underlying §§3–5 are sourced in the register and the companion working papers, not re-cited here.

*Methods this paper is positioned against (§1.4).*

- MacKinlay, A. C. "Event Studies in Economics and Finance." *Journal of Economic Literature*, 1997, 35(1): 13–39. JSTOR 2729691.
- Wolfers, J. & Zitzewitz, E. "Prediction Markets." *Journal of Economic Perspectives*, 2004, 18(2): 107–126. DOI: 10.1257/0895330041371321.
- Habegger, B. "Strategic Foresight in Public Policy: Reviewing the Experiences of the UK, Singapore, and the Netherlands." *Futures*, 2010, 42(1): 49–58. DOI: 10.1016/j.futures.2009.08.002. (Representative of the regulatory horizon-scanning / strategic-foresight tradition.)
- Schoemaker, P. J. H. "Scenario Planning: A Tool for Strategic Thinking." *Sloan Management Review*, 1995, 36(2): 25–40. (Representative of the scenario-analysis tradition.)

*Author's portfolio (self-published; persistent identifiers).*

- Fan, Y. *Cross-Border Stablecoin Register (CBSR)*, v0.10.0. Data CC-BY-4.0; code Apache-2.0; archived to Zenodo (DOI 10.5281/zenodo.20730358; cite the version DOI). ORCID 0009-0005-6762-084X · SSRN Author ID 11463068. The event calendar, time engine, and per-jurisdiction signal table rendered in §§3–4 are drawn from this register.
- Fan, Y. *Citable by Construction: An Evidence Model for Machine-Readable Comparative Law*, v0.1.0 (working paper): the companion methodology paper and the epistemic guarantor of this forward map; §§3.1, 5.1, and 6 defer to its evidence model and labelled-preview discipline.
- Fan, Y. *Multi-Jurisdiction Stablecoin Compliance Matrix*, v0.9.7 (working paper): the node facts.
- Fan, Y. *Cross-Border Stablecoin Architecture*, v0.2.8 (working paper): the eight-constraint vocabulary and the six interaction sets.
- Fan, Y. *Cross-Border Digital-Finance Corridor Atlas*, v0.2.5 (working paper): the directed-edge model, the feasibility classes, and the infrastructure-overlap attribute.

---

*Companion works.* This paper defers to and builds on: Fan, *Multi-Jurisdiction Stablecoin Compliance Matrix* (v0.9.7); Fan, *Cross-Border Stablecoin Architecture* (working paper, v0.2.8); Fan, *Cross-Border Digital-Finance Corridor Atlas* (v0.2.5); and the Cross-Border Stablecoin Register (CBSR, v0.10.1; data CC-BY-4.0, code Apache-2.0; archived to Zenodo for a DOI per release). The forward events, the per-jurisdiction signal table, and the dated transitions rendered in §§3–4 are drawn from the register's event calendar and time engine. The evidence model on which this paper's discipline rests, comprising the orthogonal claim-kind, provenance, and binding-status axes, the labelled-preview posture of the computed layer invoked in §3.1 and §6, and the documentary application of the citable-purity discipline in §5.1, is set out in full in Fan, *Citable by Construction: An Evidence Model for Machine-Readable Comparative Law* (working paper, v0.1.0), which is the epistemic guarantor of this forward map and to which §§3.1, 5.1, and 6 defer. Policy positions referenced as out of scope are developed in separately-labelled advocacy work.

*Version 0.1.0 · June 2026 · Pre-release working-paper draft for community review.*