Issuing an EMT under MiCA Title III: The 2026 Operating Playbook

The single most common mistake we see among 2026 EMT-issuance projects is treating the build as a CASP-style licence application. Founders read MiCA cover-to-cover, see Title V crypto-asset service provider authorisation, and map their token plan onto it. The reserve language at Article 36 gets noted. The Title III references get filed under "also relevant". Six months later the project hits an authorisation gate and the discovery happens: EMTs are not a CASP service, they are an issued instrument, and the regime that governs them is structurally different from the one governing exchanges and custodians.

The structural fact is this: EMTs are issued under MiCA Title III — Articles 48 to 58. CASPs are authorised under MiCA Title V — Articles 59 to 74. The two titles share a regulation, a competent-authority architecture, and a passporting framework, but the operative regimes are different. An EMT issuer must be authorised as a credit institution or as an electronic money institution under EMD2 (and, prospectively, PSD3) with a MiCA Title III overlay. A CASP merely needs Title V authorisation. The capital, governance, reserve, redemption, and disclosure architectures bear almost no resemblance to each other.

This article maps what we call the EMT issuance stack — the four-layer build that an EMT issuer must complete to ship a token into the EEA market in 2026: Layer 1 — Authorisation (Articles 48–50); Layer 2 — Reserve assets (Article 36); Layer 3 — Redemption mechanics (Article 39); and Layer 4 — Ongoing disclosure. Each layer is a separately-evidenced build. Each has its own NCA and EBA-side review threshold. None of them are operationally similar to a CASP application — which is the point.

The Markets in Crypto-Assets Regulation¹^[1] splits the universe of regulated crypto-asset activity into three structural regimes. Title II covers offers and admissions to trading of crypto-assets other than ARTs and EMTs. Title III covers electronic money tokens — single-fiat-referenced tokens. Title IV covers asset-referenced tokens. Title V covers crypto-asset service providers. The architecture matters: issuers are not service providers. An EMT issuer is shipping a financial instrument; a CASP is intermediating one. They sit in different titles because they are different businesses.

The operational consequence is that the EMT issuer build looks nothing like a CASP build. A CASP application is dominated by service-process documentation: custody segregation, order-execution policy, best-execution evidence, market-abuse surveillance, complaints handling. An EMT issuer application is dominated by prudential architecture: who holds the reserve, what is in it, how it is segregated, how it is valued, how redemptions are funded, and what the firm publishes about its composition and stress posture on an ongoing basis. The CASP is a process firm; the EMT issuer is a balance-sheet firm with a token attached.

A second structural difference: the authorisation prerequisite. A CASP is authorised directly under Title V — no prior banking or EMI licence is required, and Annex IV-class capital can be as low as €50,000. An EMT issuer cannot be authorised standalone under MiCA: it must already be a credit institution or an EMI, and then layer MiCA Title III on top. This means the EMT issuer build is in effect two licence projects — first an EMI authorisation (or credit-institution licence), then the MiCA Title III overlay — sequenced and pricedaccordingly.

We map the build as four sequential and operationally distinct layers. They are not optional and they are not parallel — Layer 1 gates the activity, Layer 2 funds it, Layer 3 stabilises it, and Layer 4 keeps it disclosable. Skipping or under-building any single layer is the most common reason an EMT project either fails authorisation or fails the EBA's significance review later.

Article 48 of MiCA sets the gating rule: only a credit institution or an authorised electronic money institution may issue an EMT in the EEA. Article 49 specifies the notification and white-paper requirements an authorised issuer must satisfy before going live. Article 50 covers ongoing supervisory obligations and the conditions under which authorisation can be withdrawn. The combined effect is to make EMT issuance a regulated overlay on a regulated entity — never a standalone permission.

For a green-field project the practical sequencing is: incorporate the operating entity, apply for the underlying EMI licence under EMD2 (typical timeline 9–15 months at the faster EU NCAs, 12–18+ months at the conservative ones), then prepare and submit the EMT white paper to the home-state NCA at least 40 working days before the planned offer date, as required by Article 51. The NCA can object to the white paper inside that window. Once cleared, the issuer may offer the EMT and request admission to trading on regulated venues.

Article 36 — incorporated into the EMT regime by reference — is the prudential heart of the stack. It requires that the reserve of assets backing an EMT be at least 1:1 with the outstanding token supply, denominated in the reference currency, segregated from the issuer's own assets, custodied with authorised credit institutions, CASPs providing custody services, or investment firms, and invested only in highly liquid financial instruments with minimal market and credit risk. The European Banking Authority²^[2] has issued Regulatory Technical Standards specifying liquidity requirements, the precise composition limits, concentration ceilings, and the stress-testing the issuer must run on the reserve.

For an EMT specifically — as distinct from an ART — the reserve composition is more constrained. The reserve must reflect the single referenced fiat: a EUR-EMT's reserve sits in EUR, a USD-EMT's reserve in USD. The instrument universe in practice converges on central-bank deposits where access exists, short-dated sovereign debt of the reference jurisdiction, and short-term deposits at authorised credit institutions — with EBA-defined concentration limits on per-issuer exposure. The issuer cannot reach for yield by running duration or credit risk on the reserve; the regime is designed to be conservative by construction.

Article 39 of MiCA establishes the redemption right that is unique to EMTs and central to the regime's stability claim: any holder may redeem at par value at any time. Redemption is to be effected free of charge or against a fee proportional to actual costs, and within standard business-day execution windows. There is no minimum redemption size that can override the right, no holding period the issuer can impose, and no discretion to settle at anything other than par.

The operational consequence is that the reserve must be continuously redemption-funded. The issuer's liquidity-risk framework must size the same-day-liquid sleeve to cover plausible peak redemption days. That is where the EBA's RTS on reserve-asset liquidity bites in practice — the supervisor expects the issuer to demonstrate that the liquidity profile of the reserve matches the demand profile of the token. A reserve invested in 12-month sovereign paper at par is not, by itself, Article 39 compliant; the redemption SLA has to be backed by reserve composition, not just by the existence of paper assets.

The white paper is not a one-off document. Under MiCA — and under technical standards developed by the European Securities and Markets Authority³^[3] jointly with the EBA — the issuer is on a continuous-disclosure cycle. White paper updates are required when any material element changes. Periodic reserve attestations — composition, valuation, custody — are published on a recurring basis. Outstanding token supply, redemption volumes, and any breach of liquidity thresholds are reported to the NCA on the cadence the supervisor sets.

Crucially, the disclosure layer is what the EBA uses to monitor whether a given EMT is approaching the significance thresholds of Article 43 — the point at which direct EBA supervision and an own-funds uplift kick in. An issuer that under-builds Layer 4 is functionally invisible to its own significance trajectory and can find itself reclassified retroactively.

The table is the single best diagnostic for a board reviewing an EMT plan. If the team has scoped the project on a CASP timeline and CASP capital, the plan is mis-scoped by an order of magnitude. The structural fact that issuers and service providers sit in different titles drives everything downstream — including the supervisor that will be in the room.

Article 43 of MiCA — applied to EMTs via Article 56 — sets out the quantitative criteria for classifying an EMT as "significant". Significance triggers two consequences: direct EBA supervision alongside the home NCA, and an uplift in own-funds requirements from the baseline 2% of reserve to 3% of the average amount of reserve assets. The thresholds were calibrated by the EBA⁴^[4] and apply cumulatively — an issuer that meets any three of the criteria is presumptively significant.

Significance is not a fixed entry condition — it is a dynamic supervisory state. An EMT that launches under home-NCA supervision can be reclassified as significant mid-life if it grows into the thresholds, at which point the EBA takes over the supervisory dialogue. Our companion piece on EBA significance designation walks the full pipeline including the EBA-side review steps. The practical guidance for a board is: model the significance trajectory before launch, and pre-build the 3% own-funds capacity into the funding plan, rather than scrambling for it after the EBA letter lands.

Article 48 names two eligible issuer types and the practical choice between them shapes the whole project. The credit-institution route — full banking licence under CRR/CRD — gives the issuer access to central-bank deposit facilities, a clean Tier-1 capital base, and the most defensible reserve-custody architecture. It is also the slowest, most expensive, and most operationally demanding path: a banking licence is typically a 24–36 month project and pulls in CRR own funds, ICAAP, ILAAP, and recovery-and-resolution planning.

The EMI route — authorisation as an electronic money institution under EMD2, with the MiCA Title III overlay — is the route most non-bank issuers take. It is operationally lighter than a banking licence, the minimum initial capital is €350,000, and our companion piece on EMI safeguarding architecture walks the underlying segregation regime in detail. The trade-off is that the EMI cannot self-custody the reserve — it must place the reserve with a credit institution or other authorised custodian, introducing a counterparty layer that the credit-institution route avoids.

The decision turns on three factors: token scale ambition (a multi-billion-EMT issuer typically converges on the credit-institution route eventually because of central-bank-access economics), the founding team's prudential bandwidth, and the time-to-market constraint. Most 2026 EMT projects launch on the EMI route and migrate to credit-institution status only after significance designation. The MiCA-compliance companion guide for CASPs covers the parallel question on the service-provider side for issuers that also intend to operate exchange or custody services.

Three operational artefacts dominate the post-authorisation reality. First, the white paper — a living document, not a launch artefact. Article 51 requires notification ≥ 40 working days before any offer, and material changes trigger update obligations. The white paper carries the token's economic, technical, governance, reserve, and redemption disclosures and is the single most-scrutinised document the issuer publishes.

Second, the reserve attestation cadence. The EBA RTS framework expects regular third-party attestation of reserve composition, valuation, and segregation. Practitioner cadence is converging on monthly attestations for live EMTs, with quarterly deeper audits, and ad-hoc reporting on any breach of concentration or liquidity thresholds. The audit-firm bench in the EEA for EMT reserve attestation is narrow — building the auditor relationship is a Layer-4 prerequisite that founders consistently underestimate.

Third, the redemption SLA. Article 39 sets the legal right; operationally the issuer needs intraday liquidity arrangements with the reserve custodian, automated reconciliation between on-chain redemption requests and off-chain reserve drawdowns, and a stress-tested process for handling concentrated redemption days. Practitioner targets are T+0 for in-window requests and T+1 worst-case, with a documented backstop process if the same-day-liquid sleeve is exhausted.

A fourth, often-overlooked operational layer: the interaction with secondary-market venues. An EMT admitted to trading on regulated CASP venues lives in a market the issuer does not control. The issuer is on the hook for the white paper and the reserve, but the venue determines order-flow patterns, the on-chain wallet topology, and the bridges between centralised and decentralised liquidity. The supervisory expectation is that the issuer maintains a market-integrity dialogue with each admitted venue — covering reference-price disclosure, suspension protocols, and incident communication — even though MiCA does not impose direct issuer-side market surveillance the way it does for trading-platform CASPs.

Governance is the throughline. Board-level reserve oversight, a designated reserve-management committee, segregation of duties between token-issuance and reserve-investment functions, and an internal audit cycle that touches every layer of the stack at least annually are now table stakes for the EBA's significance-readiness expectations. An EMT issuer that runs the build like a fintech sprint will trip on the prudential expectations; an issuer that runs it like a small bank will not.

EMT issuance does not happen in an AMLR vacuum. The EU AML package — and our AMLR 12-month readiness roadmap covers the broader compliance build — applies to the issuer as a regulated obliged entity. KYC at primary distribution, Travel Rule on inter-CASP transfers, and sanctions screening across the secondary market are operational layers that sit alongside Layers 1–4 of the issuance stack rather than inside any of them. An EMT project that scoped the AML overlay as "covered by the EMI licence" will find AMLR-specific requirements added on top, and should plan accordingly.

You need both. The EMI authorisation under EMD2 is the underlying regulated status; MiCA Title III layers on top by requiring a white-paper notification to the home-state NCA at least 40 working days before offer, plus ongoing compliance with the Article 36 reserve, Article 39 redemption, and the EBA RTS framework. The EMI licence alone is not sufficient to issue an EMT into the EEA market.

No. Article 48 limits EMT issuance to credit institutions and authorised EMIs. A Title V CASP authorisation does not, on its own, authorise issuance. A group can hold both — many integrated platforms have a CASP entity for exchange / custody services and a separate EMI entity for the EMT — but the two regulated statuses are distinct and the issuance activity must sit in the EMI or credit-institution leg.

Two consequences. The EBA assumes direct supervision of the issuer in parallel with your home NCA, and your own-funds requirement steps up from the baseline 2% of average reserve assets to 3%. You also become subject to enhanced reporting and stress-testing obligations defined by the EBA. The designation is dynamic — an issuer can become significant mid-life and should model the trajectory before launch.

The reserve sits in highly liquid, low-risk instruments, which in the 2026 rate environment do earn yield — but the regime is built around capital preservation and redemption-funding, not yield optimisation. Yield from the reserve accrues to the issuer subject to MiCA's prohibition on interest on the EMT itself (Article 50): the issuer cannot pay interest to EMT holders, because doing so would convert the instrument into something economically closer to a deposit or a money-market fund unit. Reserve yield economics for the issuer therefore differ structurally from EMI Method 1(b) yield economics on customer e-money funds.

For a green-field project on the EMI route: typically 18–30 months from incorporation to live token. The EMI authorisation itself runs 9–18 months depending on jurisdiction; the MiCA Title III overlay (white paper, reserve and custody build-out, attestation framework, NCA notification window) adds another 6–12 months in parallel and sequence. The credit-institution route adds 12–24 months on top.

• EBA Significance Designation under MiCA Articles 43 & 56: the pipeline by which an EMT or ART crosses the thresholds and becomes EBA-supervised.

• MiCA Compliance Guide for CASPs: the Title V counterpart — full authorisation walkthrough for EEA crypto-asset service providers.

• EMI Safeguarding Architecture: the EMD2 segregation regime that the EMT reserve sits on top of.

• AMLR 12-Month Readiness Roadmap: the AML obligations overlay that applies to EMT issuers as obliged entities.

• Regulatory Legal Opinions: legal-opinion support for the white-paper, reserve, and redemption architecture submissions.

The EMT issuance stack will keep evolving. The EBA's RTS framework on reserve liquidity, the ESMA-led disclosure technical standards, and the PSD3 / PSR conclusion of the EMI-licensing reform are all in motion through 2026 and into 2027. The structural fact will not change: EMTs are issued, not provided, and the projects that succeed are the ones that scope the four-layer build from day one rather than discovering it six months in.