Many crypto founders enter the EU market thinking they need to find out whether a “staking license” exists. That is usually the wrong starting point. Under MiCA, the real issue is not the product label but the legal nature of the service. The regime applies to crypto-asset service providers from 30 December 2024, while transitional rules may allow some pre-existing providers to continue until 1 July 2026 or until an authorisation decision is made, whichever comes first.
So the better question is this: are you merely giving users access to a protocol, or are you actually providing a regulated service involving client assets, control, safekeeping, or return obligations? That is where MiCA starts to bite.
The EU does not issue a standalone “staking licence”
MiCA does not create a separate licence category called staking. That neat little fantasy folder does not exist. But ESMA’s published Q&A makes the position much clearer for real businesses: when a provider offers staking services and holds the client’s crypto-assets or the private keys giving access to them, that service is treated as ancillary to custody and administration of crypto-assets on behalf of clients. ESMA states that such staking services therefore require MiCA authorisation for custody and administration.
That means the legal analysis turns on control, not branding. A page that says “earn rewards” does not magically turn a custodial model into a neutral software tool. Regulators will look at who actually controls the wallet environment, who can move assets, who handles validator logic, and who remains responsible if something goes sideways.
Why custodial staking usually falls inside the CASP perimeter
MiCA’s custody rules are not vague wallpaper. The regulation requires providers offering custody and administration on behalf of clients to maintain position registers, establish custody policies, segregate client holdings, ensure return procedures, and bear liability for losses attributable to them. Those are not decorative compliance ornaments; they are operational obligations attached to a regulated service.
That is why staking-as-a-service usually becomes a CASP question when the provider:
- receives customer assets into wallets it controls,
- retains the private keys or another form of effective control,
- manages the staking flow on the client’s behalf,
- or contractually undertakes to return the assets after the service is performed.
At that point, the model is no longer just “access to network rewards.” It starts looking like a custody-based business with additional operational and legal consequences.
“Yield services” are broader — and more dangerous to classify lazily
The word yield is much messier than staking. Some yield products are essentially custodial staking with prettier packaging. Others rely on lending, treasury deployment, rehypothecation, collateral strategies, or a mixture of several revenue sources hidden behind a single button in the app.
That distinction matters because the EBA-ESMA joint report of January 2025 analyses staking, lending, and borrowing as separate crypto business models with different risks. The report is analytical rather than a new rulebook, but it reinforces a key point: not every product marketed as “yield” fits neatly into one MiCA box, and lending-style structures raise a different set of legal and prudential questions.
So when a company says it offers “yield,” that tells a regulator almost nothing useful. The real questions are more awkward and more important:
Where does the return come from?
Who takes counterparty risk?
Are customer assets staked, lent, pooled, pledged, or internally deployed?
Who absorbs losses if the strategy fails?
And what exactly has the client agreed to?
That is the difference between a regulated product analysis and a marketing slogan in a blazer.
Four models that often get mixed together
1. Straight custodial staking
This is the cleanest case. The customer transfers crypto to the platform, the platform stakes it, and rewards are distributed under the platform’s terms. Where the provider keeps custody or controls access, ESMA’s position strongly points toward CASP authorisation for custody and administration.
2. Assisted or delegated staking with limited control
Some businesses only provide routing, interface, validator selection tools, or support layers while the client keeps direct control of the assets. These models require a closer perimeter review because the answer depends on actual control, not on how “non-custodial” the homepage claims to be.
3. Yield generated through lending or onward deployment
Once the return comes from lending, internal treasury strategies, collateral chains, or third-party exposure, the analysis becomes more complex. The 2025 EBA-ESMA report specifically examines crypto lending, borrowing, and staking as distinct activities and highlights different risk patterns across them.
4. Structured return products
Sometimes a platform wraps several mechanics into one yield product and presents it as a simple savings-style feature. Legally, that is where trouble starts. The more layers of exposure, discretion, and embedded rights the product contains, the less useful it is to call it “just yield.”
What regulators will actually care about
A licensing review for staking or yield services is not won by sprinkling legal buzzwords over a flowchart and hoping the file becomes majestic. Authorities will want a precise explanation of how the model works in practice.
That means the application or legal analysis should clearly show:
- how customer assets are received,
- whether assets are pooled or segregated,
- who controls keys or access credentials,
- how rewards are generated,
- whether losses such as slashing can occur,
- whether assets are ever exposed to third parties,
- how the assets are returned,
- and what the customer has expressly accepted.
These points are not abstract theory. MiCA’s custody framework expressly requires client protection mechanisms such as segregation, return arrangements, custody policies, and liability standards for losses attributable to the provider.
Why founders get this wrong so often
The usual mistake is not a lack of enthusiasm. It is a lack of product decomposition.
A company builds one wallet. Inside that wallet it combines staking rewards, internal asset deployment, third-party exposure, fee spreads, and maybe some lending. Then it calls the whole thing “earn.” From a sales perspective, that sounds smooth. From a regulatory perspective, that is a clown car full of classification problems.
Another common error is assuming that a technical architecture automatically answers the legal question. It does not. The key issue is not whether the system uses smart contracts, validators, dashboards, or fancy diagrams with arrows. The issue is who controls what, who owes what to the client, and which regulated service is actually being delivered.
The practical takeaway
For a business planning to offer staking in the EU, the safest starting assumption is this: if you hold client assets or control the means of access to them while providing staking for clients, you are very likely looking at CASP authorisation for custody and administration under MiCA. ESMA’s Q&A is fairly direct on that point.
For yield services, the answer is less tidy. Some products are simply staking in nicer clothes. Others are part staking, part lending, part treasury management, and part future headache. Those models need to be mapped service by service, flow by flow, risk by risk. Otherwise the legal analysis ends up being built on vapor, and vapor is a terrible foundation for regulated business.
Conclusion
A CASP strategy for staking or yield services should begin with substance, not labels. MiCA does not hand out a separate staking licence, but it does capture custody-based staking models through the rules on custody and administration of crypto-assets on behalf of clients. At the same time, the broader “yield” category can cover very different structures, some of which go well beyond a simple staking analysis.
That is why the smart move is not to ask, “Can we call this a staking product?” The smart move is to ask, “What exactly are we doing with client assets, and which authorisation scope does that trigger?” One question leads to a licence strategy. The other leads to expensive surprises.
FAQ
Is there a separate staking licence under MiCA?
No. MiCA does not create a standalone staking licence. ESMA’s published position is that where a provider stakes client assets while holding the assets or the private keys, the service requires authorisation for custody and administration of crypto-assets on behalf of clients.
Does every yield product require a CASP licence?
Not automatically. “Yield” is a broad commercial label that may cover staking, lending, borrowing, or more complex structures. The legal analysis depends on how the return is generated and what control or exposure sits beneath the product.
When did MiCA start applying to CASPs?
MiCA applies from 30 December 2024, with transitional provisions for some providers that were already operating under applicable national law before that date.
Why is custody so important for staking services?
Because MiCA’s custody regime imposes concrete obligations around safekeeping, segregation, return of assets, and liability. Once a staking provider holds client assets or controls access to them, those rules become highly relevant.
What should be reviewed before launch?
Asset flow, wallet control, validator setup, client agreement structure, return mechanics, segregation logic, and any use of lending or third-party exposure should all be reviewed before the product is offered. Those points determine whether the model sits comfortably inside the intended authorisation scope.
Need Help Assessing Staking Under MiCA?
