Draw the plumbing and something odd falls out.
PayPal, M0 and MoonPay went live on September 9 with PYUSDx, a platform that lets any business issue its own branded stablecoin without touching reserves, custody or redemption infrastructure. Three issuers launched with it. More than $100 million already processed between them. The pitch from all three partners is that the product layer should belong to whoever is building the product, and the monetary plumbing should belong to people who do plumbing. That is a good pitch and a sensible product.
Now draw it.
$PYUSD is issued by Paxos, a federally regulated national banking association, backed by dollar deposits and Treasuries. Fine. Compliant. Boring, in the way a reserve asset should be.
PYUSDx tokens are issued by MoonPay Digital Assets Limited. Their reserve asset is $PYUSD.
So the thing backing the second token is the first token. The $GENIUS Act, signed in July 2025 and still being turned into regulations, tells you who may issue a payment stablecoin and what has to sit behind it. Cash. Insured deposits. Short-dated Treasuries. Repos against Treasuries. Money market funds holding those.
It does not say anything about a stablecoin backed by a stablecoin, issued by someone else entirely. Nobody covering the launch has asked about it. It is worth asking now, while the rules are being drafted, instead of in eighteen months when they are not.
What was actually built
Three companies, three jobs. The details matter because they decide who is on the hook for what.
M0 supplies the infrastructure. Its platform lets an issuer configure individual components of a stablecoin instead of accepting a fixed model: token name, access restrictions, reward distribution, collateral policy, and cross-chain availability are all set by the issuer. M0’s chief executive has described the design intent as making the product layer belong to the builder, and the company also works with Stripe-owned Bridge and with regulated custody firms.
MoonPay issues the tokens and holds the backing. MoonPay Digital Assets Limited is the issuing entity for PYUSDx-layer tokens and holds the $PYUSD that backs them. It also contributes onboarding and distribution.
JUST IN: PYUSDx now has $100 million in stablecoin assets under management
— crypto.news (@cryptodotnews) September 10, 2026
The PayPal-linked launchpad is running on MoonPay pic.twitter.com/Zxi0w9a3y9
PayPal supplies $PYUSD. The underlying stablecoin remains a Paxos-issued product reserved with dollar deposits, Treasuries, and similar cash equivalents. PayPal’s role is supplying the asset that sits underneath and the ecosystem connection.
The disclaimers are specific and worth reading. Tokens created on PYUSDx are not PayPal or Paxos products. They cannot be sent, received, or used inside the PayPal and Venmo applications. That is an unusual carve-out for a platform named after a company’s own stablecoin, and it tells you the partners have thought carefully about where liability sits.
At launch: Saturn, Concrete, and Cap, with more than $100 million in combined processed volume. Cap migrated part of its cUSD onto PYUSDx so that a portion of its covered-credit float would rest on $PYUSD instead of more volatile decentralised finance liquidity, which is a sensible use of the product and the clearest illustration of what it is for. USD.AI and Fairblock are next.
The two-layer question
Nothing here accuses anyone of anything. This is a question the statute has not answered, asked while there is still time to answer it.
The $GENIUS Act, enacted July 18, 2025, restricts issuance of payment stablecoins to permitted issuers across four routes and requires reserves backing outstanding tokens one to one in specified high-quality liquid assets: currency, insured deposits, short-dated Treasury bills, Treasury-collateralised repurchase agreements, and money market funds holding those instruments. Our dedicated page on the law sets out the framework in full.
$PYUSD fits that cleanly. Paxos is a federally regulated national banking association, the reserves are cash and Treasuries, and the disclosure obligations apply.
PYUSDx tokens are a different object. They are issued by a separate entity, and their reserve asset is $PYUSD, not the asset classes the statute names. Three questions follow and none has a public answer.
Is a PYUSDx token a payment stablecoin? The statutory definition captures a digital asset used for payment or settlement, redeemable at a fixed monetary value, whose issuer represents it will maintain stable value. An application-specific dollar token used inside a credit product appears to meet that description.
If it is, who is the permitted issuer? The entity issuing it is MoonPay Digital Assets Limited, not Paxos. Permitted status attaches to issuers, not to reserve assets, and the four routes to permitted status all describe entities, not backing arrangements.
Does $PYUSD count as a permitted reserve asset? The named list does not include other stablecoins. Whether a token fully backed by a compliant stablecoin satisfies a one-to-one reserve requirement is a reasonable reading and it is not the reading the text supplies on its face.
None of this suggests anyone is doing anything improper. The Act does not take effect until the earlier of January 18, 2027 or 120 days after final implementing regulations, and the agencies missed their one-year rulemaking deadline in July 2026 with proposals issued and final rules outstanding. Building a product during that window is entirely legitimate. The point is narrower: the rules that will govern this structure are being written now, and this structure is not one the drafters obviously had in mind.
Why anyone would build it this way
The commercial logic is genuinely good, which is why this structure will spread whatever the regulators decide.
Reserves are the hard part. Issuing a compliant stablecoin means holding, custodying, and reporting on cash and Treasuries, contracting an accounting firm for monthly attestation, and building redemption infrastructure. That is a bank-adjacent operation with bank-adjacent costs, and it is completely disproportionate for a company that wants a branded dollar inside its own application.
The alternative was worse. Before platforms like this, a business wanting an application-specific dollar either built the whole stack, partnered bilaterally with an issuer on bespoke terms, or used an existing stablecoin and accepted no control over its properties. All three are bad options for a small team.
Configurability is the product. Access restrictions, reward distribution, collateral policy, and cross-chain availability set per issuer is a different offering from a single stablecoin with fixed properties. Cap’s use case, resting covered-credit float on $PYUSD instead of volatile decentralised finance liquidity, is exactly the kind of thing that needs configuration and not a generic token.
And for PayPal it solves a distribution problem. $PYUSD sits around $2.81 billion, eighth in a stablecoin market near $305 billion where Tether holds roughly 60%. Growing that through direct payments means competing with incumbents on their own ground. Growing it as a reserve asset for other people’s tokens means every PYUSDx issuer that scales needs more $PYUSD behind it, expanding the footprint without PayPal operating any of those applications. That is a second lever on demand and a considerably cheaper one.
What the layering actually adds
Two layers is better than one in one respect and worse in another.
On the positive side, the backing asset is a regulated, attested, cash-and-Treasuries stablecoin instead of an ad hoc reserve. An application dollar backed by $PYUSD is substantially better collateralised than one backed by a decentralised finance yield strategy, which is precisely why Cap moved. Layering onto a compliant base is a meaningful improvement over the alternatives that existed before.
On the risk side, a holder of a PYUSDx token now depends on two entities instead of one. The issuer must hold the $PYUSD it claims to hold and honour redemption. Paxos must maintain $PYUSD’s peg and reserves. A failure at either level reaches the holder, and the holder’s legal relationship is with the upper entity, not the lower one, which is what the disclaimer about these not being PayPal or Paxos products makes explicit.
That second point deserves emphasis because it is the practical consequence of the structure. The name on the platform is PayPal’s. The underlying asset is PayPal’s stablecoin. The token in a user’s wallet is neither, and cannot be used in PayPal’s own applications. A user who does not read the documentation could reasonably form the wrong impression about whose obligation they hold.
The precedent this sets
This is not really a PayPal story. It is what happens when issuing a dollar stops being a product and becomes plumbing.
The pattern is familiar from other financial layers. Card networks did not issue cards; they let banks issue them on shared rails. Payment processors do not hold funds; they let merchants transact on shared infrastructure. In both cases the layer underneath became more valuable as the layer above proliferated, and the entity operating the rails captured economics from activity it did not conduct.
M0’s chief executive framed the fragmentation problem directly, noting that as more financial institutions get involved the landscape fragments, and that most of those institutions do not know how to engage with developers, so the middle layer abstracts the complexity. That is a rails argument, and it is the correct one.
NEW: $GENIUS Act stablecoin rules deadline missed by U.S. regulators
— crypto.news (@cryptodotnews) July 19, 2026
The legislation will still activate in January 2027 despite incomplete finalization pic.twitter.com/t3JdMON7xJ
What follows, if the model works, is a large number of application-specific dollars backed by a small number of compliant base stablecoins. That concentrates systemic importance in the base layer while distributing the customer relationships across hundreds of issuers, which is a structure regulators generally find difficult, because supervision attaches to entities and the entity holding the reserves is not the entity facing the customer.
Our stablecoin status page sets out the three gaps the $GENIUS framework left open, and this sits squarely in the space between two of them.
The three issuers, and what they reveal
Three issuers is a small enough list to go through one by one, and they tell you more than the press release does.
Cap is the clearest case. It migrated a portion of its cUSD onto PYUSDx so that part of its covered-credit float would rest on $PYUSD instead of more volatile decentralised finance liquidity. That is a treasury decision: a credit product needs its float in something stable, and a regulated, attested, cash-and-Treasuries-backed stablecoin is materially better collateral than a yield-bearing position in a lending protocol. Cap did not want a branded token for marketing. It wanted better backing for an existing liability.
Saturn and Concrete have been named without the same public detail, though the three together account for the more than $100 million in processed volume the partners cite. MoonPay’s executive has argued the figure matters because the same issuance stack is already supporting credit, vault, and Bitcoin-linked products, which suggests three different applications instead of three variations on one.
JUST IN: PayPal’s $PYUSD stablecoin supply shrinks 31% since March ATH of $4.2B. Now stands at $2.92B, a loss of more than $1 billion pic.twitter.com/vMWPcDSjdU
— crypto.news (@cryptodotnews) June 6, 2026
USD.AI and Fairblock are next, with no announced timing.
Two observations follow. First, the early adopters are crypto-native firms building financial products, not consumer brands wanting a loyalty token. That is a more demanding customer set and a better signal, because a credit protocol choosing your stablecoin as its float has done diligence a marketing department would not.
Second, the $100 million figure is processed volume, not market capitalisation and not revenue. The distinction is worth holding, since processed volume measures throughput over a period and says nothing about how much of the token is outstanding at any moment. A payments product cycling the same dollar repeatedly produces a large volume figure and a small float. Both are real; they measure different things, and only one of them determines how much $PYUSD sits in reserve.
The redemption chain
The question that matters if you hold one of these: what happens when you want actual dollars back. The answer goes through two companies, in order.
A holder of a PYUSDx token redeems with the issuing entity, MoonPay Digital Assets Limited, which holds $PYUSD as backing. To deliver actual dollars, that $PYUSD must itself be redeemed with Paxos, which holds the cash and Treasuries. So a full redemption to bank money traverses two independent obligations, each with its own terms, timing, and operational capacity.
In ordinary conditions this is invisible and fast. Stablecoin redemption at both layers is routine, and the whole point of building on a regulated base is that the lower layer is dependable.
The interesting case is the stressed one, and it has a specific shape. If a large number of holders redeem simultaneously, the upper issuer must convert $PYUSD to dollars at the same time its own customers are converting tokens to $PYUSD. Those are sequential operations, and the second one is not under the upper issuer’s control. Paxos’s redemption capacity and terms become the binding constraint on a token whose holders have no relationship with Paxos.
None of this is unique to PYUSDx and none of it suggests a defect. Every layered financial structure works this way, and layering onto a well-reserved base is precisely what makes it safer than the alternatives. But it is the reason the disclosure that these are not PayPal or Paxos products is doing real work, not lawyerly throat-clearing. A holder’s claim runs to the entity that issued their token, and that entity’s ability to pay depends on an entity the holder cannot call.
The practical instruction for anyone evaluating one of these tokens is to read the upper issuer’s redemption terms specifically, since those are the terms that bind, and to understand that the quality of the backing asset and the reliability of the redemption path are two separate questions with two separate answers.
What a regulator would ask
Forget whether it is permitted. Here is what an examiner would ask, and every one of these is answerable today.
Is the backing segregated and verifiable? The upper issuer holds $PYUSD backing its tokens. Whether that $PYUSD sits in an identifiable, segregated arrangement, whether it is attested to on any cadence, and whether the reporting is public are the first questions in any reserve examination. M0’s platform advertises support for on-chain reporting and reserve validation, which is the right capability; whether each issuer uses it is a separate matter.
Who bears redemption obligation and under what terms? The disclaimers make clear the tokens are not PayPal or Paxos products, which answers the question negatively for two parties without answering it positively for the third. Published redemption terms from the issuing entity would.
What happens on issuer failure? The $GENIUS Act gives holders of permitted payment stablecoins a priority claim in insolvency ranking above administrative expenses. Whether a holder of a token backed by such a stablecoin, issued by an entity that may not itself be a permitted issuer, inherits any comparable protection is unresolved and is the single most consequential open question for a holder.
Which jurisdiction supervises the issuing entity? MoonPay Digital Assets Limited is the named issuer. Its regulatory status and home jurisdiction determine which authority examines it and under what standard, and that is a fact rather than a judgment call.
None of these are difficult to answer and none is commercially sensitive. That they are not currently prominent in the launch materials is unremarkable for a product three days old, and it is also the gap between a product announcement and the disclosure a supervised financial instrument eventually requires.
The broader point for readers tracking stablecoin regulation is that this is where the next round of rulemaking pressure will land. The first round addressed who may issue a dollar. The obvious second question, once platforms like this proliferate, is who may issue a claim on someone else’s dollar, and the answer is not in the statute.
Who actually owes you money
Worth being blunt about this, because the branding and the obligation point at different companies.
The platform is called PYUSDx. The backing asset is PayPal’s stablecoin. PayPal’s name is on the announcement. And if the token in your wallet fails, your claim is against MoonPay Digital Assets Limited, an entity most holders will never have heard of and cannot call.
The partners say this clearly. Tokens created on the platform are not PayPal or Paxos products. They cannot be sent, received or used inside PayPal or Venmo. Read that second sentence again, because it is genuinely strange: a token backed by PayPal’s dollar, launched on a platform carrying PayPal’s name, that PayPal’s own applications will not accept.
That is not sloppiness. It is a boundary drawn on purpose, and it is drawn to keep liability where the issuing entity is. Which is fine, correct even, and also exactly the kind of thing a user skims past when the logo at the top says PayPal.
The general lesson travels beyond this product. In any layered financial arrangement, the recognisable brand and the counterparty are frequently not the same entity, and the gap between them is where retail confusion lives. Card networks, white-labelled banking, payment facilitators, and now stablecoin issuance platforms all have this shape. The name sells it. Someone else owes you.
If you are evaluating one of these tokens, the only question that matters is which legal entity issued it and what its redemption terms say. Everything else on the page is marketing.
What breaks first
Every new financial structure has a most-likely failure mode, and it is usually not the one the launch coverage worries about.
For PYUSDx the risk is not $PYUSD depegging. Paxos runs a regulated, attested, cash-and-Treasuries reserve, and that is about as solid as this asset class gets. If the base layer goes, the problem is considerably larger than one platform.
The realistic failure is at the upper layer and it is mundane. An issuer scales faster than its operational capacity. Redemption requests arrive in a cluster. The issuer holds the $PYUSD it says it holds, but converting it to dollars at speed depends on Paxos’s redemption process, which the issuer does not control and which was built for a different volume profile. Nothing is insolvent. Everything is slow. And slow, in a product marketed as a dollar, looks identical to broken from the outside.
The second realistic failure is configuration. M0’s platform lets issuers set access restrictions, reward distribution and collateral policy individually. Flexibility is the selling point, and flexibility means a hundred issuers making a hundred different decisions about parameters that determine whether their token behaves like a dollar under stress. Some of those decisions will be wrong. The base asset being sound does not save a token whose issuer configured redemption badly.
Neither of those is a reason not to build this. They are a reason to read the specific issuer’s terms rather than the platform’s, which almost nobody does, and which is the entire practical takeaway from every layered financial product ever launched.
What to watch
Whether final rules address layering. The OCC and FDIC proposals are drafted and comment periods have run. Whether the final text addresses tokens backed by other stablecoins is the single most consequential detail for this structure, and it is answerable within months.
Whether any PYUSDx issuer seeks permitted status. If the answer to the layering question is that the upper issuer needs its own permitted status, the economics of the platform change substantially. Watch for applications.
$PYUSD’s supply against PYUSDx growth. The indirect demand mechanism is testable. If Saturn, Concrete, Cap, and their successors scale, $PYUSD outstanding should grow to back them. The partners have published no targets, so the correlation is the only available evidence.
Whether the carve-out holds. PYUSDx tokens currently cannot be used inside PayPal and Venmo. If that changes, the liability and regulatory picture changes with it, because the distance the disclaimers create would narrow.
Who else launches one. Stripe’s Bridge works with the same infrastructure provider. A competing platform from another payments incumbent would confirm that this is the direction of the category and not one company’s experiment.
Disclaimer: This article is for information and educational purposes only and does not constitute financial, investment, or legal advice. It describes a recently launched product and raises regulatory questions that have not been resolved by implementing rules, and nothing here alleges non-compliance by any party. Always do your own research. Information is accurate as of September 10, 2026.
What is PYUSDx?
A platform launched September 9 by PayPal, M0, and MoonPay that lets businesses issue their own application-specific stablecoins backed by PayPal USD. Issuers configure the token’s name, access restrictions, reward distribution, collateral policy, and cross-chain availability instead of accepting a fixed model. Three issuers went live at launch with more than $100 million in combined processed volume.
Who actually issues the tokens?
MoonPay Digital Assets Limited issues PYUSDx-layer tokens and holds the $PYUSD backing them. Paxos Trust Company separately issues the underlying $PYUSD, reserved with dollar deposits and Treasuries. Tokens created on PYUSDx are not PayPal or Paxos products, and PayPal’s role is supplying $PYUSD and the ecosystem connection.
Can I use a PYUSDx token in PayPal or Venmo?
No. Tokens created on the platform cannot currently be sent, received, or used inside the PayPal and Venmo applications. That carve-out is stated explicitly by the partners and is worth noting, because the platform carries PayPal’s name and is backed by PayPal’s stablecoin while the tokens themselves are neither.
How does this interact with the $GENIUS Act?
That is the open question. The Act requires permitted issuers to back payment stablecoins one to one in named high-quality liquid assets: currency, insured deposits, short-dated Treasuries, Treasury-collateralised repos, and money market funds holding those. A token issued by a separate entity and backed by another stablecoin is not obviously described by that framework, and final implementing rules have not been issued.
Does that mean PYUSDx is non-compliant?
No, and nothing here suggests it. The Act takes effect on the earlier of January 18, 2027 or 120 days after final rules, and the agencies missed their one-year rulemaking deadline in July 2026. Building during that window is legitimate. The point is that the rules governing this structure are being written now and were not obviously drafted with it in mind.
Why would a business want its own stablecoin?
Control and configuration. Issuing a compliant stablecoin independently requires holding and reporting on cash and Treasuries, contracting monthly attestation, and building redemption infrastructure, which is disproportionate for a company that wants a branded dollar inside its own application. Cap’s use case, resting part of its covered-credit float on $PYUSD rather than volatile decentralised finance liquidity, illustrates the appeal.
What does PayPal get out of it?
A second lever on $PYUSD demand. $PYUSD sits around $2.81 billion in a stablecoin market near $305 billion where Tether holds roughly 60%. Every PYUSDx issuer that scales needs more $PYUSD behind its token, expanding $PYUSD’s footprint without PayPal operating those applications. The partners have published no issuance or reserve targets.
What is the risk to a holder?
Dependence on two entities instead of one. The PYUSDx issuer must hold the $PYUSD it claims and honour redemption, and Paxos must maintain $PYUSD’s reserves and peg. A holder’s legal relationship is with the upper issuer, not with PayPal or Paxos, which is what the disclaimers make explicit. This is educational analysis, not investment advice.