PYUSD is a payments-native stablecoin with a centralized control plane

PYUSD is built to behave like a dollar that can move through PayPal’s consumer and merchant stack, then exit to public chains when you want composability. PayPal’s own product framing is explicit: it targets payments and transfers, and it is redeemable at $1.00 per PYUSD inside PayPal and Venmo.

Issuance sits with Paxos, not PayPal. PayPal’s help documentation states that PayPal USD is issued by Paxos on multiple blockchains and is fully backed by U.S. dollar deposits, U.S. Treasuries, and similar cash equivalents.

The microstructure point that matters is simple. PYUSD is not “decentralized money” with an emergent monetary policy. It is a tightly administered liability with on-chain rails. Our tokenomics methodology focuses less on narratives and more on liquidity gates: who can mint, who can burn, who can redeem at par, and how quickly inventory can migrate across venues when demand spikes or risk sentiment flips.

Supply mechanics: elastic float, no emissions schedule, no max cap

CoinGecko currently shows PYUSD’s circulating supply and total supply as 4,203,027,232 PYUSD and the max supply as ∞ (unbounded). That is consistent with what you want from a fiat-backed stablecoin: supply expands and contracts based on issuance and redemption, not a pre-committed emission curve. For a contrast with a non-fiat, issuer-administered asset token, compare this model with our PAX Gold review.

On the Paxos side, the legal structure is explicit about one-for-one backing and segregation. Paxos’ stablecoin terms state that for every USD stablecoin issued and in circulation, Paxos will hold either one U.S. dollar or an equivalent amount of permitted USD-denominated assets in segregated accounts for the benefit of token holders.

Redemption is also spelled out as a mechanical rule, not a “best effort peg.” Paxos’ terms state that all redemptions are one-for-one (regardless of any change in the market value of reserve assets), and that stablecoin amounts are rounded down to the nearest U.S. cent at redemption.

That sounds like a clean model. The market-structure caveat is that “supply” is not the same as “tradable float.” In stablecoins, the effective float is whatever inventory is sitting in venues where two-sided markets actually clear, and whatever inventory is held by actors who can do fast basis trades against redemption or trusted OTC convertibility. In practice, this is where stablecoins can feel stable for months, then gap a few basis points when a single venue de-lists, pauses deposits, or sees a sharp inventory imbalance.

Where PYUSD exists on-chain (and why chain fragmentation is a liquidity event)

PayPal’s help documentation lists issuance on Ethereum, Solana, and Arbitrum.

PayPal also publishes the canonical token identifiers for end users verifying authenticity. On March 5, 2026, PayPal’s help page lists:

PayPal announced PYUSD availability on Solana on May 29, 2024. That was a structural change because Solana shifts the cost and latency profile for transfers and market-making. Lower settlement friction tends to tighten spreads, but it also makes it easier for inventory to stampede between venues when incentives change.

PayPal also announced plans to make PYUSD available on Stellar, pending NYDFS regulatory approval, on June 11, 2025. Even before it ships, the announcement matters because each new rail introduces a new set of liquidity pools, custody pathways, and potential dislocations between “canonical” and “wrapped” representations.

On the Paxos platform side, Paxos documentation for its dashboard workflow lists PYUSD supported networks as Ethereum, Solana, Arbitrum, and Stellar for minting.

For Stellar specifically, Paxos’ API reference states the PYUSD issuer address on Stellar mainnet as GDQE7IXJ4HUHV6RQHIUPRJSEZE4DRS5WY577O2FY6YQ5LVWZ7JZTU2V5 and the asset code as PYUSD.

My bias here is mechanical. Every time PYUSD expands to a new chain, you get a “soft unlock” event even though nothing is vesting. Inventory that was effectively trapped behind a UX, a custody policy, or a settlement cost suddenly becomes mobile. That changes where spreads compress, where basis trades concentrate, and which venues become price leaders during stress.

Smart-contract control plane: mint/burn roles, optional rate limits, pause, freeze, upgradeability

The PYUSD contract architecture is designed around explicit administrative levers. Paxos’ contract repository describes a separate SupplyControl contract that manages minting and burning via role-based access control. Only addresses with the SUPPLY_CONTROLLER_ROLE (managed by a SUPPLY_CONTROLLER_MANAGER_ROLE) can mint and burn. Supply controllers can optionally have rate limits that cap how many tokens can be minted over a given timeframe.

The same repo states Paxos can pause transfers and approvals in the event of a critical security threat, and that the ability to pause is controlled by a single owner role using the OpenZeppelin Ownable/Pausable pattern. It also states that while paused, the supply controller retains the ability to mint and burn.

The repo also describes an ASSET_PROTECTION_ROLE that can freeze and unfreeze balances of any address and can wipe a frozen address’s balance to allow authorities to seize backing assets.

From a market microstructure angle, this is not a footnote. Freeze and pause controls change the tail-risk distribution for holders and liquidity providers. If you are warehousing inventory in a venue, you are implicitly long the issuer’s operational competence and compliance gating, not just the dollar peg narrative.

PYUSD is also designed to be upgradeable. Paxos’ repo describes a proxy pattern where a proxy contract represents the token, and calls are delegated to an implementation contract, using an AdminUpgradeabilityProxy pattern.

Etherscan’s verified contract page for the Ethereum PYUSD address shows proxy-related admin and upgrade functions like upgradeTo, upgradeToAndCall, changeAdmin, and admin, consistent with an upgradeable proxy setup.

Finally, the repo states PYUSD implements EIP-3009 and EIP-2612 to support gas-less style authorization flows (transfer with authorization, and permit-based approvals). That feature matters because it compresses payment UX friction and can shift on-chain transfer velocity, which in turn changes how quickly liquidity shocks propagate.

Fees, rewards, and reserve economics: who gets paid (and who doesn’t)

Stablecoin tokenomics usually gets mis-modeled because analysts fixate on supply levels and ignore cashflows. For PYUSD, the important economic flows are off-chain: fees and spreads in PayPal’s product layer, and reserve yield net of operating and compliance costs.

PayPal’s help documentation states there are no fees to buy, sell, hold, or send PYUSD inside PayPal, while noting that conversion between PYUSD and other cryptocurrencies carries a fee, and that network fees may apply when sending PYUSD to external wallets on supported networks.

PayPal also describes a PYUSD rewards program where rewards accrue based on the account’s average daily PYUSD balance and are paid monthly in PYUSD, with a variable reward rate viewable in-app.

On Paxos’ side, the stablecoin terms are explicit that the stablecoins are not designed to create returns or accrue financial benefit to holders by default.

Paxos also states it may apply a fee to reserves backing USD stablecoins, while also stating that it will not impose a fee that reduces reserves below the amount of USD stablecoins outstanding.

For redemption, Paxos’ terms state it will not charge fees for redeeming USD stablecoins, while noting that banks and wallet providers may charge fees, and that Paxos may deduct banking fees charged to Paxos by financial institutions from transferred amounts.

Put together, the picture is clean: PYUSD is not an “incentive token.” There is no native on-chain burn that reflexively boosts holders. The economic engine is reserve management plus distribution and payment UX. That is fine. It also means the stability story is only as good as the redemption plumbing and the issuer’s operational discipline.

Liquidity plumbing: the peg is a market outcome, and redemption access is the main lever

In calm markets, PYUSD trades like other fiat-backed stablecoins. Convertibility at par inside PayPal and Venmo anchors retail expectations, and Paxos’ mint/redeem stack anchors professional arbitrage for accounts that can access it. PayPal’s launch materials emphasize that PYUSD can be bought or sold through PayPal at $1.00.

The stress-path is different. In PayPal’s June 11, 2025 disclosure, PayPal explicitly warns that while PYUSD is designed to maintain a stable value of one U.S. dollar per token, stability is not guaranteed in all circumstances. It also states that while PayPal, Paxos, and select partner account holders may be able to redeem PYUSD at one U.S. dollar, external self-custody wallet holders and third-party platform holders do not have guaranteed access to redemption at $1.

This single disclosure drives most of the real tokenomics. If redemption is segmented, then the peg for “everyone else” is a secondary-market peg. It is supported by liquidity providers and by whatever subset of players can warehouse inventory and reliably convert it back to dollars. During a shock, price can move a few bps because the marginal seller is forced to cross a spread without being able to redeem, while the marginal buyer demands compensation for operational and compliance risk.

Paxos’ own terms also make clear it can engage market makers to provide liquidity for USD stablecoin markets on Paxos and on third-party platforms. That is a direct admission that secondary liquidity is an engineered input, not a naturally guaranteed property.

There is another microstructure tension. PYUSD’s contract toolkit includes pauses, freezes, wipes, role-based supply control, and upgradeability. Those features are defensible for compliance-grade dollars. They also introduce jump risk for on-chain venues that cannot price “issuer discretion” continuously. When markets get nervous, liquidity pulls first. The peg breaks second.

Risk register

Top 3 risks

  1. Redemption-access segmentation (dominant). Trigger: a market-wide risk-off move, a venue-level suspension (deposits/withdrawals), or a policy tightening that reduces who can redeem at par. Mechanism: if external holders cannot reliably redeem at $1, they are forced into secondary-market sales, widening spreads and allowing PYUSD to trade below peg until arbitrage channels reopen or inventory rebalances. Who bears it: external wallet holders, DeFi LPs, and market makers carrying inventory when flows flip. Measurable indicators: persistent discounts to $1 on major venues, withdrawal/deposit status changes on large exchanges, and widening basis between PayPal’s internal convertibility and on-chain spot prices.

    Dominant risk: This is the risk that actually sets the ceiling on how “systemically stable” PYUSD can be in crypto-native markets. If you can redeem at par on demand, you can run a tight arbitrage book and keep price glued to $1 with modest balance sheet. If you cannot, then PYUSD becomes a credit-and-liquidity instrument whose value is mediated by intermediaries.

    PayPal’s own wording matters because it highlights a two-tier market. Tier one is PayPal, Paxos, and select partners with a path to $1 redemption. Tier two is everyone else, who may have to sell at market prices that can move “materially” above or below $1. That is not an academic nuance. It is the core reason stablecoins sometimes de-peg without any reserve impairment. The peg can fail as a microstructure event, not a solvency event.

    Chain expansion increases this risk surface. Each new chain and integration increases the number of venues that can become “inventory sinks” during a shock. If redemption remains gated, then cross-chain liquidity is not a single pool. It is a set of pools stitched by market makers, bridges, and custody workflows. Any seam can tear under stress.

    One practical mitigation is transparency and predictable operations. Paxos publishes monthly reserve reporting and attestations, with a clear distinction between self-reported portfolio composition and independently issued attestations. Paxos states it self-reports portfolio composition 5 business days after month-end, and that these reports are not independently reviewed. Paxos also states that attestations posted on or after February 28, 2025 are issued by KPMG LLP, while earlier attestations were issued by WithumSmith+Brown, PC. For related monitoring frameworks, see our crypto research reports.

    That helps modelability. It does not eliminate the dominant risk because the binding constraint is still convertibility access and how quickly it can be exercised in size when markets get ugly.

  2. Issuer-admin control and address-level intervention. Trigger: compliance action, court order, or a security incident that prompts issuer intervention. Mechanism: Paxos’ own contract documentation states an ASSET_PROTECTION_ROLE can freeze/unfreeze balances and wipe balances after freezing, and that Paxos can pause transfers and approvals via a single owner role. This can create sudden, discontinuous loss for affected addresses and second-order liquidity shocks if LPs or venues become cautious about holding inventory that could be frozen. Who bears it: sanctioned or mistakenly flagged users first, then any pool or venue that becomes exposed to frozen inventory paths. Measurable indicators: on-chain pause events, freeze/wipe-related contract calls, and changes in venue collateral eligibility.

  3. Upgradeability and operational error risk. Trigger: a contract upgrade, a misconfiguration in supply control, or an internal operational failure during mint/burn workflows. Mechanism: upgradeable proxies concentrate upgrade authority, and operational mistakes can create transient but severe shocks to perceived integrity even if corrected quickly. Who bears it: DeFi integrators, market makers, and any venue relying on automation keyed to totalSupply or on-chain event streams. Measurable indicators: unexpected supply jumps, proxy upgrade events, and emergency pauses. For example, Bloomberg reported that on October 15, 2025, Paxos mistakenly minted $300 trillion of PYUSD before burning the tokens minutes later.

If you are integrating PYUSD into a product, treat “peg risk” as a function of operational access, not just reserves. The biggest wins usually come from mapping convertibility paths, venue inventories, and failover routes, then stress-testing your flows against pauses, freezes, and withdrawal halts. If you need outside help, this is the narrow slice of tokenomics consulting that tends to pay for itself because it reduces surprise liquidity shocks in production.



This article is part of our Tokenomics Deep Dive series.