USDz in the Anzen stack

USDz is built as a composable, RWA-backed dollar that tries to behave like a stablecoin in DeFi while routing its economic engine through a private credit balance sheet. CoinGecko describes it as “an RWA-backed stablecoin” backed by “a portfolio of tokenized private credit assets.”

The key architectural choice is that USDz holders do not automatically earn the underlying yield. Anzen’s docs explicitly separate holding from “staking.” Holding USDz is yieldless. Staking USDz on Ethereum mainnet produces sUSDz, and sUSDz is the object that accrues rewards.

Collateral-wise, Anzen positions USDz as backed by a diversified private credit portfolio and underwritten with an RWA partner. The FAQ states those assets are underwritten in collaboration with a US licensed broker-dealer and points readers to an on-chain transparency view for reserves.

The on-chain representation of that backing is a permissioned token called SPCT (Secured Private Credit Token). Anzen’s transparency docs say USDz is minted via audited smart contracts and is backed 1:1 with SPCT.

Two immediate implications follow.

First, USDz is structurally closer to a “synthetic dollar with RWA collateral plumbing” than a fully redeemable money-market share token. Anzen’s own OTC Market post says most users are not allowed to redeem the underlying RWA directly, and frames a market mechanism to bridge that gap.

Second, tokenomics analysis has to include (a) mint and redeem rules in the contracts, (b) the staking wrapper mechanics, and (c) stability tooling that substitutes for a hard 1.00 redemption guarantee.

Mint, burn, and supply constraints

USDz supply is not pre-allocated. It expands and contracts through mint and burn around deposits and redemptions. CoinGecko lists max supply as infinite, which is directionally correct for a mintable stablecoin.

Mechanically, the canonical source of truth is the verified USDz contract. On Ethereum mainnet, the USDz contract is deployed at 0xA469B7Ee9ee773642b3e93E842e5D9b5BaA10067.

The contract implements a direct deposit path from USDC. In code, deposit(uint256) transfers USDC in, converts the 6-decimal USDC amount into an 18-decimal internal amount by multiplying by 1e12, and mints USDz to the depositor net of fees.

Fees exist, and they are bounded by constants. The contract defines maxMintFeeRate and maxRedeemFeeRate as 1% (using a fee coefficient of 1e8).

Where do those fees go. Mint-side, the contract mints the fee amount to a treasury address when mint fees are non-zero.

Redeem-side, the contract computes a fee, burns the post-fee amount, and transfers the fee to the treasury.

Redemption is explicit in the contract. There is a redeem(uint256) path that returns USDC to the caller, subject to a reserve sufficiency check against spct.reserveUSD() and the contract’s collateral checks.

There is also a redeemBackSPCT(uint256) path that returns SPCT to the caller, again subject to collateral checks and blacklist restrictions.

One subtle but important rule sits in the collateral guard. The contract defines a constant collateralRate = 1 and enforces oracle.getPrice() / 1e18 >= collateralRate via a checkCollateralRate modifier on deposits and redemptions.

From a mechanism design angle, this is a deterministic “circuit breaker,” but only to the extent the oracle feed and the oracle admin are credible. If you want more depth on how we evaluate these control surfaces, we publish ongoing research notes on crypto mechanism design.

Supply is also multi-chain. Anzen documents USDz deployments on Ethereum, Base, Arbitrum, Blast, and Manta, with contract addresses published in their documentation.

Cross-chain behavior is built around LayerZero’s OFT standard, per Anzen’s bridging guidance and product positioning.

Finally, there is an explicit “cap” concept in the docs, even if the cap value is not publicly parameterized. The FAQ states that minting can hit a cap limit and further minting requires waiting for the next cap increase.

This is one of those places where modelability drops. A cap exists, but if the cap-setting process is discretionary and not on-chain, it becomes a governance lever, not a rule.

Staking: sUSDz as a yield-bearing wrapper

The staking wrapper is where USDz tokenomics becomes “product tokenomics.” Staking is currently only supported on Ethereum mainnet according to the staking guide.

When a user stakes USDz, they receive sUSDz. The staking guide says the sUSDz balance stays constant while the value of sUSDz in terms of USDz increases when rewards are distributed.

Anzen’s “Getting Started” page describes the same structure: rewards accumulate in the staking contract, increasing the value of sUSDz over time rather than paying out. When unstaking, users receive USDz equal to principal plus appreciation.

Operationally, rewards are emitted on weekly epochs. The rewards distribution page states epochs last 7 days and distributions start every Monday (US time).

Rewards vest linearly over the epoch. Users staking for partial epochs earn pro rata.

Compounding is automatic. The rewards page states rewards automatically compound into each user’s balance, so users do not need to claim.

There is a liquidity cost. Unstaking triggers a mandatory 7-day cooldown before withdrawal, per the staking guide.

And cooldown funds do not earn staking rewards, again per the rewards distribution page.

This is a classic design trade. Cooldowns protect the reward pool from instantaneous “in-and-out” behavior around reward snapshots. They also make sUSDz less money-like. In stress, the cooldown becomes a source of basis risk between USDz spot liquidity and sUSDz exit liquidity.

Anzen also publishes oracle infrastructure for the sUSDz exchange rate, including an exchange-rate feed used for integrations.

Stability tooling: AMM arbitrage and the OTC buyback market

Anzen’s public explanation of peg behavior relies heavily on arbitrage incentives rather than a guaranteed retail redemption. The homepage frames a model where DEX discounting creates buy pressure, and qualified market makers can mint at par when USDz trades above $1 to create sell pressure.

The missing piece in that story is what happens when arbitrageurs want out but cannot access a clean $1 redemption, or when the market wants more USDz liquidity than the system can mint at par. Anzen’s answer is an OTC market layer.

On January 13, 2025, Anzen announced an OTC market for USDz, positioned as a permissionless venue to buy and sell USDz with “zero slippage,” deployed on Base and Ethereum.

It explicitly acknowledges the redemption constraint. The same post states that most users are not allowed to redeem the RWA backing directly, and frames the OTC market as a way to decentralize market making and enable arbitrage.

The most mechanism-relevant part is the protocol’s discretionary role in buybacks. When the protocol deploys buyback capital, the OTC system uses a dual-queue allocation. The post specifies 70% of capital to a Priority Queue and 30% to a Standard Queue.

Capital allocation across queues is governed by three criteria in the post: (1) veANZ balance, (2) ask price, and (3) order creation date.

This is a hybrid design. The market is permissionless, but the “stability backstop” is a budgeted, prioritized buyback program. That is not purely algorithmic monetary policy. It is an explicit policy lever.

Anzen also discussed a 1:1 buyback concept in its “USDz Vaults” announcement on October 9, 2024, describing an upcoming protocol buyback feature at a 1:1 rate to USDC with no slippage for vault participants.

Collateral composition can also evolve. On January 29, 2025, Anzen announced expanding USDz’s collateral framework to include tokenized fund products from Superstate.

That may be good for liquidity and risk dispersion. It also reinforces the point that the backing set is a managed portfolio, not a single transparent on-chain strategy.

For a contrasting “redeemability-first” mental model in DeFi, compare this peg framing with our xDAI design review.

Governance and parameter authority

If you like deterministic systems, USDz is a mixed bag. The contract has explicit, machine-checkable constraints. Fee ceilings are hard-coded to 1%.

Core operations are gated by collateral checks using an oracle.

But the contract also has strong administrative powers.

On-chain, the USDz contract uses role-based access control. A DEFAULT_ADMIN_ROLE can pause and unpause the contract.

A POOL_MANAGER_ROLE can set mint and redeem fee rates (within the 1% cap), set the treasury address, set the oracle address, and add or remove addresses from a blacklist.

Even if you assume benevolent operators, these controls matter because they are not bounded by an objective on-chain “constitution” beyond a few caps. Pausing is binary. Blacklisting is subjective. Oracle replacement is a high-leverage action.

On the incentive side, Anzen also runs a points system. For example, their “How to Buy USDz” page states holding USDz generates 5 z-points per day, and the staking guide states staking USDz generates 1 z-point per day.

Points are not token supply, but they are part of the token’s effective incentive stack. They also come with explicit jurisdiction restrictions, and the FAQ lists the United States among restricted jurisdictions for Anzen’s points program.

The governance token layer is ANZ and veANZ. Anzen’s ANZ docs describe ANZ as a utility token deployed on Base and describe a vote-escrow (ve) model where locked tokens can govern the allocation of token rewards and liquidity incentives.

That matters for USDz because veANZ shows up directly inside the OTC buyback allocation criteria.

ANZ allocations (relevant because veANZ influences stability tooling and incentive routing):

If you want a quick refresher on interpreting allocations, cliffs, and vesting mechanics in tables like this, see our tokenomics FAQ.

The veANZ documentation is directionally clear about “gauge”-style incentive routing and fee capture, but it is thin on hard parameters like lock duration, decay curve, and distribution formulas.

For a design that leans on veANZ in the OTC buyback priority, missing parameterization reduces predictability. You can’t reason about control concentration if you can’t model the voting power function.

Security posture is documented but not equivalent to economic robustness. Anzen states it completed audits in May 2024 with Halborn, Zellic, and PeckShield, and links the reports.

If you are integrating USDz into a lending market, a perp venue, or a collateral system, this is the point where a token economy design review should treat admin powers and redemption tooling as first-class risk factors, not footnotes. For governance-heavy stability systems, our Reserve Rights review is a useful comparison framework.

Risk analysis

USDz is trying to achieve three things at once: (1) keep a tight dollar peg, (2) distribute private credit yield to on-chain users via sUSDz, and (3) remain composable and multi-chain. The design is coherent, but it concentrates risk in the exact places that deterministic engineers dislike: portfolio opacity, admin discretion, and “stability budgets” that are not protocol-enforced redemption rights.

Top 3 risks

  1. RWA-liquidity peg breaks under risk-off flow. Trigger: sustained sell pressure pushes USDz below $1 while exit demand rises faster than available backstop liquidity. Mechanism: most users cannot redeem underlying RWA directly, and peg support depends on secondary market arbitrage and an OTC buyback program that only deploys capital “when available.” Who bears it: USDz holders (mark-to-market depeg), LPs (inventory risk), sUSDz stakers (basis risk during cooldown). Indicators: USDz spot price on major DEX pools, OTC queue depth and fill latency, on-chain reserve sufficiency checks such as the contract’s reserve gating on redeem.
  2. Admin control and censorship surface leaks into “stablecoin” expectations. Trigger: regulatory pressure, exploit response, or internal risk decision leads to pausing, oracle swapping, or blacklisting. Mechanism: DEFAULT_ADMIN_ROLE can pause and unpause, and POOL_MANAGER_ROLE can blacklist addresses and change the oracle and fee rates (within caps). Who bears it: affected addresses (fund mobility), DeFi integrators (sudden collateral disablement), and market makers (inventory trapped during pauses). Indicators: pause status on-chain, RoleGranted/RoleRevoked events, changes to oracle address, blacklist updates.
  3. Cross-chain and oracle dependency risk. Trigger: messaging failure, misconfigured peers, or oracle feed disruption on a destination chain. Mechanism: USDz is deployed across chains and uses OFT bridging, so supply and liquidity become a cross-domain system. Oracles are explicitly part of the system design, including an on-chain USDz price oracle documented by Anzen. Who bears it: bridge users (stuck transfers), protocols relying on price feeds (bad debt), and LPs (chain-specific depeg). Indicators: bridge completion times, oracle update cadence and deviation, chain-specific USDz price dispersion.

Dominant risk: RWA-liquidity peg breaks under risk-off flow

This is the one that dominates because it is not a smart contract bug risk. It is a balance-sheet and market microstructure risk that compounds during stress.

Anzen’s public peg narrative is explicitly market-based. The homepage leans on arbitrage incentives and qualified market-maker minting at par when USDz trades above $1. That is reasonable on the way up, when demand is growing and issuance is the constraint.

The harder direction is down. When USDz trades below $1, “buy the discount” only works if arbitrageurs believe they can later exit at or near $1 with enough size. In fully redeemable stablecoins, that belief is anchored by a hard redemption channel. In USDz’s own framing, most users are not allowed to redeem the underlying RWA directly.

Anzen responds with an OTC market that claims zero slippage and allows anyone to participate in market making and arbitrage. That helps, but it still does not create a deterministic redemption floor. It creates a venue.

The protocol also positions itself as an actor in that venue through buybacks. But the buyback mechanism is explicitly conditional: the protocol may participate when capital is available, and then allocate that capital through the dual-queue system. Conditional buybacks are a policy tool, not an obligation.

Even the queue design embeds governance weight. In the allocation criteria, veANZ balance is one of the three determinants of priority. That is not automatically bad, but it does mean “stability support” can be preferential in who it serves. In a stress event, this can turn into an adverse selection loop. Sophisticated actors can position veANZ to front-run access to buyback liquidity. Retail and passive LPs become the residual claimants on volatility.

Now, you might argue the on-chain contract includes redemption to USDC, so there is a deterministic route. It’s true that redeem(uint256) transfers USDC out, but it is gated by reserve sufficiency (spct.reserveUSD()) and collateral checks tied to an oracle price threshold. In practice, this means the “redemption floor” is only as strong as the system’s liquid USDC reserves and the operational willingness to keep the contract unpaused and the oracle functional.

That is the essence of the dominant risk. USDz’s backing is a private credit portfolio. Private credit is not instant-liquidity collateral. It is cashflow collateral. So the design must either (a) maintain meaningful liquid buffers, or (b) rely on time, queues, and market structure to smooth exits. Anzen’s own product roadmap gestures in this direction, including an OTC buyback pathway and upcoming buyback features in vault products. But unless the liquidity backstop is parameterized as a binding rule, it remains a governance decision in the moment you need it most.

For a more traditional issuer-style stable asset with clearer redemption expectations (in a different currency context), compare with our EURCV model review.

From a mechanism design engineer’s perspective, the cleanest improvement would be to publish and bind more of the “stability budget” into deterministic constraints. Examples include explicit buyback budgeting formulas, transparent rules for cap increases, and a governance-minimized oracle and pause policy with tight, auditable conditions. As it stands, the system mixes rules with discretion. That can work in calm markets. It tends to underperform when the distribution of states gets fat-tailed.

If you need help pressure-testing these flows and parameter authorities, targeted tokenomics services are usually cheaper than learning via liquidation cascades.



This article is part of our Tokenomics Deep Dive series.