USDe is a delta-neutral basis book wrapped in an ERC-20

USDe’s core economic claim is simple. Ethena holds spot crypto assets and simultaneously runs offsetting short derivatives so the portfolio’s USD value is designed to stay roughly stable, even if the underlying crypto price rips or gaps down. Ethena calls this “delta-neutral stability,” and it’s the mechanism doing the heavy lifting for the peg as described in how USDe works.

That design choice forces a very TradFi-looking framing. USDe is not “cash.” It behaves more like a structured product referencing (i) spot crypto custody and (ii) the operational reliability of a multi-venue short book with frequent settlement. The token’s value proposition is not redemption from a bank account. It is the expectation that the hedge stays tight and that authorized arbitrage keeps secondary-market price close to $1 under delta-neutral stability.

Ethena’s own documentation is explicit that minting and redeeming is engineered to be atomic around the hedge, and that backing assets are routed to “Off-Exchange Settlement” custody providers rather than deposited onto exchanges.

One detail worth keeping front-of-mind when you model USDe as an instrument: Ethena states it “earns no profit from the minting or redeeming of USDe.” If you are looking for classic issuer spread on primary issuance, the docs are telling you it is not the story.

Supply mechanics: mint, burn, and what “infinite max supply” really means

USDe’s supply is demand-driven. There is no hard cap. CoinGecko lists USDe’s max supply as ∞, and shows circulating supply and total supply moving together as the onchain supply changes in its USDe supply data.

As of March 7, 2026, CoinGecko shows 5,955,159,756 USDe circulating (and the same number as total supply).

On Ethereum mainnet, CoinGecko lists the USDe contract as 0x4c9edd5852cd905f086c759e8383e09bff1e68b3.

Primary creation and redemption is not open to everyone. Ethena’s docs and API documentation describe mint and redeem access as restricted to whitelisted users, with KYC/AML gating. In our tokenomics methodology, we treat this access layer as part of peg design, not a footnote.

That whitelist matters because it is the functional “authorized participant” layer that enables peg arbitrage. Ethena’s “Peg Arbitrage Mechanism” describes arbitrage against the protocol’s mint/redeem contract as the currently available strategy when USDe trades away from $1 in external markets.

In the USDe Terms and Conditions, Ethena also draws a clean legal line: only “Mint Users” can redeem USDe directly with the issuer entity (Ethena BVI) under the terms and conditions.

For Mint Users, Ethena states tokenizations are credited at 1 USDe per $1 of notional stablecoin value, less applicable fees and transaction costs.

Also in those terms: USDe held in a wallet is not covered by deposit insurance protections like FDIC or SIPC. That is not a cosmetic disclaimer. It’s a reminder that this is a crypto-native instrument with market and operational risk, not an insured bank deposit.

Cashflows: protocol revenue, reserve fund capture, and who actually gets paid

The important tokenomics fact is that USDe itself does not auto-accrue yield. Ethena’s docs are explicit that if you want rewards you must stake USDe and receive sUSDe.

sUSDe is implemented as an ERC-4626 “token vault” style receipt token. You stake USDe into the StakedUSDe contract and receive sUSDe. Rewards come in as additional USDe transferred into the vault over time, so the USDe-per-sUSDe exchange rate rises.

Ethena also makes a strong promise about the staking wrapper: rewards transferred into the staking contract are only positive, and “the USDe value of stUSDe can only increase or stay flat over time.” You are taking protocol risk, but the accounting mechanism aims to avoid negative rebases in the staking contract itself.

Mechanically, Ethena says reward payments into the staking contract occur every 8 hours and vest linearly over that 8-hour window to reduce “lumpy” distribution and sandwich dynamics.

Ethena’s accounting and distribution layer matters too. The “sUSDe Rewards Mechanism” page describes weekly internal APY accounting tied to a StakingRewardsDistributor contract that then “drips” rewards into sUSDe.

On the revenue side, Ethena attributes protocol revenue to three sources: (1) staking rewards on staked assets, (2) funding and basis spread from the delta-hedging derivatives book, and (3) fixed rewards on liquid stablecoin holdings.

The same page gives a snapshot allocation for early 2025. It states staked ETH was 6% of backing assets as of early 2025.

It also describes the derivatives-driven component as the dominant piece, with ~92% of backing assets as of January 2025 associated with funding and basis spread (including staked assets).

And it cites 7% of backing assets as of January 2025 in “liquid stables.”

On historical context, Ethena’s “How USDe Works” page publishes averages for 2024: BTC funding rates averaged 11%, ETH funding rates averaged 12.6%, and sUSDe APY averaged 19% in 2024. These numbers are useful, but they should be treated as descriptive of an environment, not a promise.

The other critical sink is the reserve fund. Ethena’s docs describe the reserve fund as an “additional margin of safety” used to cover periods of negative funding and as a potential bidder of last resort in open markets.

Ethena states the reserve fund is funded with a portion of protocol revenue, with the exact allocation subject to governance. It is also capitalized with funds raised from private placement investors.

Ethena’s reserve fund page gives one hard datapoint for sizing history: $46.6m as of Q4 2024.

The reserve fund is described as controlled by a 4/10 multisig with keys held by contributors within Ethena Labs.

Governance and control points: this is not “set-and-forget” money

Ethena’s governance docs are unusually candid about why it cannot be pure onchain governance. Because the system requires offchain infrastructure to manage hedges and custody relationships, Ethena states that “fully on-chain governance is not a practical or viable option at present,” and instead delegates most decisions to committees appointed by ENA tokenholders.

The Risk Committee is one of the key bodies because it touches the parameters that directly determine USDe’s risk profile. Ethena’s Risk Committee page states that committee member terms lapse every 6 months and are put up for governance confirmation or replacement.

In practice, the most important “governance” for USDe holders is not a vote. It is the set of permissions around minting, redeeming, supported collateral assets, custodian routing, and emergency stops.

Ethena’s mint/redeem contract documentation describes strict allowlists for supported assets and custodian addresses, plus per-block mint and redeem limits.

The same page lists contract roles, including an ADMIN multisig with powers like adding or removing supported collateral assets and custodian addresses, and setting max mint/redeem per block. It also includes a GATEKEEPER role shared between Ethena Labs and external security firms that can disable mint/redeem.

Ethena also documents a “Mint and Redeem Contract V2” feature called a stables delta limit, which prevents direct mint/redeem when the divergence between USDe and USDT/USDC exceeds a defined limit adverse to the protocol. It also supports distinct per-asset (and asset-group) per-block limits.

From a TradFi risk lens, these are circuit breakers and issuance controls. They reduce certain attack surfaces. They also introduce a form of discretionary control that looks more like a risk desk than a fully permissionless stablecoin.

History that matters for tokenomics: backing mix, attestations, and Proof of Reserves

USDe’s tokenomics has already evolved in ways that affect modelability. Ethena’s “Liquid Stables: Dynamic Allocation” page describes an initial design starting in April 2023 focused primarily on ETH and ETH LSTs hedged with perps and dated futures, and it explains why stablecoin buffers and other assets were incorporated as liquidity and risk management tools.

The same page describes a launch environment with elevated funding, stating that at public launch funding rates reached as high as 60% annualized in some weeks and that Ethena generated over $8m in weekly protocol revenue on launch week.

It also describes an early stress event in April/May 2024, when the protocol handled over 100m USDe redeemed during a sharp open interest reduction, with USDe’s market price staying within 20 bps of $1 for most of the sell-off.

Transparency is part of the system’s credibility story because the product is only as good as its backing and hedge discipline. Ethena maintains monthly custodian attestations and states these validate the existence, control, and value of backing assets, and that they demonstrate none of the backing assets reside directly on exchange partners. The documentation page lists attestations through January 2026.

Ethena also launched a Proof of Reserves program. In its announcement dated April 11, 2025, Ethena states it launched USDe Proof of Reserves with inaugural attestors including Harris & Trotter, Chaos Labs, LlamaRisk, and Chainlink, and that proofs are published weekly.

That same announcement states the PoR process answers whether backing assets’ USD value meets or exceeds USDe supply, whether assets are restricted to governance-approved assets and the corresponding derivatives positions, and whether backing is delta-neutral in USD terms.

It also lists “Governance Assets” included in the PoR scope as: BTC, ETH, stETH, mETH, WBETH, SOL, USDT, USDC, USDtb, sUSDS.

One nuance that matters if you are trying to price risk: the announcement states the reserve fund assets are not included in the USD value of the backing assets for PoR. In other words, PoR is aimed at validating the core backing and hedges, while the reserve fund remains a separate buffer.

Risk analysis: USDe’s peg is a balance sheet, not a slogan

Ethena publishes a fairly comprehensive risk taxonomy. It includes funding risk, liquidation risk, custodial risk, exchange failure risk, backing asset risk, stablecoin-related risks, and margin collateral risks.

The design has two genuine strengths that are easy to miss if you only compare it to fiat-backed stablecoins. First, the backing is intended to be transparent and mark-to-market, with hedges that mechanically offset directional risk under normal market functioning. Second, Ethena has built explicit buffers and controls: a reserve fund, issuance limits, emergency stops, and attestations.

The weaknesses are structural. USDe depends on (i) centralized venues for derivatives liquidity and (ii) institutional custody and settlement plumbing. Ethena explicitly frames “exchange failure risk” as exposure to unsettled PnL between settlement cycles, even when backing is not deposited on exchanges.

Dominant risk: a correlated failure of the hedge-and-settlement stack during a fast market. This is the risk that actually causes principal loss, not just lower yield.

Here is the mechanism. USDe stability assumes the short book remains open, adequately margined, and closely tracks spot. If an exchange becomes unavailable, Ethena says the protocol’s exposure is limited to the outstanding PnL between settlement cycles and that it can redelegate assets to another venue and re-hedge. That is the intended playbook.

The stress is in the seams. In a sharp move, you can get simultaneous pressure on (i) derivatives pricing and liquidity, (ii) margining rules, (iii) custodian delegation and settlement workflows, and (iv) secondary-market liquidity for USDe itself. If any piece lags, the portfolio can temporarily become non-neutral in USD terms. That is when a “stablecoin” starts trading like a risky asset, because the market is repricing the time-to-rehedge and the probability-weighted haircut on backing.

Liquidation dynamics are part of this dominant risk, not a separate footnote. Ethena’s liquidation risk page describes that exchanges can forcibly close positions if there is a material divergence between the value of backing assets used for margin (including LSTs and stables) and the underlying of the derivatives position, especially under “no loss” exchange policies. Even if Ethena runs minimal effective leverage, exchange risk engines can still impose discretionary closures in edge cases.

The reserve fund is the backstop intended to absorb periods where net revenue is negative or where the system needs capital to defend hedges. Ethena explicitly links the reserve fund to stepping in when combined revenue sources turn negative.

But a sudden correlated operational shock is not “negative funding.” It is basis risk, liquidity risk, and settlement risk hitting at once. In that scenario, the reserve fund is still useful, but its sizing becomes a solvency question, not a marketing bullet. Ethena’s own reserve fund page anchors a historical size of $46.6m in Q4 2024. That number is not automatically “enough” if the system is doing multi-billion notional hedging and the market is discontinuous.

So the real evaluation task is monitoring whether the transparency stack and committee-driven risk controls are keeping pace with scale. The weekly Proof of Reserves and monthly custodian attestations are directionally the right tooling. They reduce hidden-leverage risk and provide external checks. They do not eliminate tail events.

Funding risk still matters, mostly because it is the demand engine. Ethena explicitly acknowledges exposure to persistently negative funding rates and frames the reserve fund as the intended absorber when combined revenue is negative.

Stablecoin-related risk is the third leg. Ethena’s docs note that incorporating fiat-backed or RWA-backed stablecoins introduces confiscation and censorship risks, and that the dynamic allocation methodology only favors stablecoin basket allocations during periods of market distress that may threaten the reserve fund and overall backing. For a more familiar crypto-collateral stablecoin design, contrast this with the Dai tokenomics model.

Put differently, USDe is trying to diversify away from a single failure mode (funding/basis) by adding exposures that are closer to TradFi and regulatory risk. That is a rational trade. It also makes the “censorship-resistant money” narrative harder to underwrite without reading the fine print.

Top 3 risks

  1. Exchange or settlement disruption. Trigger: a major derivatives venue outage, insolvency event, or forced position closure during high volatility. Mechanism: the hedge leg cannot be maintained or PnL cannot be settled inside the expected window, leaving the backing temporarily non-neutral and/or crystallizing losses via exchange risk engines. Who bears it: USDe holders (peg dislocation risk), sUSDe holders (yield interruption and potential confidence shock), and the reserve fund (buffer draw). Measurable indicators: exchange concentration and settlement cadence assumptions in Ethena’s exchange failure risk framework, abnormal growth in unsettled PnL exposure, widening and persistent secondary-market USDe discounts, and operational pauses via gatekeeper actions.

  2. Funding regime shift. Trigger: a prolonged period where perp funding turns persistently negative or basis compresses across the assets Ethena hedges. Mechanism: the protocol must pay funding, reducing or eliminating distributable rewards, and potentially requiring reserve fund support if net revenue turns negative. Who bears it: sUSDe holders first (yield collapse), then USDe holders if demand-driven supply contraction causes liquidity stress and secondary-market discounts. Measurable indicators: rolling funding rates and basis spreads on the venues used for hedging, sUSDe APY trending toward zero, reserve fund drawdowns, and declining share of supply staked (which can amplify perceived instability).

  3. Governance and key-risk events. Trigger: compromised keys, misconfigured parameters, or oracle and pricing divergences that cause mint/redeem to execute at incorrect prices or allow unwanted issuance/redemption flows. Mechanism: incorrect routing, unsupported asset enablement, or adverse mint/redeem execution can create protocol losses or force emergency pauses that impair arbitrage. Who bears it: USDe holders through peg instability, and the reserve fund through capital loss absorption. Measurable indicators: role and permission changes on the minting contract, frequency of mint/redeem pauses, changes to max mint/redeem per block, and activation or adjustment of stables delta limit controls.

If you are integrating USDe into a treasury, a lending market, or a margin system, the practical work is building a monitoring and control loop around these indicators, and maintaining a living set of internal notes and research reports on venue concentration, settlement cadence, and funding regimes.

If you need tokenomics consulting on whether USDe’s cashflow and risk distribution fits your product, treat it like underwriting a short-vol carry strategy with operational dependencies, because that’s what the mechanism resembles. If you want help with that underwriting, see our tokenomics design services.



This article is part of our Tokenomics Deep Dive series.