Ethena and USDe: Deconstructing the Delta-Neutral Synthetic Dollar

Most dollar-denominated crypto assets keep their value by holding dollars. USDe does not. It keeps its value by holding a crypto asset and simultaneously selling the same amount of that asset in the derivatives market, so that the two positions offset each other and the net value stays close to a dollar regardless of where the market goes.

Ethena's own documentation draws the distinction sharply: "USDe is not the same as a fiat stablecoin like USDC or USDT. USDe is a synthetic dollar, backed with crypto assets and corresponding short futures positions."

That single sentence contains the whole risk profile. A fiat-backed stablecoin fails if the bank fails or the reserve is misreported. A synthetic dollar built on a basis trade fails if the hedge fails. This article works through exactly what "the hedge fails" means in practice, using the mechanics and the risk data Ethena publishes itself.

1. The delta-neutral structure

The core position is the classic cash and carry trade: long the asset in spot, short the same notional in perpetual futures. The two legs move in opposite directions by design, so the combined value barely moves.

A worked example

Suppose $1,000,000 of USDe is minted, backed by $1,000,000 of spot ETH and a $1,000,000 short ETH perpetual position.

What this example actually tells you

The peg is insensitive to the collateral's price, and that is the entire point. A 30% crash in ETH does not threaten USDe's backing, which is what makes this design different from an overcollateralized stablecoin that needs a buffer to absorb the same move. Asking "what happens to USDe if ETH crashes" is asking the wrong question.

The right question is what happens if the hedge stops working. The offset only holds while the short is open, margined, and settleable. That gives exactly three failure paths, and every real risk in the system maps to one of them:

1. The short costs money to hold, because funding turned negative.
2. The short cannot be held, because its margin was insufficient during a rally.
3. The short cannot be settled, because the venue holding it failed.

Ethena's published risk categories map cleanly onto those three: funding risk, liquidation risk, and exchange failure risk, alongside custodial, collateral, and stablecoin-related risk. The rest of this article takes them in turn.

2. Where the yield actually comes from

USDe itself does not pay anything. Yield accrues only to sUSDe, the staked form. Ethena's documentation describes sUSDe as "the staked form of USDe and the protocol's dollar savings asset," and notes that "as rewards accrue to the staking contract, the USDe value of sUSDe increases over time, with no further action required by the holder."

The revenue funding those rewards comes from a diversified backing portfolio rather than from a single source. Ethena lists crypto perpetual and futures funding rates, non-crypto derivatives funding, overcollateralized DeFi lending, institutional lending revenue, tokenized real-world assets including government debt and credit, and liquid stablecoin holdings.

What this composition actually tells you

The advertised APY is a blend, and the blend is actively rebalanced. Ethena can shift allocation toward lending and real-world assets when crypto funding rates decline. That is genuine risk management, and it also means the historical yield of the product is not a stable estimate of its future yield. The strategy mix behind the number changes.

The rate per staked dollar depends on how much of the supply is staked. Revenue is generated on the entire backing portfolio, but rewards accrue only to the staking contract. That is a structural consequence of the two documented facts above: the same revenue pool is divided among a smaller staked base whenever the staking ratio falls, and among a larger one when it rises. A rising sUSDe APY is therefore not always evidence of improving conditions. It can equally reflect holders unstaking.

Unstaked USDe earns nothing while carrying all of the same risk. Holding USDe without staking gives you full exposure to every failure path described in this article and none of the compensation for it. This is the least defensible position in the system.

3. Negative funding: the documented failure mode

The short perpetual leg earns funding when longs are paying shorts, which is the normal state in a bull market. When positioning flips, the same leg pays out instead.

Ethena publishes historical data on how often this happens. Over the period it analyzed, "17.5% and 15.9% of days had a sum negative return for ETH and BTC perpetual futures respectively." Once staking yield on the ETH collateral is included, "only 8.84% of days had a sum negative revenue" for ETH. And critically for anyone modelling duration: "the longest streak of consecutive days with negative funding lasted just 13 days."

Two protocol mechanisms sit behind that exposure:

• The reserve fund. Ethena states that a reserve fund "exists and will step in on occasions when the combined revenue is negative," protecting the spot backing behind USDe.
• No negative pass-through. The documentation states that "Ethena does not pass on any 'negative revenue' to users who stake USDe for sUSDe." Staker balances are not reduced during negative periods.

What this data actually tells you

Negative funding is a frequent, normal condition rather than a tail event. Roughly one day in six was negative for ETH perpetuals before staking yield. Anyone treating sUSDe as a product that always yields has misread the instrument.

The relevant risk is regime, not incidence. A 13-day streak is a drawdown that a reserve fund is well suited to absorb. A structural shift in market positioning that keeps funding negative for months is a different problem entirely, and the reserve fund is a finite buffer measured against an open-ended drain. The published streak statistic describes the past sample. It is not a bound on the future.

Yield of zero and loss of backing are different outcomes. Because negative revenue is absorbed rather than passed through, a normal negative-funding period shows up to a staker as a low or zero APY, not as a falling sUSDe balance. Backing only comes under pressure if the drain exceeds what the reserve fund can absorb.

Where exactly does "zero APY" sit on the funding-rate axis? Ethena does not publish a staking-yield percentage or a funding-rate magnitude to answer that with real numbers, but the composition described in section 2 pins down the shape of the answer. Staking income on the ETH collateral is one part of the blended revenue, and it runs largely independent of the funding rate; the funding leg is the part that turns negative. Figure 1 plots that relationship directly, using an illustrative staking yield since Ethena does not publish one, so total yield is staking yield plus the funding rate itself.

Reading Figure 1

Figure 1 turns the yield composition from section 2 into the number section 3's day-count statistics do not give you: the funding rate at which the position stops paying. Work through it in three steps.

1. Start at the flat component. At a funding rate of 0%, the line sits at the illustrative 3.5% staking yield. That is section 2's point about the collateral's staking income made concrete: it keeps paying regardless of what the short leg is doing, because it is a different revenue source entirely.
2. Follow the line as funding turns negative. Total yield is staking yield plus the funding rate, so the line falls one point for every point the funding rate falls. It reaches zero at −3.5%, marked with the dashed clay line and the dot on the axis — the break-even shown by default before you hover anything.
3. Read the shaded region as where the trade pays nothing or worse. Past that crossing, the position's income no longer covers the short leg's funding drain. The reserve fund described above is what keeps a staker's balance from following the line down here; the chart shows the funding regime where that mechanism is doing the work, not a level it guarantees to survive.

The break-even moves with the staking-yield assumption, not with anything else. A higher blended or realized staking yield pushes the crossing further left, so funding has to be more negative before total yield turns negative. A lower one pulls the crossing toward zero. Because Ethena does not publish a current staking-yield figure, the useful discipline is to substitute your own estimate of the collateral's yield and re-run this arithmetic, rather than treating −3.5% as a real number.

Duration and magnitude are separate questions, and the 13-day streak only answers one of them. Section 3's statistic tells you how long a negative-funding stretch has lasted historically. It says nothing about how far below zero funding got during that stretch. A short streak deep past this break-even can drain a position faster than a long streak that stays a fraction of a point below it. Before sizing a position, know roughly where your own break-even sits and check both how long and how negative funding has run in the past, not just one of the two.

4. Custody: how off-exchange settlement works

The 2022 exchange failures taught a specific lesson: assets sitting on a venue can be lost when the venue fails, regardless of what your position said. Ethena's design responds to that directly.

What this structure actually tells you

OES converts an unbounded credit exposure into a bounded, time-limited one. Depositing collateral on an exchange means the entire deposit is at risk if the venue fails. Delegating through a custodian with daily settlement means only one settlement period's unrealized profit and loss is at risk. That is a real and substantial risk reduction, and it is the single most important design decision in the protocol.

It relocates the risk rather than eliminating it. The custodian is now a critical dependency. A failure, freeze, or operational error at the OES provider affects the collateral directly, and the protocol's ability to move margin quickly depends on that provider functioning during exactly the market conditions that stress everyone. Ethena lists custodial risk as its own category for this reason.

Venue diversification is a live operational requirement. Ethena trades on multiple venues including Binance, Bybit, Bitget, Deribit, and OKX, and describes actively monitoring the ecosystem and de-risking when threats increase. That is discretionary human risk management on a continuous basis, not an automated property of the system. Its quality is a real input to the product's safety and it cannot be verified on-chain.

5. How the peg is actually defended

This is the section most retail analysis skips, and it changes how you should interpret any USDe price on a secondary market.

Ethena's peg arbitrage mechanism works the way an ETF creation and redemption basket works. If USDe trades below $1 on a decentralized exchange, an arbitrageur buys it cheaply and redeems it with Ethena at $1. If it trades above $1, an arbitrageur mints at $1 and sells into the premium. Each action pushes the market price back toward par.

The constraint is who is allowed to do this. Ethena's documentation restricts mint and redeem to "authorized, whitelisted users," and its FAQ states that "all addresses will need to be whitelisted by the Ethena Protocol after satisfying KYC/AML checks," adding that "US users are not able to access the application."

What this structure actually tells you

A discount on USDe is information about arbitrageur capacity, not just about backing. A persistent gap below par can mean the market doubts the backing, but it can equally mean whitelisted participants are at their size limits, are unwilling to warehouse the position, or cannot redeem fast enough. The two look identical on a price chart and imply very different things.

Your realistic exit liquidity is the on-chain pool depth, not the protocol's redemption capacity. Size any USDe or sUSDe position against the depth of the pools you would actually sell into, plus the sUSDe unstaking period if you are staked. Redemption capacity you cannot personally access is not part of your exit plan.

Jurisdiction is a live constraint, not a footnote. Access to mint and redeem requires passing KYC, and the documentation states US users cannot access the application. Confirm your own eligibility before assuming any redemption path is available to you.

6. Liquidation on the hedge leg

Ethena defines liquidation risk as arising "when a user no longer has sufficient collateral to meet the margin requirements of the position," and characterizes it as extremely unlikely for the protocol given the minimal leverage it employs. The backing is held in spot ETH and BTC alongside staked ETH and liquid stablecoins, in proportions the documentation discloses and periodically updates.

The mitigations Ethena lists are operational: systematically delegating additional backing assets to strengthen margin, cycling backing assets between exchanges to support a stressed position, deploying reserve funds to hedging positions, and in extreme cases closing derivatives positions or disposing of affected assets.

What this actually tells you

The binding constraint is speed of delegation, not quantity of collateral. The protocol holds far more collateral than any single margin call requires, but that collateral sits with an off-exchange custodian. In a sharp upward move, the short leg loses value quickly and margin must be delegated before the exchange liquidates it. The risk is therefore an operational latency risk, and it is highest during exactly the violent rallies that also congest venues and custodians.

Partial liquidation is the realistic failure shape. Ethena notes that liquidation results in incremental position closure rather than instant loss of all collateral. A partial closure leaves the protocol under-hedged rather than unhedged, meaning some of the backing becomes directionally long until the hedge is restored. That is a smaller problem than total loss and a real one, because it introduces price exposure into an asset whose entire premise is having none.

Leverage discipline is the mitigation that matters most. Every other mitigation is a reaction. Keeping the hedge close to unlevered is what makes the reactions rarely necessary, and it is the assumption most worth monitoring over time.

7. The full risk stack

8. Conclusion

USDe is a well-documented instrument that is frequently misclassified. It is not a stablecoin with reserves in a bank. It is a delta-neutral basis trade, productized and tokenized, with the operational risk of running that trade at scale transferred from the holder to the protocol.

Judged as what it is, the design is coherent. Delta-neutrality genuinely removes collateral price risk from the peg. Off-exchange settlement genuinely converts a total exchange exposure into a single settlement cycle's worth. The reserve fund genuinely absorbs the frequent, short negative-funding periods that Ethena's own data shows are a normal feature of perpetual markets.

The risks that remain are the ones that structure cannot remove. A regime change in funding that outlasts the buffer. A custodian that fails at the moment margin needs to move. A rally fast enough that delegation lags liquidation. And for anyone who has not been whitelisted, no redemption path at all, which makes secondary market depth the only exit.

The practical discipline is short: stake if you hold, because unstaked USDe carries the risk with none of the compensation. Size against the depth of the pool you would actually sell into rather than against the protocol's total backing. Watch the reserve fund against supply rather than the headline APY. And treat a sustained funding regime change, not a bad week, as the scenario that actually matters.

For how this strategy sits alongside other advanced on-chain yield structures, see advanced DeFi yield and hedging strategies. For trading the funding rate directly as its own market rather than embedding it in a synthetic dollar, see funding rate trading on Boros.

Frequently asked questions

Sources and further reading

Primary protocol documentation:

• Ethena Documentation — Protocol overview
• Ethena Documentation — Funding risk
• Ethena Documentation — Liquidation risk
• Ethena Documentation — Exchange failure risk
• Ethena Documentation — Peg arbitrage mechanism
• Ethena Documentation — How USDe works

Related CoinBeaver articles:

• Advanced DeFi yield and hedging strategies
• Trading funding rates on margin with Boros
• Funding rates explained
• Perpetual futures explained

This article is educational and is not financial advice. The dollar figures used in the worked example are illustrative. Ethena's backing composition, venue list, custody providers, reserve fund size, and access restrictions change over time, and the historical funding statistics quoted describe the period Ethena analyzed rather than a guarantee about future conditions. Verify current figures and your own eligibility in the protocol's documentation before committing capital.