Stablecoins in DeFi fall into three established categories: custodial (USDC, USDT — centralized fiat reserves), CDP-based (DAI, crvUSD, Liquity — overcollateralized crypto loans), and algorithmic (UST RIP, FRAX, Reflexer — algorithmic stabilization).

Ethena USDe is a different category entirely. It is a synthetic dollar backed by a market-neutral hedge: a long position in ETH spot combined with a short position in ETH perpetual futures. The concept is a TradFi cash-and-carry trade, implemented on-chain.

USDe grew to over $5B in supply within a year. sUSDe (the yield-bearing form) delivers 10–30% APY — significantly higher than any other on-chain dollar yield. The natural question follows: what's the catch?

This article examines the mathematics of USDe: where the peg comes from, how yield is generated, the mechanics of funding rates, and what unfolds in adverse scenarios. No marketing wrappers — just financial engineering.

Part 1. Cash-and-Carry: A Classic Strategy in a New Wrapper

The Core Idea

In TradFi, a cash-and-carry trade works as follows:

• Buy the spot asset S (e.g., oil in barrels)

• Simultaneously sell a futures contract F with delivery at time T

• If F > S (contango), lock in profit = F − S − storage costs

This is arbitrage: price risk is eliminated (long spot + short futures = delta neutral), and the trader captures the basis F − S.

In crypto the logic is identical, but with perpetual futures (no expiry date):

• Long ETH spot

• Short ETH perpetual

• Net exposure = 0 in dollar terms

• Earnings = the funding rate that shorts collect from longs (when funding > 0), plus ETH staking yield when the spot leg uses stETH

The End Product: USDe

USDe is a synthetic dollar backed by the protocol's delta-neutral position. For every $1 of USDe minted, Ethena holds $1 of crypto assets in spot plus an equal-notional short perpetual position.

The spot leg serves a dual purpose: collateral for USDe and cross-margin collateral for the short — so only $1 of capital is deployed per $1 of USDe, not $2.

Changes in the spot asset's price are offset by the short's P&L, keeping the synthetic dollar value at ~$1.

The protocol's net economic position (simplified for intuition):

Asset side: 1 ETH spot ($3500)
Liability:  3500 USDe minted
Hedge:      short ETH perp $3500 notional
Net delta:  +1 ETH (spot) − 1 ETH (perp short) ≈ 0 ETH

If ETH rises to $4,000 — spot +$500, short −$500. Net P&L ≈ 0 (excluding funding payments). The same holds on the downside.

That is the intuition. What the simplified model obscures:

1. The collateral is an adaptive mix, not purely ETH. Ethena uses ETH (often as LSTs: stETH, wstETH — capturing native staking yield), BTC, SOL, and liquid stables (USDC at Coinbase, Treasury yield via BlackRock BUIDL). During periods of low or negative funding, the stable allocation increases automatically — reducing dependence on the perp market and substituting a portion of yield with the Treasury rate.

2. Collateral is held in OES (off-exchange settlement), not on CEXs. This is a critical architectural detail. Assets are held with off-exchange settlement providers — Copper, Ceffu, Fireblocks. The exchange (Binance, OKX, Bybit, Deribit) accepts them as margin via the OES provider's credit line: it grants the ability to trade perps, but does not hold the assets. Settlement at Copper occurs daily. This gives Ethena access to CeFi exchange liquidity without FTX-style counterparty risk.

3. Mint/redeem is restricted to a KYC whitelist. Direct interaction with the Ethena contract is open only to verified "Mint Users" — institutional partners from permitted jurisdictions following KYC/KYB. Retail holders purchase USDe on secondary markets: Curve, Uniswap, CEXs.

4. The USDe peg holds "in most scenarios" — not "always." Stability is maintained by several layers of protection:

• Reserve Fund (~$45M as of Q1 2026) — a buffer for negative funding periods

• Insurance Fund (~$10–20M) — a secondary layer

• Dynamic rebalancing toward liquid stables when funding is not paying

The primary systemic risk, which Ethena acknowledges publicly: sustained negative funding combined with the risk of cascading liquidations (short squeezes) in DeFi, with a reserve fund on the order of 1–2% of TVL. The buffer exists, but it is not unlimited — that is the honest characterization of USDe's peg stability.

Part 2. Yield Mathematics

Source 1: Funding Rate

Perpetual futures carry a funding rate — a periodic payment between longs and shorts (typically every 8 hours on CEXs). Its purpose: keep the perp price anchored to spot.

When the perp trades above spot (contango, bullish market) → funding is positive → longs pay shorts. This is the typical regime in a bull market.

When the perp trades below spot (backwardation, bearish market) → funding is negative → shorts pay longs.

Ethena is always short. Therefore:

• Positive funding → Ethena receives payments → positive yield

• Negative funding → Ethena pays → negative yield

Funding in Practice

Historically, on BTC and ETH:

• Median funding in a bull market: 10–15% APR

• Median in a bear market: −5% to 5% APR (can go negative)

• Peak funding during bullish euphoria: up to 100% APR (brief periods)

This explains why USDe yield is so high: in a positive-funding regime it is naturally elevated. And why it is so volatile: funding resets every 8 hours and can shift dramatically.

Source 2: ETH Staking Yield

Ethena holds stETH (or another LST) rather than raw ETH. This adds approximately 3–4% APY from Ethereum staking — income that is independent of the funding rate.

The full protocol yield formula:

Total_yield = staking_APY + funding_yield − operational_costs

staking_APY ≈ 3-4% (changes slowly)
funding_yield = (1/365) * Σ funding_payments * 3 (volatile, can be negative)
operational ≈ 0.5-1% (primarily execution costs)

Yield Distribution: USDe vs. sUSDe

An important asymmetry applies here:

• USDe — a stablecoin with no yield. Simply $1.

• sUSDe — the yield-bearing form. It receives all protocol yield minus the reserve allocation.

Holding USDe effectively subsidizes the yield for sUSDe holders. This is analogous to the Maker DSR or USDC from Circle: a non-yielding stablecoin — except unlike Circle, the destination of that yield is verifiable on-chain.

This asymmetry is Ethena's business model: the more users hold USDe (or deploy it in DeFi as an ordinary stablecoin), the more yield concentrates in the hands of sUSDe holders.

Part 3. When the Model Works and When It Doesn't

Positive Funding (Operating Scenario)

When funding is positive and robust, the model performs excellently. Ethena earns 15–25% APY from funding and 3% from staking. Minus ~1% in costs = 17–27% net APY on sUSDe.

This is the regime of an early-to-mid bull market: new capital enters crypto via perp longs, funding the shorts — i.e., Ethena.

Sideways / Mild Bear Market (Mixed Scenario)

When funding oscillates near zero, yield drops to 3–5% (staking only, plus negligible funding). USDe continues to function (the peg holds), but yield becomes worse than Aave USDC. Demand for sUSDe falls.

In this regime, a portion of USDe is redeemed for ETH and program size contracts. Ethena reduces the hedge proportionally.

Sustained Negative Funding (Stress Scenario)

This is where the model faces a fundamental challenge. If funding is persistently negative, Ethena is paying the longs (i.e., longs are receiving payments from Ethena). Net yield turns negative. The reserve fund ($45M at the last checkpoint) provides a buffer, but it depletes.

What follows:

• sUSDe yield goes to zero or negative

• Holders begin exiting — redeeming sUSDe → USDe → ETH

• Program size contracts

• Ethena closes short positions (forced buying of ETH perp)

• This creates buying pressure on the perp → funding goes further negative (self-reinforcing feedback loop)

This is a feedback loop. In the worst case — a death spiral where the program contracts faster than Ethena can manage, putting the USDe peg at risk of breaking.

Historical Periods of Sustained Negative Funding

Post-2022 (post-LUNA, post-FTX): BTC and ETH funding was slightly negative or near zero for 6+ months. Ethena would have performed poorly in that environment.

The model does account for this: hedges are distributed across multiple CEXs, and collateral can be redirected toward markets with positive funding. Some altcoins (SOL, AVAX) may carry positive funding even when BTC/ETH is negative. However, short-side liquidity on altcoin perps is shallower, limiting program scale.

Part 4. Custodial Risk: The Primary Constraint

Where the Assets Are Physically Held

This is the critical question for USDe holders. The structure:

• ETH spot (portion in stETH): held with off-exchange settlement providers (Copper, Ceffu, Fireblocks)

• Margin for perp shorts: at CEXs (Binance, OKX, Bybit, Deribit)

• Reserve in USDC: stablecoins for redemption liquidity

Off-exchange settlement means: ETH is technically held by a custodian, but via an MPC scheme one of the keys resides on the CEX side. The exchange's matching engine can use these assets as margin for perp positions but cannot withdraw them. In the event of a CEX bankruptcy, assets are returned to Ethena — in theory, subject to jurisdiction.

Risks of This Architecture

1. CEX bankruptcy

If Binance or OKX were to fail with open Ethena perp positions, the resolution could take years. In the best case, ETH is recovered through bankruptcy proceedings. In the worst case, partial loss. The FTX collapse is the illustrative precedent.

2. Custodian risk

Copper and Ceffu are centralized entities. Their own compromise represents an additional attack vector.

3. Forced position liquidation risk

If several major CEXs simultaneously recognize that Ethena is the largest short in the system, an incentive exists to manipulate the market — pumping ETH to trigger a margin call on Ethena. Practically difficult (Ethena is well-collateralized), but not impossible.

4. Regulatory action against CEXs

If the SEC or CFTC determines that margin-backed perpetual futures in USDC constitute an illegal derivative for U.S. users, they could compel CEXs to freeze positions. Ethena must maintain activity on non-U.S. CEXs, which constrains venue selection.

Part 5. Comparison with Other Stablecoin Types

USDe vs. USDC

• USDC: 100% custodial fiat-backed collateral. Risks: Circle insolvency, U.S. sanctions.

• USDe: synthetic delta-neutral. Risks: funding reversal, CEX failure, custodian failure.

The adoption ceiling for USDe is similarly constrained by legal acceptability as USDC: both depend on centralized intermediaries.

USDe vs. DAI

• DAI: overcollateralized by crypto loans (ETH/USDC via Maker). Risks: collateral liquidation cascades, governance.

• USDe: 1:1 collateralization instead of 150–200%. No cascading liquidations as a mechanism, but with direct dependence on funding rates.

USDe vs. crvUSD / Liquity

• crvUSD: soft liquidations (LLAMMA) — a specific mechanism for a gradual liquidation process

• USDe: an entirely different risk class — market volatility in funding rates, not price/liquidation dynamics

How USDe Fits Into the Stablecoin Ecosystem

USDe is not a "better" stablecoin — it represents a different tradeoff:

• Full decentralization required → DAI/Liquity (but overcollateralization → high capital cost)

• Maximum liquidity and adoption → USDC/USDT

• High on-chain yield on a stablecoin, with tolerance for funding volatility → sUSDe

USDe fills a specific niche: a high-yield on-chain dollar for the DeFi-native who understands funding and CEX risks. It is not designed for pension funds, and Ethena does not claim otherwise.

Part 6. Practical Scenarios

Scenario 1: Holding sUSDe for Yield

If you believe a bull market will continue for 6–12 months, sUSDe delivers 15–20% APY with minimal volatility (the USDe peg holds). This outperforms most DeFi yield aggregators.

Exit signals: BTC/ETH funding persistently drops below 5% APR for a week — time to consider rotating into Aave USDC or alternative yield sources.

Scenario 2: Using USDe in DeFi

USDe is accepted on Aave, Morpho, Pendle, and dozens of other protocols. It can be used as an ordinary stablecoin for lending, liquidity provision, and leveraged trading. Note: your risk is additive — you carry USDe risk plus the risk of whichever protocol you deploy it in.

Scenario 3: Shorting USDe (Bearish View)

One can borrow USDe (on Aave) → sell it → receive USDC. If USDe loses its peg under stress, buy it back cheaper and capture the spread. Cost: Aave borrow rate (often 8–15% annualized — expensive).

This is a popular strategy among skeptics of the model.

Scenario 4: Arbitrage Between USDe and sUSDe

If USDe trades at a discount to $1 (e.g., depeg event), but sUSDe yield remains elevated — one can buy USDe at a discount, convert to sUSDe, and hold until the peg recovers. The spread can reach 0.1–0.5% during stress periods.

Part 7. Open Questions and Unresolved Uncertainties

1. Sustained Negative Funding Durability

No one — including the Ethena team — knows precisely how long the protocol can absorb sustained negative funding before the program contracts below a viable scale.

2. Concentration Risk in the Stablecoin Ecosystem

USDe already accounts for a significant share of TVL in DeFi. A major depeg event could produce cascading effects across all protocols that accept USDe as collateral.

3. Regulatory Uncertainty

USDe is a synthetic dollar not backed by fiat. Regulators have not settled on a classification. It may be classified as a security, a commodity, or an entirely new asset class — each carrying a different regulatory framework.

Conclusion

USDe is the most interesting stablecoin experiment of 2024. It is a TradFi cash-and-carry trade, elegantly translated to an on-chain setting. Profitable in a bull market, but with genuine path dependency: the program's survival depends on sustainably positive funding, which in turn depends on market regimes that cannot be controlled.

For the end user: sUSDe is a high-yield stablecoin with well-defined but real risks. Not a free lunch — a structured product with a complex risk/return profile that most retail participants do not fully understand.

For DeFi protocols: integrating USDe as an ordinary stablecoin is an implicit bet on the continuity of Ethena's model. If you use USDe as collateral, you are exposed to a funding regime shift without necessarily being aware of it.

For DeFi risk infrastructure (which is what CoinYield.org addresses): funding rates are becoming a key risk dimension. USDe should not be treated as a standard stablecoin; it warrants a dedicated category with dynamic scoring based on funding metrics.

"USDe is not a conventional stablecoin. It is a packaged structure: yield from long stETH plus funding rate income, wrapped in a form convenient for the DeFi user."