Ethena drops to $4.6B in USDe — does it work when the market goes down?

Ethena created the world's third-largest stablecoin — $4.6 billion in USDe as of May 2026 — without dollar reserves. Its mechanism: buying spot ETH/BTC and selling perpetual futures for the same amount. The difference between these two positions (the funding rate) is what generates the yield for sUSDe — which reached 27% in 2024 and is currently around 9.4% (7-day moving average). But when the market dropped in Q4 2025, USDe fell from $14 billion to $6 billion in weeks. And in May 2026, it continues to contract: today it's around $4.6 billion — 67% below its peak. On October 10, 2025, it lost its peg, touching $0.65 on Binance. It recovered — but the question remains: does USDe work when everything else is falling?

This article explains how Ethena's delta-neutral strategy works, why the yield fluctuates between 3% and 35%, what happened during the October 2025 depeg, how the Kelp hack affected the protocol, and what real risks it poses for an investor in 2026.

What is Ethena and how does the "synthetic dollar" work without dollars?

A traditional stablecoin (USDC, USDT) works like this: you deposit $1, and the issuer holds that dollar in T-Bills. Ethena works differently: you deposit ETH or BTC, and the protocol opens a short position in perpetual futures for the same value. The result is a delta-neutral position — which neither gains nor loses from price movements of the underlying asset.

Concrete example: you deposit $10,000 in ETH. Ethena buys $10,000 in spot ETH and simultaneously sells $10,000 in ETH perpetual futures. If ETH goes up 20%, you gain $2,000 on the spot position but lose $2,000 on the short position. Net: $0. If ETH goes down 20%, you lose $2,000 on spot but gain $2,000 on the short. Net: $0. Your capital is protected from price movement — it's "delta-neutral."

So where does the yield come from? From funding rates — the fees paid by traders who are long (betting it goes up) to those who are short (betting it goes down). Historically, in bull or neutral markets, funding rates are positive: more people are betting it goes up, and they pay for the privilege. Ethena collects these payments and distributes them to sUSDe holders (the yield-bearing version of USDe).

How much does sUSDe yield and why does it vary so much?

sUSDe's yield has fluctuated between 3% and 60% depending on market conditions:

The lesson: sUSDe yields more when the market is bullish and volatile (more people pay funding to be long). It yields less — or nothing — when the market is bearish (nobody wants to be long, funding reverses). It's not a "fixed yield" — it's a bet that the market will continue to pay for leverage.

The important comparison: a tokenized T-Bill like Ondo's USDY pays ~4.5% consistently, without volatility. sUSDe pays ~9% now but paid 3.6% two months ago and could pay 0% tomorrow. Risk-adjusted yield depends on whether you value stability or can tolerate variability.

What happened during the October 2025 depeg?

On October 10-11, 2025, a massive crypto market correction caused USDe to touch $0.65 on Binance and $0.92 on Bybit. The depeg was caused by an oracle failure: the internal price feed relied on the order book of a single platform, which collapsed during the massive liquidation.

The critical point: on decentralized DEXs, USDe remained at $0.99 — processing $790 million in volume without issues. The protocol managed $1.9 billion in redemptions without using the Reserve Fund. The underlying collateral was intact — only the price on centralized exchanges deviated.

The response: Ethena implemented weighted index-based oracles (not relying on a single exchange) and a "dynamic cooldown period" for sUSDe withdrawals — if the collateral is predominantly perpetuals, the cooldown can extend to 7 days to prevent bank runs.

How did the Kelp/rsETH hack affect Ethena?

In April 2026, the KelpDAO exploit ($292M) indirectly affected Ethena because USDe uses LayerZero's OFT standard for cross-chain interoperability — the same protocol that was compromised in the attack. Ethena suspended its OFT bridge as a preventive measure and kept it closed until the end of May.

Real impact: USDe's collateral was not affected (collateralization ratio remained >100%). But the bridge suspension prevented users from moving USDe between chains for weeks. For a protocol operating on Ethereum, Arbitrum, Base, Solana, and more — interoperability is critical. The reliance on LayerZero as a messaging layer proved to be a point of fragility that Ethena did not control.

From $14 billion to $4.6 billion — what left and why?

USDe reached a market cap of $14 billion in September 2025. By May 2026, it's at ~ $4.6 billion — 67% below its peak, and the contraction has not stopped. What happened?

• The "Aavethena loop": users deposited sUSDe as collateral on Aave, borrowed USDC, bought more USDe, deposited again. Recursive leverage that artificially inflated TVL. When funding rates fell, the loop unwound in a cascade.
• Points farming: part of the capital was seeking the ENA airdrop — not the yield. After airdrop distribution, that capital exited.
• Funding compression: from 19% average in 2024 to 3.6% in Q1 2026. The product became less attractive compared to simpler alternatives (T-Bills at 4.5%).

What remained: ~ $4.6 billion in capital that believes in the long-term model. More than 50% is staked in sUSDe — indicating conviction, not transient speculation.

What has changed in the model for 2026?

Ethena is evolving from a "funding arbitrage protocol" to "stablecoin infrastructure as a service":

• Diversified collateral: no longer just ETH/BTC. Now includes USDtb (backed by BlackRock's BUIDL), USDC, and plans to add tokenized gold and equity derivatives to generate yield independent of the crypto market.
• Whitelabel: MegaETH launched USDm, Jupiter launched JupUSD, Sui has suiUSDe — all using Ethena's delta-neutral infrastructure as a yield engine.
• HyENA (perpetual DEX): exchange built on Hyperliquid where collateral (USDe) generates yield while in position. 45% of fees go to the Ethena ecosystem.
• Reduced Risk Committee: from 5 to 3 voting members (including Blockworks Research and LlamaRisk). Current solvency: 101.11%.

What are the real risks of Ethena in May 2026?

The regulatory risk has already materialized: BaFin (Germany) ordered the liquidation of Ethena GmbH in June 2025. Ethena no longer operates in the EU/EEA. In Spain, the CNMV has MiCAR's transitional period until June 2026 — after that, Ethena's services could be further restricted for European residents.

How does Ethena compare to other yield-bearing stablecoins?

Ethena's position: higher yield than regulated alternatives (USDY, USDC on Aave) but with greater variability and no regulated access in Europe. The collapse of Resolv's USR — which used a similar delta-neutral model — demonstrates that the mechanism is not immune to infrastructure failures. The difference: Ethena has $62.5M in its Reserve Fund, real-time audits from Chaos Labs, and diversification of custodians. Resolv had a single admin key without validation. Architecture matters more than the model.

Is Ethena an investment or a tool?

Two ways to participate — with very different risk profiles:

As a tool (sUSDe): you deposit dollars, receive variable yield (3-15% depending on the market). It's an alternative to stablecoin lending on Aave/Morpho with higher potential yield but greater variability. Capital is in "dollars" (with quotes — USDe is not a dollar, it's a derivative that tries to track the dollar).

As an investment (ENA): the governance token that could activate a fee switch — redirecting part of the protocol's revenue to ENA stakers. The protocol generated $65M in Q1 2026 ($260M annualized). If the fee switch redirects 20%, that's ~ $52M/year for ENA holders. With a circulating market cap of ~ $700M, that implies a yield of ~7.4% — significant if it materializes.

But there's a huge caveat: over 3 billion ENA tokens will be issued in 2026 (selling pressure of ~ $375M at current prices). Governance has preferred to capitalize the Reserve Fund before activating the fee switch — a conservative decision that protects sUSDe but frustrates ENA holders. The dilemma: activating the fee switch attracts capital to the token but reduces the protocol's reserves against negative funding.

The comparison with Lido (LDO -96%) and Ondo (ONDO -88%) is direct: protocols with successful infrastructure and tokens that don't capture value. All three have a promised or debated fee switch. All three trade as if the market doesn't believe it will be activated. If Ethena activates first, it could be the catalyst that revalues the entire category of "infrastructure with useless token."

The most honest assessment: Ethena works — it has survived 50% corrections, a depeg, and an indirect hack via LayerZero. The delta-neutral mechanism is robust when there is demand for leverage. The risk is not that it "explodes" — it's that funding rates compress so much that the yield doesn't justify the complexity. In a world where USDY pays 4.5% with no crypto risk, sUSDe needs to consistently yield above 7% to justify its existence. In May 2026, it achieves this (~9.4%). Will it achieve it in the next bear market? That's the test it hasn't passed yet.