USDe has no bank deposits at all, yet holds $1 — is it actually safe?
First, it relies on a hedge, not a promise: USDT's $1 is backed by Treasuries Tether claims to hold; USDe's $1 is backed by “one spot long + one equal-notional perpetual short.” Say you mint 3,000 USDe with $3,000 of ETH. Ethena simultaneously opens a $3,000 ETH short on an exchange. If ETH falls to $2,000, your spot is down $1,000 but the short is up $1,000 — the two net to $3,000. Collateral value doesn't move, so USDe keeps matching $1.
So its “safety” is conditional: the hedge works most of the time, but it swaps price-volatility risk for three new risks — exchange counterparty (the short sits on a CEX; Ethena uses off-exchange settlement to soften this), funding rate turning negative (in bear markets the short side pays), and forced liquidation of the short in violent moves. It is far safer than UST (it holds real 1:1 hedged collateral), but it is nowhere near “risk-free like USDC.” Treating it as “a yield-bearing synthetic dollar with structural risks” is more accurate than treating it as “another USDC.”
Why is USDe's yield (sUSDe) often higher than banks or USDC products — and can it disappear?
The real source is the funding rate: in perpetual futures, longs usually outnumber shorts, so longs continuously pay a “funding rate” to shorts to pull the perp price back toward spot. Ethena simply sits on the short side and collects that payment. In bull markets or bullish sentiment, funding stays positive for long stretches, and sUSDe's annualized yield can comfortably exceed 10% — not a subsidy, but genuine market-structure income. Add the staking yield from the spot leg (stETH, etc.) and you get sUSDe's high return.
But it is directional, variable, and can go negative: once sentiment flips bearish and shorts crowd in, funding can turn negative — now the short side pays the longs, and Ethena's yield drops sharply or goes negative. Negative-funding stretches are not rare historically. So sUSDe's high yield is not “risk-free fixed interest” — it is “you providing the other side of leveraged longs and collecting their funding.” Sweet in good markets, shrinking when they reverse. The simple gauge: check whether current perpetual funding is positive or negative, high or low.
USDe and UST (Luna) both look like they have “no real reserves” — what's the difference, and why did one survive while the other died?
UST's fatal flaw was circular self-support: UST's $1 relied on being redeemable for $1 worth of LUNA, while LUNA's value came from demand for UST. That's standing on your own shoulders — once confidence broke, UST sell pressure minted huge amounts of LUNA, LUNA crashed, which supported UST even less, forming a death spiral with no external asset to catch it.
USDe's difference: it actually holds hedged assets: every USDe is backed by a real position — spot long plus perpetual short — and does not rely on its own governance token. However price moves, the two legs offset and collateral value holds. There is no “the more it sells, the more it dies” loop.
But don't treat it as invincible: USDe trades that for a different risk set — funding turning negative, exchange counterparty, stETH discounts, short liquidation. It won't go to zero like UST, but it can briefly depeg in extreme conditions (it touched ~$0.97 during 2025's largest single-day liquidation and recovered within hours). In short: UST was an intrinsic design fragility; USDe is an external execution dependency — fundamentally different things.
Advanced: under what conditions is USDe most likely to depeg, and how should you read its counterparty and hedge-failure risk?
Trigger one: deeply negative funding for an extended period. The hedge mechanism doesn't break, but the yield flips into a cost. If bear-market funding stays negative for long, Ethena pays funding on the short while still owing sUSDe yield; the reserve (insurance) fund slowly drains, and once the market questions sustainability, it can trigger a wave of sUSDe unstaking and USDe redemptions.
Trigger two: exchange counterparty and custody risk. The short must sit on a centralized exchange. If an exchange fails, is hacked, or freezes Ethena's positions, the hedge leg instantly fails and the net position becomes a naked long. Ethena uses off-exchange settlement (OES) to keep assets with third-party custodians rather than on the exchange, and spreads across multiple venues to reduce this — during a ~$1.4B exchange hack in 2025, exposure was held under $30M.
Trigger three: stETH discount and liquidation. The spot leg holds a lot of stETH; if stETH discounts sharply versus ETH (it fell to 0.93 during the 2022 Terra collapse), a gap opens between the two legs. And in a violent upward spike, the short's margin can be liquidated before Ethena rebalances. Mitigations are low leverage (~1.0–1.2x) and active rebalancing. Your gauge: watch funding sign, insurance-fund size, the stETH/ETH ratio, and how many exchanges the hedge is spread across — these four tell you more, earlier, than “it's still $1 right now.”
Understand how USDe holds $1 through one concrete mint
Say ETH is $3,000 today and you mint 3,000 USDe with 1 ETH. Ethena takes that 1 ETH as spot collateral (long) and simultaneously opens a $3,000 ETH perpetual short on an exchange. The book now reads “long 1 ETH + short 1 ETH,” net exposure zero.
| Scenario | Spot 1 ETH | Short P/L | Total collateral | USDe backing |
|---|---|---|---|---|
| ETH falls to 2,000 | worth 2,000 (−1,000) | +1,000 | 3,000 | still backs $1 |
| ETH rises to 4,000 | worth 4,000 (+1,000) | −1,000 | 3,000 | still backs $1 |
Whatever ETH does, the two legs offset and the backing stays $3,000, so USDe holds 1:1. Over the same period, the short leg collects funding paid by longs and the spot leg earns staking yield — both flow to sUSDe.
Real stress test: during 2025's largest single-day liquidation, USDe briefly fell to ~$0.97, but because it is backed by real hedged positions rather than a circular token, it recovered within hours. That October told a different story — bearish sentiment, weakening funding, and fading yield appeal cut USDe's TVL in half within two months. That was not a depeg but the classic bear-market contraction of a yield-bearing stablecoin.
What this means for your money: if you load up on USDe for sUSDe's high yield, treat it as “collecting funding from leveraged longs,” not “a bank deposit.” It's sweet in good markets; when they reverse, both the yield and the size shrink — and that's exactly when you need to ask whether you can stomach a temporary discount and a redemption queue.
USDe's core trade-off: high native yield, scalable, censorship-resistant ↔ depends on positive perp funding, CEX counterparty, mechanism complexity
Choosing USDe/sUSDe buys you something other stablecoins can't offer: scale without banks or Treasuries, plus genuine market-structure yield (often 10%+ annualized in bull markets). The cost is that both the yield and the stability rest on the premise that perpetual markets stay long-biased and funding stays positive — and the short must live on a centralized exchange, introducing counterparty and custody risk that USDC and USDT don't have.
USDC, by contrast, gives you “boring but clean”: regulated, audited, low yield, low structural risk. Neither is simply better — the question is which risk you want.
Missing Link: many people don't realize that sUSDe's yield is essentially “you taking the other side of leveraged longs and collecting their funding.” That means your return is tightly tied to bullish crypto sentiment — the more euphoric the market, the more you earn; the moment it turns bearish, you not only earn less but in extremes can pay. It isn't “passive interest” — it's “continuously standing on the short side of perpetual markets.”