I have deeply growing concerns about the pace, process, and risk profile of this USDe supply cap increase. USDe is not a standard stablecoin — it’s a synthetic dollar backed by derivatives positions on centralized exchanges, and its issuer (Ethena) has a documented history of large-scale outflows under market stress. Doubling the cap to $800M on a new chain instance, executed 3.5 hours after posting with no community feedback, warrants more scrutiny than a routine parameter adjustment.
Concerns
1. The cap has quadrupled in six days with declining review time
The USDe supply cap on MegaETH has gone through at least three increases since early May:
| Date |
Cap Change |
Time from Post to Execution |
| May 3 |
$100M → $200M |
~31 hours |
| ~May 6 |
$200M → $400M |
Days |
| May 9 |
$400M → $800M |
~3.5 hours |
Each increase is justified by “the cap filled up.” That’s a demand signal, not a safety signal. The filling rate tells you there’s appetite for levered stablecoin yield on a new chain. It does not tell you the position is safe under stress.
The May 3 thread at least received one substantive community response (@litostarr flagging the need for firewalls around high-leverage silos). The May 9 thread received zero community responses before execution.
2. USDe is structurally different from standard stablecoins
The proposal’s core risk thesis is that because both legs of the dominant trade are dollar-pegged (USDe collateral, USDm debt), liquidation depends on relative peg drift rather than directional price movement, making liquidations constrained to “highly unlikely USDe insolvency scenarios.”
This framing treats USDe as if it has the same peg-stability characteristics as USDC or USDT. It does not.
USDe maintains its peg through a delta-neutral derivatives strategy — staked ETH on one side, short perpetual futures on the other. That peg depends on:
- Positive or manageable funding rates. When funding rates go negative for extended periods, the strategy loses money, and the insurance fund absorbs the loss.
- Functional centralized exchange infrastructure. The derivatives positions sit on exchanges like Binance and Bybit. A major exchange outage or insolvency directly impacts USDe’s backing.
- Uninterrupted custodian operations. The backing assets are held through centralized custodians, not on-chain in a permissionless smart contract.
These are not theoretical risks. Ethena’s TVL halved from $14.8B to $7.4B following the October 2025 market crash — $8B in total outflows over two months. During that period, the backing composition shifted from 93% perpetual futures to just 11% futures, with the remaining 89% moved to stablecoins as a defensive measure. USDe maintained its peg, but the mechanism that generates its yield was fundamentally restructured under stress.
Separately, Germany’s BaFin ordered Ethena GmbH to wind down in 2025. No legal recourse mechanism exists for USDe holders if the issuer faces further regulatory action.
A “highly unlikely USDe insolvency scenario” is not the same as a “highly unlikely USDe temporary depeg scenario.” The positions in this pool have health factors between 1.01 and 1.10 with a median of 1.04 (per LlamaRisk’s own data). A 2-3% temporary depeg of USDe relative to USDm would be sufficient to trigger liquidations across a significant portion of the $800M in positions.
3. One supplier controls more than half the market
LlamaRisk’s analysis states that “the largest supplier of USDe [is] responsible for more than 50% of the market share.” At $800M, that means a single entity likely controls $400M+ in USDe supply on this market.
If that entity exits or is liquidated, the impact is not proportional — it’s structural. Concentration at this level means the market’s liquidity profile depends on one participant’s behavior, and that participant has no obligation to maintain their position.
4. MegaETH is a new chain with unproven liquidation infrastructure
This is an Aave instance on MegaETH, a relatively new L2. Cap increases on established deployments (Arbitrum, Optimism, Base) at least have years of data on liquidator behavior, oracle reliability under congestion, and bridge throughput for emergency exits.
Deploying $800M in synthetic-dollar exposure on a chain where the Aave instance has been live for weeks compounds asset-level risk with infrastructure-level uncertainty. The question isn’t whether liquidation should work on MegaETH — it’s whether we have enough operational data to be confident it will work at this scale under stress.
What I’d recommend
-
Minimum cooling-off period between cap increases. At least 7 days between successive doublings, so community members have time to review, question, and challenge. The Risk Steward process is valuable for efficiency, but serial doublings of high-risk asset exposure in under a week is not what that process was designed for.
-
Liquidation infrastructure assessment for MegaETH. How many active liquidators operate on MegaETH? What’s the maximum liquidation throughput in a 24-hour stress window? What’s the oracle update latency under network congestion? These are standard questions for any new deployment, and they become critical at $800M in concentrated, tightly-leveraged positions.
-
Concentration limits. If a single supplier holds 50%+ of the cap, further cap increases should be paused until position concentration improves. Doubling a cap that one entity controls half of doesn’t diversify risk — it doubles one entity’s exposure.
-
Stress-test disclosure. What happens to the USDe/USDm E-Mode positions if USDe depegs by 2%, 5%, or 10%? LlamaRisk has the position-level data. Publishing liquidation cascade modeling before a cap increase — not after — would give the community the information it needs to evaluate the proposal.
Bottom line
LlamaRisk’s analysis is technically sound at the narrow question it asks: is the stablecoin-loop structure internally consistent? Yes. Health factors are tight, peg correlation bounds liquidation risk under normal conditions, and demand is real.
But USDe is not USDC. It’s a synthetic dollar that has already been through one major stress event ($8B in outflows, 85% deployed-capital drawdown, a complete restructuring of its backing composition) and whose issuer faces active regulatory enforcement in at least one major jurisdiction. Scaling exposure from $200M to $800M in six days on a new chain, via a process that executes changes in hours with no community feedback, warrants more caution than the current pace reflects.
Robert Greenfield IV (@robtg4)
*Author, TOKEDEX: The Tokenomic Bible
