[NIP-105] Decrease May 2025 sUSDe PT min collateral ratio

[NIP] Notional Improvement Proposal
1

We propose to decrease the min collateral ratio on the May 2025 sUSDe PT vault on Mainnet to 0.15 and the max deleverage collateral ratio to 0.4.

This change will enable users to take out more leverage against sUSDe PTs. It will increase the max available leverage from 4x to 6.6x and represents a max LTV of ~87%.

Reasoning:

  • The remaining time to maturity on the sUSDe PT has decreased substantially and it is now ~3 months to maturity instead of the original ~6 months. This decreases worst case scenario PT illiquidity risks.

  • We believe this leverage ratio still provides Notional depositors adequate protection from negative funding and perp unwind slippage scenarios.

We draw heavily from the analysis provided by Llama risk in their Ethena dashboard for this recommendation.

We believe that a reasonable worst case scenario is the following:

  1. Ethena short perp positions begin to accrue -50% funding rates.
  2. Ethena users begin to redeem USDe in large size.
  3. Ethena becomes a forced seller of their perps positions with insufficient offsetting perps liquidity.
  4. Perps basis moves against Ethena and they take temporary losses on their positions.
  5. They realize those losses on any positions they close out at those levels. This means they will likely gate redemptions to enable them to slow down and realize better exit levels.
  6. They push all losses directly onto sUSDe holders in order to maintain the USDe peg to one USD.

All in all, we believe the losses could be as bad as:

  1. -50% funding for two weeks.
  2. -10% funding for two weeks.
  3. -5% redemption hit on 40% of their total size.

Currently, there are ~$6B in USDe, ~$4B in sUSDe, ~$4B in short OI, and ~$60M in the reserve fund. These losses would amount to roughly:

  1. ~$77M
  2. ~$15.3M
  3. ~$80M

That equates to ~$132.3M in total losses. After eating through the reserve funds, that would imply a ~$112.3M loss for sUSDe holders. Out of a total ~$4B holder base, that amounts to a ~2.8% haircut to sUSDe holders.

That loss is well within the safety margin implied by this updated min collateral ratio. In order for depositors to have losses, sUSDe would need to receive a haircut of 10% - 15% which is substantially higher than what we would see here.

It’s still possible that sUSDe receives this haircut due to the centralization and trust involved in depositing into Ethena. But what this analysis shows is that it is highly unlikely for a haircut of this size to result from a significantly adverse scenario assuming that none of the trusted actors here engage in fraud.

Centralization is still a risk for this strategy but it can be better dealt with by capping deposits.