Thanks @MarcZeller for this post. As we have stated previously and in our original risk recommendations for v3 Gnosis, Gnosis liquidity between stables and non-stables is limited. Managing borrow and supply caps ultimately involves balancing recursive LST strategies and recursive SDAI strategies with native demand for WETH/WSTETH collateralized stablecoin borrowing and vice versa. As such, Gauntlet recommends increasing caps conservatively in line with cap usage, in order to align with liquidity and borrow usage changes.
The previous risk steward raised wstETH, SDAI, WXDAI, USDC, and EURe caps. Should the community wish to raise caps despite lower current usage, Gauntlet provides the following parameters. These parameters are intended to be considerate of stablecoin borrows becoming increasingly collateralized by WSTETH/WETH/GNO. See the below appendix for more details.
| Asset | Current Supply Cap Usage | Current Supply Cap | Recommended Supply Cap | Current Borrow Cap Usage | Current Borrow Cap | Recommended Borrow Cap |
|---|---|---|---|---|---|---|
| WSTETH | 68% | 5,000 | 6,000 | 0 | 400 | 400 |
| WETH | 53% | 4,000 | 5,000 | 34% | 3,500 | 4,500 |
| GNO | 52% | 30,000 | - | - | ||
| SDAI | 67% | 3M | 6M | - | - | |
| WXDAI | 47% | 2.3M | 3.5M | 42% | 2.3M | 3.5M |
| USDC | 34% | 1.5M | 2M | 23% | 1.5M | 2M |
| EURe | 33% | 750k | 1M | 26% | 750k | 1M |
Increasing supply caps for stablecoins disproportionately may result in excess WETH borrowing against stables in light of current liquidity. As v3 Gnosis matures, WETH borrowing against stables may increase. For instance, v3 Arbitrum, Optimism, Polygon USDC show ~15% of nonstable debt as a percentage of USDC supplied.
Appendix
The v3 Gnosis loanbook shows significantly more correlated (i.e. stable ↔ stable, or WSTETH ↔ WETH) positions as compared to other v3 deployments on sidechains. Caps should be considerate of v3 Gnosis maturing and increased amount of stable debt collateralized by WSTETH/WETH/GNO, as well as non-stable debt collateralized by stablecoins.
Current stable debt is ~75% collateralized by SDAI and other stables. We target global stablecoin borrow caps such that should this ratio drop to 50%, repayment of liquidated non-stable collateral during liquidation incurs ~25% slippage.
As of 2023-11-20, a ~1200 WSTETH (~$2.8m) swap incurs ~25% slippage.
Current loanbook
- Analysis of current loanbook reveals one large account supplying ~60% of WSTETH and the large majority of all other assets, excluding SDAI. This account also bootstraps ~90% of the liquidity between WSTETH and stablecoins.
- Borrow distribution against stables shows small borrow against USDC and xDAI, and ~$1.1m stables against SDAI
- Borrow distribution against WSTETH shows ~$2.7m borrowed against, 90% in emode, against WETH shows ~$90k in stables
On a borrow-asset level,
- $950k WXDAI borrows are ~90% collateralized by SDAI, ~10% by WSTETH/WETH
- $350k USDC borrows are ~50% collateralized by SDAI, ~50% by WSTETH/WETH
- $200k EURe borrows are ~60% collateralized by stables, ~40% by WSTETH/WETH
As such, currently the majority of users are engaged in recursive WSTETH or recursive SDAI strategies. These are naturally sensitive to WSTETH yields, DSR, and stablecoin borrowing rates. Ultimately, borrow usage and behavior may evolve in the future on v3 Gnosis as demand for WETH leverage increases.