DAI borrow rates at Compound and Aave have been very volatile over the last month and ended at a level between 2% and 3% after including farming rewards on supplied collateral and borrowed DAI. Aave has recently decreased farming rewards by 30% but we don’t expect this to affect DAI rates significantly as the reward distribution between supplied and borrowed DAI changed from 50:50 ratio to 33:66. While rates at two largest Maker competitors are on average in line with current ETH-A rate, note that the rate volatility on these secondary lenders platforms is very high, implying that additional larger borrow amounts could not be sustained at these low rates at Aave and Compound as easily.
This was also one of the main reasons why we thought a smaller increase in ETH-A rate should be sustained by the large borrowers. According to the last study performed, most of the recent debt increase on ETH-A was performed by two entities who represent already two thirds of ETH-A debt exposure.
Risk premiums have remained at similar levels over the last months, except for some smaller assets that have seen outflow of liquidity from Sushiswap lately - TVL dropped by more than 1bn in the last 2 weeks, potentially due to different allocation of Sushi liquidity mining rewards. Assets that are being affected are mostly tokens that are currently being offboarded at Maker. YFI liquidity dropped significantly on Sushiswap, but 25m out of 31m YFI-A debt is represented by Yearn treasury and YFI vault strategy, which is normally treated as a safe and well managed position.
WBTC vaults still present a threat at current exposure due to potentially long WBTC redemption cycle issues, but after knowing the largest two borrowers are two large and well capitalized centralized lenders, we feel a bit more relaxed about the state of WBTC vault risk. Additionally the upcoming safer and cheaper WBTC-C vault should offload some of these risks.
The calculated ETH-B risk premium is an outlier currently and has a high value because of the large amount of recently newly opened vaults that the risk model by default treats as medium or high risk, since we don’t have any behavioural inputs to use.
- increase Stability Fee by 0.25% from 2.5% to 2.75%
Competitive rates have decreased slightly in the last month. The proposed rate of 2.75% will remain below the competitive rates but bring us closer to them. This continues our policy of proposing small rate changes where possible.
- increase Stability Fee by 0.5% from 6.0% to 6.5%
Similar adjustment like with ETH-A. ETH-B users are traditionally less sensitive to rate changes so we propose a bigger increase of the Stability Fee here.
- increase Maximum Debt Ceiling by 500 MM from 1500 MM to 2000 MM
- increase Target Available Debt by 20 MM from 60 MM to 80 MM
The DAI-minting from WBTC-A has slowed down in the last week, maybe also related to the Mid-month rates increase. The primary driver for the accelerated DAI-minting has been identified - Celsius - so we feel a lot more confident now with WBTC in general which is why we propose changing
gap to allow more DAI from this vault type.
- increase Stability Fee by 1.0% from 1.5% to 2.5%
We have been keeping the rates for LINK-A artificially low to stimulate DAI-minting. Within the last 6 weeks ~25 MM new DAI have been minted so we think it is time to bring this to a level we have for other similar collateral types.
- increase Stability Fee by 3.0% from 3.0% to 6.0%
With the currently set Stability Fee MANA-A is not a profitable product. We propose to increase the Stability Fee with the target to breakeven with the oracle costs. If this product starts working again, we can decrease the fees again later - if it doesn’t the community should think about offboarding it like we already did with a couple of other collateral types.
- increase Stability Fee by 2.0% from 1.0% to 3.0%
Similar situation as with MANA-A - the rates are artificially low but the intended effect of increased DAI minting is not happening. We applied a similar growth strategy here as we did with LINK-A, but in this case it did not work out. The proposed changes will barely cover the oracle costs.
- increase Stability Fee by 1.0% from 0.0% to 1.0%
The GUSD-PSM is going to launch soon. As a first step we propose to increase the Stability Fee slightly with the goal to increase motivation to unwind here and move over to PSM.
- decrease Stability Fee by 0.4% from 0.5% to 0.1%
The APR on this pool has decreased significantly so the DAI-minting has slowed down a lot. This is related to the new 0.01% UniV3-pool which already has a much higher volume than the underlying 0.05% pool of this collateral. With the currently set Stability Fee it is not profitable to stay here, we propose to decrease the Stability Fee to reduce the loss for vault users.
Note that there is already a Signal Request in the forum for onboarding the corresponding Gelato-Token for the new pool.
The proposed changes will decrease our annualized income by ~0.75 MM - we still think this is a reasonable change as the alternative would be that the vaults will unwind.
- UNIV2DAIETH-A - increase Stability Fee by 0.5% from 1.5% to 2.0%
- UNIV2WBTCETH-A - increase Stability Fee by 0.5% from 2.5% to 3.0%
- UNIV2USDCETH-A - increase Stability Fee by 0.5% from 2.0% to 2.5%
- UNIV2UNIETH-A - increase Stability Fee by 2.0% from 2.0% to 4.0%
We propose adjusting the Stability Fees to closer align with the rates we have for the underlying collateral types.
We anticipate the effect on annualized revenue - assuming stable levels of utilization - to be
- ETH-A: increase by ~7.8MM
- ETH-B: increase by ~0.77 MM
- LINK-A: increase by ~0.68 MM
- MANA-A: increase by ~0.16 MM
- UNI-A: increase by ~0.25 MM
- GUNIV3DAIUSDC1-A: decrease by ~0.75 MM
- Uniswap V2-LPs combined: increase by ~0.58 MM
The proposed changes, if confirmed, should increase yearly revenue by about 9.6 MM or from 153.2 MM to 162.8 MM DAI.
Proposed changes will get included into next week’s on-chain poll on 2021-11-28T23:00:00Z, and if passed, will be included in an executive vote on 2021-12-02T23:00:00Z.