Note: I’ve deployed the CAP for this proposal. If you support this, please delegate to this address: 0x5576a4db81a44cb7158fc8d5ae752cb44f57be76
Proposal to Increase DAI Reserve Factor from 5% to 15%
I am working on a series of proposals to 1) reduce the risk in the cDAI market, and 2) compensate users who were impacted in the DAI liquidation event. I am working with active community members, stakeholders, and users of the protocol to enact these changes.
Reserve Factor Overview
The Reserve Factor in Compound is the parameter that represents the percentage of the interest paid by borrowers that accrue to that cToken’s Reserve Pool. The Reserve Pool protects lenders against borrower default.
Raising the Reserve Factor for DAI from 5% to 15%
Following the DAI liquidation event, many community members have voiced concerns regarding the state of the DAI market on Compound: the cDAI market accounts for 47% of all of the assets on Compound and exceeds the total DAI in circulation by 590M, denoting the over-concentrated and over-levered risk in DAI market. Various changes have been proposed including a revamp of the oracle system. In the interim, given the additional time and discussion needed for a new oracle system and/or liquidation mechanics, adjusting existing parameters such as the reserve factor can immediately start to de-risk the DAI market.
I propose raising the DAI reserve factor from 5% to 15% for the following reasons:
De-risk the DAI market: A higher reserve factor would decrease the economics of recursively borrowing DAI and in turn reduce the concentration, leverage, and risk of the DAI market.
Increase DAI reserves: This would help build additional reserves as a safety measure for Compound and its user-base.
Standardize Reserve Factors: The reserve factors across stablecoin markets are inconsistent. For example, USDC which has a fixed price of $1.0, has the same 5% reserve factor as DAI, a stablecoin with relatively higher price volatility. A higher reserve factor would adequately move borrow/lend activity from higher-risk assets to lower risk assets to improve the overall protocol risk profile. I’ve consulted with various active participants and believe that a 15% reserve factor is adequate, since a 10% reserve factor may not be enough to de-risk the market, and 20% may be overly aggressive.
Relatively quick change to improve security: Adjusting the DAI reserve factor requires no changes to the contracts. Ideally, a reserve factor change would be the first of many changes to improve the security of Compound.
The community has already signaled a strong interest in raising the reserve factor through a forum poll.
If you support this change, please delegate COMP to this Autonomous Proposal to:
For full transparency, I was one of the 142 Compound users affected in the DAI liquidation event. I am also currently working on a compensation plan for users affected which I will be sharing in the forums next week.