title: [TEMP CHECK] Integrating MakerDAO’s DSR into Aave V3 Ethereum Pool
author: @MarcZeller Aave-Chan Initiative (ACI)
created: 2023-05-31
Abstract
This TEMP CHECK proposes integrating MakerDAO’s Dai Savings Rate (DSR) into Aave V3 ETH pool. This would be achieved by integrating sDAI, an ERC-4626 Vault token that includes the DSR yield, into the Aave protocol.
Motivation
The DSR is a variable rate set by MakerDAO governance that allows DAI holders to earn savings by locking their DAI into the DSR contract. Recently, MakerDAO governance plans to increase the DSR rate to 3.49%. By integrating the DSR into Aave via sDAI, we can allow Aave liquidity providers to benefit from the DSR rate without additional cost to the Aave DAI borrowers.
Specification
At the protocol level, when a user deposits DAI, the protocol will wrap their DAI in the form of sDAI and deliver aDAI to the user. When a user borrows DAI, the protocol will unwrap the stored sDAI, deliver DAI, and accrue the debt amount in DAI.
Limitations
This integration will have the consequence of increasing the gas cost of using DAI in Aave for every user. The exact impact of these gas cost increases should be reviewed and considered by the community during the ARFC stage of this proposal.
Next Steps
If this TEMP CHECK receives positive feedback and a Snapshot vote outcome is YAE, the next step would be to draft an ARFC
detailing the technical implementation of this integration.
Disclaimer
The ACI is not paid by makerDAO or any third party to publish this TEMP CHECK.
Update from MakerDAO Governance that DSR will now be set to 3.49% rather than 3.3%. It just ended the polling stage and will move to an executive vote soon.
I believe it’s a great idea to provide DAI depositors with access to the risk-free yield offered by the DSR. However, I propose an other approach that avoids the need for an Aave upgrade and minimizes complexity in the core protocol. Instead, I suggest listing sDAI, the synthetic representation of DAI deposited into the DSR, as collateral, borrowable, or both, alongside the other assets.
Enabling sDAI as collateral would offer users the opportunity to earn a 3.49% yield on their DAI collateral. On a side note, this could potentially incentivize users to borrow other assets at a higher rate, thereby increasing borrowing activity. Enabling sDAI as borrowable would allow DAI lenders to earn a minimum of 3.49% interest rates, in addition to the interest paid by borrowers. If sDAI is listed as both collateral and borrowable, both use cases are fulfilled. In this case, the DAI market could be gradually deprecated in favor of its new alternative.
By ticking these boxes, the proposed alternative accomplishes the same goals as the initial proposal while avoiding additional complexity.
Onboarding sDAI directly means borrowers have to pay DSR yield + aave APR.that’s highly unattractive, and we’ll end up with fragmented liquidity, less overall revenue for aave users & for protocol, and higher overall cost for Aave borrowers.
That’s precisely right. Therefore, it is crucial to determine the Interest Rate Model (IRM) of sDAI accordingly.
In general, when the DAI borrowing rate remains below the DSR rate, it implies that DAI lenders would earn less by lending their funds to borrowers compared to depositing them into the DSR, which is a risk-free and liquid investment. As a result, lenders who do not utilize DAI as collateral would likely choose to withdraw from Aave and instead deposit their funds directly into the DSR. Unfortunately, I don’t have specific data available regarding the market share of these “pure lenders.”
Considering these factors, it appears reasonable to conclude that the DAI borrowing rate should remain higher than the DSR rate. Therefore, listing sDAI should be acceptable.
@MathisGD and @MarcZeller make good points. Looking at the current DAI utilisation rate I am still concerned about the cost of deployment vs total returns in the short term, however I do believe this is net positive over the longer term if the base DSR rate continues to outcompete lending interest on V3.
We support the proposal to integrate Maker’s DSR into the V3 Ethereum pool. This is key in preventing the outflow of DAI supply out of Aave in light of the recent DSR increase from 1% to 3.49%.
We have already observed users withdrawing DAI from Aave V3 Ethereum and depositing into Maker.
To illustrate the impact of this integration, we demonstrate the projected supply rate for DAI on V3 Ethereum, while utilizing the DSR on idle supply.
Given the current borrow rate and utilization, the DSR integration would lead to a similar rate for DAI suppliers on Aave compared with depositing to DSR directly:
Aave projected supply rate (at the time of writing) = (3.49*0.21)+2.84 = 3.57%
2.84% is the current supply rate for 100% of the supplied funds
With a utilization rate of ~79%, 21% of the supplied DAI is idle capital that could be supplied to DSR and generate a 3.49% yield.
9.34% is supply cap utilization and the one that matters is unutilized DAI supplied at 22% which gives extra 0.77% yield and not 2.5%. This would then somehow equal DSR, but the spread between effective supply and borrow rate would be the lowest among stablecoins at Aave making it the most efficient.
We could take 10BP off the top to make up for the difference. For example, sDAI could yield 3.39% + lending fees and the protocol would take 0.1% + the regular fees.
