Aave announcing GHO stable, Celsius repaying $247m to DeFi platforms, MakerDAO gov wars, Deep Dive Interview with Getty Hill on Interest Protocol, and more…
Issue #3 of The State of DeFi Lending newsletter
Welcome to issue #3 of The State of DeFi Lending, a bi-weekly newsletter covering the highlights of lending markets in DeFi.
In this issue we cover -
Celsius repays $247m of stables debt to leading DeFi platforms.
MakerDAO governance wars escalate and drive ATH participation.
Aave announcing GHO, flames a DAO-2-DAO discussion.
Bastion’s USDT liquidations incident post-mortem published.
Compound announcing Comet - Compound v3 to take the protocol multichain.
A deep dive interview into Interest Protocol and DeFi lending past, present, and future with co-founder Getty Hill.
Read below for more…
Arbitrum lending TVL dropped 76% in the last 2 weeks, most of it from dForce TVL on Arbitrum shrinking from $122m to $3m.
News Recap
Celsius deleverage - repays $247m in stablecoins loans to MakerDAO, Aave, and Compound. Lenders are concerned repayments were done using their locked funds.
Since the beginning of July, Celsius has repaid over $247m of its loans to the 3 leading DeFi lending protocols MakerDAO ($224.8m DAI), Aave ($19.5m USDC), and Compound ($2.8m DAI) over 10 different transactions.
According to Zapper which tracks Celsius wallets - Celsius still has $235m in loans on Aave ($150m) and Compound ($85m), after closing its WBTC vault on Maker on July 7th. It seems in its current search for liquidity, Celsius has preferred to repay its MakerDAO’s loan first, as the 145% collateral ratio freed up more liquidity (WBTC) than their positions on Compound or Aave would. The source of the funds which was used for the repayment was not mentioned anywhere and Celsius users, who are still waiting for withdrawals to be unfrozen, were concerned the debt repayment is done using their locked funds.
While DeFi plays its part in the Celsius collapse, CeFi entity Tether announced its liquidation of a loan given to Celsius without mentioning the exact amount, something that can’t be done on decentralized platforms
Game of Delegations - MakerDAO gov wars escalate, driving participation to ATH.
The last few weeks saw MakerDAO’s governance process hits new ATH participation rates, as the discussions over the protocol’s future zoomed into one proposal, making MKR holders signal their support of 1 of 2 factions -supporting outsiders’ overview vs keeping things within inner circles.
ICYMI, here are two write-ups that cover the rolling drama from 2 different perspectives -
Monet Supply covering some of MakerDAO’s history led to the recent governance conflict, presenting the two approaches, the End Game plan supported by Rune (MakerDAO founder) which aims to maintain the reins within MKR holders vs Hasu and a few other VCs who support a more simple, some may say centralized, approach to Maker’s governance -
And Luca Prosperi who’s led the LOVE proposal which created the drama, giving his PoV on why this vote results are weakening the decentralization of Maker and why users and MKR holders should be concerned -
Aave announcing their native stablecoin GHO, invokes an open DAO-2-DAO discussion.
Aave has announced GHO, its new overcollateralized native stablecoin, backed by collateralized assets in Aave.
Though inviting the community for a discussion on the new addition to the protocol, Aave made the public announcement while the new protocol development was completed and the code was already set for audit -
The news was brought up into Maker’s forum which flamed a discussion about the competitive nature of DeFi, as concerns were raised about how GHO will affect DAI distribution in the market.
Leading figures from the Aave community, including Stani himself, posted on the forum discussion, emphasizing Aave’s desire for collaboration rather than competition. Some were not impressed by it…
Time will tell if these two DeFi giants are to tighten their long-time collaboration or will GHO set them as fierce competitors.
Bastion USDT liquidation incident
Though nearly a month old, Bastion released a post-mortem on their USDT liquidation incident that was triggered through oracle misconfiguration, taking into account a spike in the price of USDT-MXNT (Mexican Pesos) pair on Bitfinex. The 20 accounts that got affected will be fully rembursed by the Bastion team to cover the total of $81k which got liquidated due to this event.
https://bastionprotocol.medium.com/usdt-liquidation-incident-report-a33af2ab1de2
Compound announced Compound v3 - Comet - that aims to take the protocol multichain.
With the goal going multichain leading the way, Compound has announced its v3 - Comet.
The announcement met with some confusion as previous development work made by Compound Labs on its multi-chain strategy used Gateway (Compound Cash), where a CASH, a native Compound stable unit of account was discussed. It seems these plans were put aside for now, not before confusing some leading features in the space -
We will dive deeper into Compound v3 in coming issues so stay tuned…
5Qs (turning into 7Qs)
For issue #3 of the State of DeFi Lending we are taking a deep dive interview with Getty Hill, co-founder of the recently launched Interest Protocol. He will cover his perspective on the current state of DeFi lending and how IP improves it.
Interest Protocol (IP) is a borrow/lend protocol that adapts the fractional reserve system of banks to a DeFi protocol. Instead of posting houses as collateral and borrowing dollars, Interest Protocol users post wETH, wBTC, and UNI as collateral and borrow a stablecoin called USDi, and instead of depositing dollars into a bank and earning next to nothing in interest, Interest Protocol users deposit USDC into the protocol and receive the lion’s share of the interest paid by borrowers. While they build on top of all the core pieces of a borrow/lend protocol, such as the price oracle system, the liquidation mechanism, and the collateral system, Interest Protocol’s main innovation is the capital efficiency that comes from lending out its own stablecoin. Interest Protocol combines the capital efficiency of the fractional reserve system with the safety of over-collateralized loans.
Current DeFi lending platforms are either based on open market interest rates (e.g., Compound and Aave) or on a debt-backed stablecoin with an interest rate that is set by the protocol governance. IP protocol offers a hybrid solution that combines the two. Could you share your view on what is broken with the existing platforms, and how IP fix it?
The existing dominant lending platforms, MakerDao, Compound, and Aave are only the first few versions of borrow/lend protocols. MakerDao is different from Compound and Aave because its goal is to take volatile assets and derive a stable asset. Having only begun with single collateral and transitioning to the multi-collateral system they are running today that has a breadth of collaterals and tranches. Maker’s system functions more like a federal bank whose job is to manage the stability of its currency over the profit of the protocol. Borrowers deposit collateral with the protocol and pay fees directly to the DAO. In addition to managing the risks of collateral, the DAO must maintain DAI’s peg. Whereas Compound and Aave (currently) act as a middle ground between borrowers and lenders and profit from a portion of the interest paid by borrowers.
Maker’s SAI was the first attempt at any kind of a borrow/lend protocol and Compound v1 was the next major milestone for the industry. However, since then almost every borrow lend protocol launched has followed one of the two economic models. Compound & Aave style protocols are limited by the number of assets they can lend out because they are lending the underlying assets. MakerDao is limited by the assets they can lend out because an oversupply of Dai would hurt the peg. Interest Protocol mends the gap by lending out USDi, stablecoin hard pegged to 1 reserve asset (USDC) but manages the supply of USDi through a variable interest rate. The economic model uses the best part of Compound & Aave and the efficiency of a protocol asset like Maker without the same risk of de-pegging.
In the traditional world, most of the lending activity is for fixed-rate/term loans. Why is this not the case for DeFi? Can the USDi concept be extended to support fixed-rate loans? Are there plans to do it?
I think there is a bit of misunderstanding for both DeFi and traditional participants when it comes to variable versus fixed-rate loans. Traditional participants lack an understanding of variable rate protocols and their history which is important to understand for future rates. Similarly, I think DeFi participants are used to the volatility and have seen enormous costs of using fixed-rate platforms in DeFi due to the industry’s volatility which makes them very hard for the current participants to take an interest in them. Notably, most fixed-rate options available in DeFi have fairly short tenures when compared to traditional loans. Perhaps when DeFi protocols have the ability to compete on 30-year fixed-rate loans we’ll see a change.
In traditional lending and borrowing, a fixed rate is more complex than “overnight” lending paired with on-demand liabilities. The risk of liquidity mismatches is high unless lending and borrowing follow the same cadence and risk appetite. For traditional lenders – both bank and non-bank – matching is the primary solution to this. It’s inherently risky from a liquidity perspective to lend out at fixed maturities while simultaneously being responsible for borrowing at overnight rates and maturities. That liquidity and interest rate risk increases as the lender moves out to longer maturities.
To mitigate this risk, assets and liabilities are typically matched by their maturity and rate. For a bank, this is the difference between a CD and a simple savings account. The former allows the bank to confidently make a loan at around the same maturity as the CD, while the latter is much riskier for the bank to use for fixed-term lending.
DeFi has limited experience with regard to locking up an asset for a fixed term. A few protocols have begun to experiment, but it’s a tougher chicken-and-egg problem than simple overnight lending/borrowing. Each maturity (if it is an asset that can be traded) is segmented into its own market, which fragments liquidity. In order to safely lend at fixed terms, the protocol needs to also source funds with similar maturity or have enough capital to lend on its own.
As for Interest Protocol’s participation in the world of fixed-rate loans, we do not have anything planned at this time. Personally, my interest is in variable rates and building a generalized money market that solves the liquidity fragmentation problem.
Compound was extensively forked by other projects. Do you think their v3, with a Business Source Licence, is the right way to mitigate future forks?
While Compound was forked numerous times, a successful fork of the protocol has yet to be seen. In my mind, a BSL is going to do very little to protect a protocol from a malicious fork. Instead of playing the unwinnable game of code copyright, protocols and their community should be welcoming input, collaboration, and innovation. If someone forks the protocol and meaningfully improves it the parent protocol can make the same improvements and continue its lead thanks to its existing position in the market. If someone forks the protocol and does something poorly, not only is the risk isolated to that instance, but the lessons learned could be invaluable. Borrow/lend protocols are particularly difficult to fork because the risk management of those governing the protocol is key to its success. I think that is why we haven’t seen a successful fork of Compound or any of the major borrow/lend protocols.
Does IP have a multichain strategy for the future?
I think getting Interest Protocol onto more chains would be great. Aave did a great job being a first-mover on many protocols and proved that there is substantial opportunity off of Ethereum. As for a strategy, we’re deployed on Ethereum and coming up on our first month. Once IP has navigated its first couple of months, I think governance can consider additional deployments. Something unique to interest protocol is its ability to support governance tokens as collateral assets and retain their voting power. For that reason, I am excited by the prospect of deploying on Optimism and supporting OP as collateral.
MakerDAO governance is making decisions on a monthly basis and recently had a mini civil war. Compound governance, on the other hand, is taking relatively few decisions every year. What is your view on the ideal governance model for lending platforms?
I think the substance of governance is much more important than the frequency of governance. Protocols like Compound and Maker should have a continuous pace of governance since the market they operate in continuously changes. However, while dealing with recent political issues, Maker has lost sight of the bigger picture. Protocol revenue has dropped off a cliff, Dai generated by non-stablecoins collateral has fallen significantly, and the PSM has seen significant withdrawals over the last two months. While the DAO has +100 people on payroll it seems many have lost sight of the bigger picture while thinking about RWA and other issues. Historically, the protocol’s breadwinner has been ETH and WBTC posted as collateral and Dai borrowed. Today the protocol has ~1 billion Dai borrowed from non-stablecoin collateral which is down from a high of 5.6 billion in December 2021. The current levels have not been seen since January of 2021.
Similarly, Compound has seen a significant drawdown of borrows from its highs. While the protocol did add several collateral assets and generally increased the listed assets’ collateral factors the protocol has followed the industry trend.
Despite interest rates being a critical component of Maker, Aave, and Compound, each protocol has done very little work to optimize the interest rates despite massive macro changes over the last several months. In an ideal world, a regular cadence of governance similar to Maker’s is positive, but without focus and effort on the core components of the protocol we’ll likely continue to see them shrink. Ultimately is a part of the opportunity that IP will get to capitalize on.
What changes have you made to your governance process that differ from Compound’s Governor Bravo and other governance systems?
We built upon Compound’s popular Governor Bravo and introduced Governor Charlie. Charlie includes three major improvements. First, we merged the timelock contract and the governor contract to solidify the timelock as a core component of the governance module. Second, we introduced emergency proposals. Emergency proposals function the same way as a traditional proposal but have a separately configurable quorum, voting period, and timelock period. After Compound proposal 62 introduced a bug into the protocol the DAO had to wait the full seven-day period to implement a fix. Emergency proposals introduce a path to make rapid changes while still securing the protocol. Lastly, we introduce Optimistic proposals. Optimistic proposals follow the same flow as traditional proposals and have a separate configurable review period and quorum. Optimistic proposals are best used by frequent proposers performing uncontroversial improvements. In the event, that a bad actor makes an Optimistic proposal a quorum of Against votes could veto the proposal. While Charlie was developed for IP I hope Charlie is adopted by more DAOs.
Do you think DeFi lending will eventually become a winner takes all market with one big platform having 99% of the market?
I think over the next few years will see a lot of changes in the top lending protocols. In that period of time, we’ll likely see one or two rise to dominate the industry due to taking on extra risk and then will likely suffer some kind of collateral issue or other exploits. Although, over the long run, I expect there to be a market of DeFi lending protocols much like the banking sector. A few large banks exist, but a wide range of smaller banks are also profitable. As DeFi continues to mature and become a core component of traditional finance we’ll see the line between DeFi and Tradfi blur.
Until next time…