Lodestar Finance exploit, MakerDAO expansion into RWAs & transparency initiatives, 8 Questions for Aiham from Silo Finance…
Issue #19 of The State of DeFi Lending newsletter
Welcome to issue #19 of The State of DeFi Lending, a newsletter covering the highlights of lending markets in DeFi.
In this issue we cover
Lodestar Finance, a lending protocol on Arbitrum, suffers an oracle exploit for $6.5m.
MakerDAO is stepping up its focus on Real World Assets (RWAs) in a partnership with BlockTower Credit & Centrifuge worth $220m.
Silo Finance’s growth lead Aiham sat down with us to share insights into Silo’s unique features and views on the DeFi lending sector.
Read below for more…
News
Lodestar Finance, a lending protocol on Arbitrum, got exploited for $6.5m on 10 December. At the center of the attack was a vulnerability of the GLPOracle which enabled the exploiter to inflate the value of plvGLP and drain the lending market’s available liquidity.
The attack was highly complex and involved a number of transactions, including 8 flashloans worth $70.5m. The attacker deposited USDC as collateral on Lodestar, loop-borrowed plutus staked GLP (plsGLP) and then lent it for iplsGLP. In the process, the attacker managed to grow the difference between plsGLP and GLP which was arbitraged for profit.
Certik published a detailed analysis about the incident.
According to Certik, the Oracle vulnerability is as follows:
By manipulating the exchange ratio, the attacker was able to push up the price by 1.7x.
Lodestar also published a summary on the events and expects to recover $2.4m in lost funds.
MakerDAO deepens its focus on real world assets (RWAs) in a partnership with BlockTower Credit and Centrifuge. These new vaults have an investment capacity of $220m and aim to bring tradFi yield to DeFi investors.
In an effort to scale the protocol, MakerDAO is expanding its focus on RWAs. A recently passed vote enables the protocol to allocate up to $150m to RWAs. The collateral acquisition and portfolio management is carried out by BlockTower Credit who also provides $70m in capital. The required tech infrastructure will be sourced from Centrifuge which is a crypto platform specialized in RWAs. The stability fee is set at 4%.
BlockTower Credit will provide $70m in capital to invest alongside MakerDAO. This will be in the form of a junior tranche that is the first loss piece in case the collateral quality deteriorates.
With this initiative, MakerDAO aims to bring more “real world yield” to DeFi investors. Already today, 75% of all MakerDAO revenues are generated by RWAs.
Given the off-chain nature of RWAs, MakerDAO community members are advocating additional tools to create transparency and trust. The current RWA exposure is described in detail in MakerDAO’s monthly RWA report.
As RWAs become more relevant to DeFi protocols, data platforms like rwa.xyz target to fill an important gap by presenting loan & asset details in one dashboard.
For anyone curious in financial information of MakerDAO: Steakhouse Financial is providing detailed financial reports that break down key financial items, in an open and transparent way.
Announcements and short news
TheDefiant reports about bad debt across DeFi, referencing RiskDAO dashboard
Analysis by Crypto Risk Assessment on ApeUSD: $APE-collateralised stablecoin
Eigenphi explores how $1.6m of bad debt was created on Aave through the $CRV incident
Q&A with Silo Finance
Silo is a recently-launched lending protocol which is distinguished by isolated lending markets (aka “silos”) which are designed to minimize platform-wide bad debt. Risks are isolated to these standalone silos with no cross-contagion should individual assets be exploited. This is achieved by pairing individual tokens with a bridge asset such as ETH. The SiloDAO is also issuing its own stablecoin $XAI that is used as a bridge asset across the protocol. Silo has attracted a TVL of $21.5m since launch in August 2022 with a noticeable jump in November when $XAI launched.
Interviewee:
Aiham Jaabari - Founding contributor, handling growth for Silo with a background in economics and growth hacking
Silo launched its mainnet beta in August this year, in one of the most chaotic crypto environments given the Terra/UST/3AC implosion a few weeks earlier. What do you think is the "secret sauce" of Silo that helped you get the adoption you have seen?
Shared risk or isolated risk - it comes down to one single feature that drastically changes the way we do lending/borrowing in DeFi. We are still far from the adoption we target because the DeFi community has not yet understood the risk they assume when they use traditional lending protocols.
Silo issues its own stablecoin called $XAI. What is the motivation for $XAI and how will it support protocol adoption and growth? How well has the peg held up and how do you support price stability?
XAI is a stablecoin that allows users to amplify the capital efficiency of the protocol. When the SiloDAO extends XAI to select silos, XAI becomes a revenue generator when it is borrowed into existence. Adding a stablecoin as a bridge asset alongside ETH was an obvious choice as it opens up unique use cases like leverage, shorting, and delta-neutral strategies. XAI can be viewed as a CDP stablecoin like DAI but it has a key advantage in that it can be used as collateral across the entire silo markets, making it the premier asset for borrowing out assets.
So far, the backing of XAI is made up of ETH and USDC, which combined with our incentivized liquidity, has created a very healthy pool. Our peg has performed strongly thus far, trading just above a dollar - meaning we have ample space to extend XAI further. It is worth mentioning that 1 XAI always equals $1 in the protocol, offering position stability for borrowers and creating arbitrage opportunities.
Governance is a huge discussion point for lending protocols. How is Silo approaching this topic? Do you see best practises at other protocols?
In Silo v1, the entire protocol is owned by the SiloDAO. Token holders vote on all aspects of the lending protocol, including rolling out new silos, setting up borrowing factors, setting interest rates, and most importantly setting/ changing price oracles for our markets. It is a governance-intensive protocol at the moment.
We have chosen this approach to establish trust in the protocol - you know beforehand what oracles are used, and you trust that oracles can only be changed in a transparent governance process where a proposal takes 5 days from initiation to execution.
In my opinion, the best approach for any project is either to be either fully permissionless protocol or DAO-permissioned as is the case with Silo. Eventually, we think protocols should also move to be as immutable as possible, and we demonstrate some of that in our contracts. We see large protocols where multi-sig wallets have manager permissions. That surely cannot be secure - "Anything that can go wrong will go wrong"
The recent weeks have been particularly volatile for DeFi lending protocols. Silo’s key promise is the isolated nature of lending markets. How has the concept worked in practice and what are the first lessons learnt?
We are yet to see how our silos will mitigate risk given their recency and shallow liquidity. There is a couple of lessons that we can learn from the economic attack on Mango Markets and Aave (CRV short squeeze):
Lesson 1: Users in traditional lending protocols share efficiencies but they share risk too. All users in the protocol are at risk and a pool is only as good as its weakest link.
Lesson 2: We have a massive knowledge gap in the community as to the risk of lending in a shared pool.
Lesson 3: Be it isolated-risk markets or shared-risk ones, lending is as safe as the amount of liquidity you can find in a given moment.
Lesson 4: Risk isolation is the only way to scale up lending. Efficiency is a workable problem.
Can you tell us why you picked Ethereum mainnet for launching where lending markets are not scarce and are there any plans to go multi-chain? And if so, what would be your main criteria to pick the chains to grow into?
Risk isolation is best illustrated in a mature ecosystem where there is a long tail of token assets that lack borrow/lend markets. We believe Ethereum has this problem, and our solution is extremely promising as it opens up markets for assets that will likely never be listed in any of the common shared-pool markets. This ability to open an infinite amount of markets is a unique advantage, and though there are lending markets on Ethereum, there is still not any good solution for the long tail assets besides our own. We are evaluating options between side chains and L2, and we are inclined to choose destinations with mature DEX and price oracle infrastructures.
Can you briefly describe Silo’s interest rate model? In how far is it different from other major lending protocols?
The model behaves similarly to the common Kink interest rate model up to an optimal utilization threshold that we set. Above the optimal utilization, interest rates start to go up predictably and slowly to a point where utilization enters a critical zone that we set. When utilization exceeds the critical utilization, the interest model employs a time factor that gradually makes borrowing more expensive as time goes on. At full utilization, borrowing rates increase at a predictably faster rate.
Dynamic interest rate models work greatly for long-tail assets but might not be as efficient for highly liquid assets. That is why we have 10 iterations of the main model currently implemented in our markets to match market rates and APYs that are produced by native staking options and induce certain user behaviors.
What is your market outlook for fixed vs variable interest rates? Why have fixed interest rates not become a DeFi standard (yet)?
I think there is not a large market for fixed-rate lending yet. APY Opportunities are abundant in DeFi, making the opportunity cost too high for a fixed-rate lender.
Do you think DeFi lending will eventually become a winner-takes-all-market with one big platform having >80% of the market?
I think we are looking at a Pure Competition market structure where a single lending protocol doesn’t control borrowing rates. Capital in Defi is mercenary. Everything being equal, capital will always look for higher APY, creating infinite arb opportunities that bring borrowing rates of different protocols closer to each other.