Within the dynamic panorama of decentralized finance (DeFi), danger administration is the inspiration on which sustainable lending protocols are constructed.
The problem lies find the fragile stability between paternalistic danger administration (i.e. lending thresholds are set by DAO governors and danger managers) and permitting the invisible hand of the free market to find out danger tolerance.
Michael Bentley is CEO of Euler Labs.
Because the area grows, it’s crucial that we absolutely perceive the tradeoffs inherent in numerous danger administration fashions.
Euler v1 serves as a thought-provoking illustration of the continuing debate between immutable code and managed code. Whereas Euler v1 adopted a paternalistic protocol design, with code managed by a decentralized autonomous group (DAOs) that might adapt to financial shifts or bug discoveries, it confronted a crucial turning level in early 2023: a $200 million exploit.
Additionally see: Hacker behind $200 million Euler assault apologizes and returns cash
Regardless of rigorous audits, assurances, and a considerable bug bounty at launch, a seemingly minor bug emerged, resulting in a code repair adopted by an extra audit and DAO vote within the months main as much as the assault. Nonetheless, this repair inadvertently uncovered a bigger assault vector, culminating within the exploit final yr.
Whereas we finally took actions that will result in one of many largest recoveries the crypto area has ever seen, the query nonetheless stays: is paternalism in DeFi inherently dangerous?
I nonetheless suppose, as I all the time have, that paternalism is all about trade-offs and private danger tolerance. In the end, customers should weigh their perceived dangers and determine what is true for them.
The complexity of dangers in credit score protocols
Think about a lending protocol the place debtors use USDC as collateral to safe loans in ETH. Figuring out the optimum Mortgage-to-Worth (LTV) ratio for this transaction might be a frightening job. The perfect LTV is consistently shifting, influenced by components comparable to asset volatility, liquidity, market arbitrage and extra. Within the fast-paced world of DeFi, it’s impractical to calculate the proper LTV at any time.
The design of credit score protocols subsequently requires heuristics and pragmatic selections. This results in three broad classifications of danger administration fashions.
International paternalism by way of DAO governance
As we speak, the preferred type of danger administration for DeFi lending protocols is the “paternalistic” mannequin, managed by DAOs and danger administration organizations comparable to Gauntlet, Chaos, and Warden. I name this the “paternalistic” mannequin as a result of it tends to imagine {that a} governing physique – whether or not a DAO or another organizational type – understands the danger tolerance its customers should undertake higher than the customers themselves.
This ‘international’ method, adopted by protocols comparable to Euler v1, Compound v2, Aave v2/v3 and Spark, implies that LTV ratios are set comparatively conservatively. If the danger atmosphere deteriorates, the board can modify the protocol-wide LTV ratios for all customers.
Additionally see: Aave group votes to implement model 3 on Ethereum
Whereas this mannequin ensures capital effectivity for debtors and prevents liquidity fragmentation, it’s not with out drawbacks. DAOs are made up of individuals with various abilities, a lot of whom might not be certified to straight vote on danger parameters themselves.
Delegating voting rights will help put management within the fingers of extra certified DAO members, however this solely helps centralize decision-making within the fingers of some people, who usually find yourself wielding important energy. Even when these specialists make “good” choices, DAO governance takes time and choices might not be carried out shortly sufficient if the atmosphere is altering quickly.
Governance additionally forces protocol customers to just accept or reject a single danger/reward end result, when in actuality customers have very completely different tolerances. It additionally probably trains customers to anticipate dangers to be managed for them, conditioning customers to depend on paternalistic danger administration, doubtlessly hindering their capacity to make knowledgeable risk-reward choices for themselves sooner or later.
The invisible hand by way of remoted swimming pools
The free market ideas underlying the ‘invisible hand’ mannequin permit lenders to actively select their danger/reward preferences. First coined by economist Adam Smith, the “Invisible Hand” is a metaphor for the invisible forces that drive a free market economic system towards optimum options. Whereas it’s actually not infallible, it’s the foundation of nearly all of free market capitalism at present.
Protocols comparable to Kashi, Silo, Compound v3, Morpho Blue, Ajna and FraxLend permit lenders to deposit funds into numerous (largely) unaudited, remoted swimming pools, offering flexibility in LTV ratios, based mostly on free market ideas. As a result of there are lots of swimming pools to select from, customers are free to borrow for a variety of potential LTV ratios (and different danger parameters). Some could take a cautious method, borrowing at low LTV ratios and attracting fewer debtors, whereas others could also be extra open to danger and leverage.
Additionally see: The subsequent era of automated settlement
This in flip permits completely different lending and borrowing use circumstances to emerge. Issues are additionally usually less complicated on the protocol layer with free market fashions. The dearth of governance permits for the development of immutable primitives that can be utilized by anybody. Complexity and product specs could be pushed to an aggregation layer or consumer interface layer (see beneath). Whereas this doesn’t essentially cut back the complexity of the system as an entire, it does simplify the complexity of the trusted code base for the subset of customers who want to handle their very own danger.
Nonetheless, this method just isn’t with out challenges, such because the fragmentation of liquidity, making it harder for lenders and debtors to attach. Remoted swimming pools not solely make it harder for lenders and debtors to search out one another, but in addition usually make borrowing dearer (even when customers can discover a match). That is as a result of in most siled credit score market protocols, debtors use collateral that does not give them returns (e.g. Morpho Blue, Compound v3, FraxLend).
In monolithic lending protocols, alternatively, debtors can concurrently use an asset as collateral and lend it out on the identical time. This could considerably cut back the price of borrowing and even borrowing worthwhile, permitting rate of interest arbitrage (by way of “carry trades”). And with extra borrowing comes better returns for lenders. However there isn’t any free lunch right here. Lenders are uncovered to remortgage danger below monolithic lending protocols in a manner that they aren’t below remoted lending protocols.
Additionally see: Rehypothecation could also be widespread in conventional finance, however can by no means work with Bitcoin
Native paternalism by way of aggregators
Collectors are an answer for the disadvantages of insulated swimming swimming pools. It’s generally argued that aggregators assist remedy the liquidity fragmentation drawback related to remoted swimming pools, as a lot of the isolation is eliminated for lenders. Nonetheless, lenders are solely half the story right here. Even when lenders use aggregators, the image for debtors remains to be fragmented. Aggregators permit customers to deposit property right into a managed pool, the place danger administration is delegated to a neighborhood danger supervisor. They summary the complexity of siled swimming pools and supply passive entry to numerous danger/reward alternatives.
Aggregators come in numerous flavors nowadays. There are impartial aggregators, comparable to Yearn and Idle, which can be usually agnostic concerning the downstream credit score markets they pour into. They’re merely attempting to maximise danger/reward for his or her customers, no matter how the rewards are achieved. And there are extra protectionist aggregators, like MetaMorpho, which can be extra opinionated about the place income comes from, and usually attempt to handle danger by retaining capital inside their very own ecosystem or merchandise.
Whereas aggregators improve flexibility for lenders, they arrive with extra prices and inherent paternalistic disadvantages. They usually do nothing to deal with the challenges going through debtors, who nonetheless should work with fragmented experiences and may have their extra methods or fashions for efficient danger administration.
The necessity for modularity and adaptability
To actually scale decentralized lending and compete with conventional finance, DeFi wants a credit score ecosystem with modularity at its core, the place completely different protocol designs serve completely different consumer wants. There isn’t any one-size-fits-all resolution for drawing up a mortgage protocol.
Managed monolithic credit score protocols present capital effectivity however lack numerous danger/reward alternatives. Insulated credit score markets, pushed by the invisible hand, supply flexibility however undergo from liquidity fragmentation and excessive financing prices. Aggregators, whereas addressing plenty of points, introduce their very own set of challenges.
That is the place protocols that leverage modularity will actually shine – by design they’ll facilitate the creation and use of extremely customizable experiences – bridging the hole between monolithic lending protocols and remoted swimming pools. They acknowledge numerous consumer preferences and allow the implementation and interconnection of customized lending vaults in permissionless environments.
Additionally see: The 5 ‘unsolved issues’ of crypto in response to Haseeb Qureshi
Modularity marks a paradigm shift within the composability and connectedness of DeFi. Protocols comparable to Euler v2 permit customers to seamlessly swap between completely different danger administration fashions based mostly on their preferences. True freedom just isn’t about selecting between paternalism versus the invisible hand; what issues is which you could seamlessly swap between whichever mannequin you need at any time.
It’s this flexibility that can drive innovation and progress by way of community results as increasingly numerous protected varieties are deployed.
On the coronary heart of Euler v2’s design philosophy is the Ethereum Vault Connector (EVC) constructed in-house. Though not but carried out, the EVC is at the moment present process rigorous critiques and audits and is supported by a big bug bounty. As soon as reside, it can present the inspiration on which customers can construct vaults on high of high vaults. This agnostic method accommodates each unchanging and managed preferences. Customers in search of the simplicity of immutable, governance-free vaults can create and use them in a permissionless method.
Alternatively, those that desire a paternalistic expertise led by a DAO, danger administration organizations, or a specialised aggregator layer could select that different. Crucially, the underlying code maintains neutrality, giving customers the liberty to specific their private preferences.