Below is an attempt at synthesizing the reasoning behind the subtle but significant push in decentralized lending markets to become less opinionated, more robust and externalize risk management outside the core system. This may turn into a multipart part series as there is a ton of nuance to be unpacked. I want to give a shout out to Paul @ Morpho Onetruekirk and Greg from Ajna for pontificating on this subject and pushing the discussion forward. A special thanks goes to Nick Pai, ctxmsg, and Zack for feedback on the article.

For sake of brevity and my desire to publish this before it becomes irrelevant, this article will be a high level overview, which can be used as a foundation for a mental model of how decentralized lending will evolve in the coming product cycle.

Paul states clearly in his article that the issues Morpho has defined to plague modern lending markets (Aave, Compound et al) can be reduced to governance oversight of internalized risk parameters. Essentially governance in DeFi has succumbed to what Eliezer Yudkowsky coined Inadequate Equilibria. Simply put: governance currently lacks the expertise, precision and proper incentive to manage risk preferences on behalf of millions of users effectively.

Static vs Dynamic

The intrinsic issue with DAO governance over lending markets is that the complex instruments used to manage risk is static once set. Given governance plays a reactive role in managing risks, many of the changes are antiquated once put in place. This lagging approach to modifying risks, results in “static” risk parameters governing “dynamic” and volatile markets.

Lack of expertise

DAO governance is meant to be inclusive, which is good in theory but poor in practice. This inclusivity results in a heterogenous governing body that does not necessarily have expertise or specialization with regards to specific functions (risk management in this case). After all, DAOs are opt-in voluntary work, which is a hard model to incentivize specialists in a domain.

Incentive misalignment

There does not necessarily need to be an overlap of DAO participants and Liquidity providers. The smaller the overlap the greater the potential for misalignment of incentives.

Risk internalization

Inherently DAO governed risk systems are “risk internalized”. What this means is the core system has internal dependencies on specific risk assumptions, that are solely up to the discretion of the DAO. This is in stark contrast with externalized risk systems (e.g. AMMs) who push risk management to the edge for users to decide how to manage risk. While this makes for inferior UX for a part of the marketplace (usually LPs) there is merit to having a system absent of internal risk.

Opinionated

Lastly the determination of risk parameters is strict enforced opinion that lacks flexibility. This includes oracles, LTVs, interest rates etc. While I believe it is nearly impossible to make a truly unopinionated system, new markets are building in a way that allows a layered approach to managing these externalities.

Given the break down of the tenants of the problem, its clear the path forward for lending protocols in DeFi to undergo, what I coin “Protocol Stratification”. This process essentially remedies the issues of current DeFi lending systems by externalizing risks, and allowing for modular components to be built on top of the core protocol to support different user experiences. This strategy is not unique to DeFi lending markets, the unbundling and re-bundling of products is integral to the maturation of markets, and can be seen in all facets of the stack (Modular chains, Uni v4 hooks etc).

Clear next steps

Ajna is a good example of what an unopinionated lending protocol looks like in practice. They have taken the approach of externalizing (removing) oracles as a core dependency, which unlocks the ability to spin up a pool for any digital asset. I expect there to be 3 core lending models moving forward

• Complete risk internalized: DAO managed (Aave v3, Maker, Compound v3)

• Hybrid: Risk parameters externalized, Oracles internalized (can be adaptable), aggregated LPs (Sentiment v2)

• Complete risk externalized: P2P, User managed, Isolated LPs (Ajna)

What users can expect

Product development and competition benefits users the most, and with the next iteration of lending, I posit that users can expect the proliferation of liquidity markets for every asset on chain. Key things I expect to come out of this are

• Social Lending: Lending that takes place between social cohorts

• Action specific lending: Allowing for increased capital efficiency on specific loan actions

• RWA Backed Credit

Stripping away the limitations from the core protocol allow for quicker iteration of different products and experiences.

What investors can expect

Mainly focusing on LPs, I expect navigating crypto credit markets to be extremely tough moving forward. The proliferation of lending markets will (hopefully temporarily) increase liquidity fragmentation, choice and opportunity costs. Self-managed loan books are worse ux, and costlier. I expect there to be an aggregation (mostly around the hybrid model) That aims to solve these pain points.

I should be clear that I am incredibly (somewhat irresponsibly) optimistic on liquidity lending markets in DeFi over the next several years. That said, I am also aware a ton of work needs to be done. There are considerable short comings to the proposed new solutions:

• Liquidity fragmentation (worse for both lenders and borrowers)

• Choice paralysis (So many pools, so little LPs)

• Increased User costs

  • A layered protocol approach invites more middle-men to tack on fees

  • Self-management requires time, expertise and money (gas costs, simulations etc)

  • Being aware of contagion risks elevate

• Value Capture

  • Foundation layers will get commoditized, and fees will trend towards 0

• Incentive alignment

I, along with the Sentiment team, have been building lending markets for long enough to realize the limitations of the passive-DAO managed lending markets. We have been researching and developing a new way to build infrastructure that is more flexible and scalable, while still optimizing for capital efficiency. We love the current discourse that other players are contributing and we cant wait to show users what V2 has in store.

If you are a power user, LP or developer interested in what we’re building please get in touch!

DeFi has the power to create equitable markets that are accessible to billions of users, but we have to be pragmatic about how it is actually done, our hope is that Sentiment v2 is a step in the right direction.