I think that having an open list of problems is a good reminder of progress for a field, nascent or old, research problems or practical ones. These also provide a benchmark to look towards and glance upon to take stock of the progress it has done over the years.
While there is no doubt that yield stacking is impressive and that we can create extensive levels of leverage by stacking things on top of one another. The stick of measure will be solving problems that either extend the ethos of decentralization to every real-world product or rectify the errors that current systems possess or fill the gaps to make the system seamlessly functional, all of this leads to more adoption or reduces the friction/inefficiencies in the ecosystem.
Following is an incomprehensive list of defi problems, I will take stock of these problems in 12 months.
- Automated risk scoring of lending borrowing pools -> Increasingly important problem
- Possible Solution: Risk assessment without historic data is really hard because it could be so that a pool of users with a good credit history will always pay the loan back. “Credit score” in tradfi.
- One alternative way of looking at the problem would be, looking at a function for calculating the probability of default given the pool of assets you have.
- Thinking of risk in the extreme terms of everyone defaulting although useful is an edge case, every time someone defaults (since these are collateralized pools) there'll be a liquidation. So this means that there is an existence of a liquidation price, so one way to view this problem (even without considering the users at all) given that I borrow X amount at Y price, What is the chance that price hits my liquidation price Z (in this case the user will have defaulted and will get liquidated).
- How do you calculate/model the price action?
- One naive and simple metric would be a weighted average(weight by asset in a pool of n assets) of historical/implied volatility (because in a highly volatile asset it’s likely that the price reaches your liquidation price) and you can do it on a basis of a monthly epoch (now this is a very sanitized way of thinking about pool risk because an asset can improve its order book/action price becoming less volatile)
- The added advantage of doing a weighted average across a function (IV/historical volatility) is you can just add another variable modelled through ML or some other way and get a more refined answer
- The assumption of modelling the problem this way is that a borrower might borrow from a pool with less risky/volatile assets.
- Managing Risk for lenders and distributing risk/ Undercollateralized Loans
- Defi Lending is not in a place where it can be used by traditional or real word entities because of over-collateralization
- Most real-world loans are undercollateralized
- This under collateralization ensures that the lender establishes the borrower’s ability of repayment
- Tradfi is plagued by NPAs but still ultimately fall back to some sort of credit score establishment [Spectral finance solving this, but still an open problem].
- But still, most credit score methods would rely on onchain history for credit establishment, we are moving towards privacy-centric defi is this approach extendable to that idea? [Homomorphic encryption could provide a solution]
- Uniswap v3 is one of the most used AMMs, with over a few billion dollars in it, What is the optimal strategy for lending for both passive and active liquidity providers? How can you formulate and model the problem?
- Most vault managers have failed and only charm finance appears to have some stronghold over this having consistently beaten full range v2 performance
- The tokenomics problem [Millenium problem of defi]
- A survey of different tokenomics designs and what they have accomplished for their respective projects
- is veToken the optimal token design for most protocols?
- What are some good boilerplate tokenomics that a project could look into?
- Solving the Liquidity dilemma for most protocols, is POL the final answer for this?
- What does a privacy-preserving CFMM look like, and what privacy guarantees can it make? From the paper A note on privacy-preserving constant function market makers
- Private Lending
- Lending practices today involve an almost on-chain registry, how do you make this more private
- Imagine knowing what every user borrowed and paid in the real world
- Derivatives
- Today’s on-chain options are not a success yet [The metric here would be compared to volume compared to centralized counterparts]
- Primitive is making some headway with this
- Compiler → I give you a payoff and you “compile” it by stringing together other primitives
- One way to think of a compiler is, you are interested in a certain payoff from the assets you have so you have a compiler that auto-selects and tries to model the desired payoff from the instruments/primitives available on-chain, “auto yield stacker”
- Gas derivatives
- Cross Margining Systems, from The Defi Prime Broker.
- Soulbound Governance & Proof of excellence, Inspired by Vitalik’s Blog Soulbound, Hard Problems in Cryptocurrency: Five years later
- Governance today especially in defi is misaligned with the ideas that push forward long term development and choices, Is there a way to recognize and give non-transferrable say in matters concerning the protocol.
- Recognizing a way of who should be soul bounded also solves some airdrop dilemmas which haven’t had much progress since 2019
- Talking about Economic problems from the above blog post, we’ve come a long way in some cases since 2019. Stable value assets have seen a lot of exploration and development, looking forward. Stablecoin health dilemma [Centralized ↔ Usage ↔ Mechanism]
- Is it dependent on incentivization for maintaining peg, how healthy is this?
- Decentralized Public Goods Incentivization
- While there are experiments with retroactive public goods incentivization and GitCoin also helps in this area, there is a lot of exploration in this space left
- Reputation Systems
- Why is this a defi problem? A good decentralized reputation system can do away with a lot of requirements that would need a private defi ecosystem to thrive
- Oracle Systems which give access to more varieties of data
- Comments from SoK: Oracles from the Ground Truth to Market Manipulation
- Two conditions seem necessary for securing oracle systems: the market capitalization of the token stays material and the token is evenly distributed.
- Oracle systems with on-chain modules are expensive to run on public blockchains like Ethereum, which prices out certain use-cases that consume a lot of oracle data but do not generate a proportional amount of revenue (e.g., Weather data).
- There have been multiple hacks/losses in the Defi space due to oracle manipulation which suggests that we aren’t there yet.
- Better Fiat on-ramps/off-ramps, are Centralized exchanges the single point of fiat on ramps to blockchains? [Source]
Updated last on the 29th of Jan 2021.
Inspired by Riva Tez’s following tweet.