Ethereum LSDs are the fastest-growing sector in the entire industry after the merge. Billions of dollars of ETH have been deposited into stETH, rETH, sfrxETH, and a bunch of other derivative wrappers and LSDfi products. For the majors, growth has been smooth, oracles are demand-driven, and DeFi protocols widely adopt the LSDs as collateral.
For smaller competitors, getting something as simple as a Chainlink oracle setup is a lengthy and tedious process. We’ve discussed here at Flywheel before how Sam K was lamenting the slowness of Chainlink to add an sfrxETH price oracle for months with little feedback. It took Sam K. making his grievances known in the group chat and reposting on Twitter for Chainlink to finally speed up its release. LSDs on the long tail have an even worse time with integrations. A small market share means fewer fees generated, so why bother even adding them? This is a chicken and egg problem for LSDs. You need DeFi integrations to spur people to mint your token, but you also need enough people minting to drive other developers to add support. And if there is no Chainlink oracle… that’s an automatic disqualification.
The Beacon Chain aka consensus layer currently has no method to interact with the application layer. If a slashing event occurs for an LSD, the data must be transmitted to the app layer and promulgated. If there is no oracle to transmit that data, pricing might not respond to a negative slashing event. This could ultimately lead to a DeFi app ending up with bad debt and being unable to liquidate the collateral holdings.
This is where Ion protocol steps in. Ion Protocol is a lending market for borrowing ETH against LSDs, restaked LSDs, and other long-tail ETH assets. What makes Ion interesting is how they are approaching oracles for their collateral underwriting.
To lend with maximum safety and efficiency, Ion is building a data provision layer that actively monitors validator balances on the Beacon chain, and then creates a risk profile and pricing data point that can be shared back to the app layer.
Ion Protocol uses a “ZK-Enabled Proof-of-Reserve” that uses “ZK state proofs (proofs that trustlessly communicate information within blockchain state) to query information directly from Ethereum's consensus layer.”
What this means is that Ion will track individual validator balances from the Beacon Chain. Instead of basing the oracle pricing on current DEX and CEX prices, each LSD or restaking provider will be tracked. The overall balances determine the value of the collateral token deposited into the protocol.
So instead of being liquidated based on pricing, Ion will liquidate based on the changes in the underlying balances of the validators. Price data can easily be manipulated. Beacon chain balances cannot, and can only lose value when slashed. By tracking reserve values, Ion thinks they can build a more efficient and safe lending protocol.
Ion will also run ZK Machine Learning Supported Risk Underwriting, which analyzes a variety of variables for each validator comprising LSDs and then determines the relative risk of a slashing event occurring. All of the data will be processed off-chain and then provide verifiable inferences on-chain using ZK proofs. Based on the data, the protocol will set the max LTV and interest rate curves.
Any long-tail LSD or new ETH-based staking variant will be eligible for access to Ion’s lending protocol immediately without having to wait for a Chainlink oracle. Ion can lend against any ETH-staked asset, as its balances will be verifiable. Using the ZKML risk engine, as the new long-tail assets grow in size, it will be able to relax the lending terms.
I think it's a novel idea. Access to lending and leverage is a key part of growth for any protocol. Going through the chicken/egg process to obtain collateral access to a money market is tedious. Ion’s ability to lend to any ETH product should open the doors for newly funded LSD entrants to hedge their positions they otherwise might not be able to in DeFi.
When Eigenlayer goes live this year or next, there are going to be millions of different combinations of services that LSDs will be providing security for. Only with an off-chain risk engine can you effectively compute all the different variables and slashing conditions. Ion Protocol will be capable of analyzing these factors and then establishing risk scores with its ZKML tech. They potentially will be the only protocol able to lend to Eigenlayer restaked LSDs at scale.