Ethereum co-founder Vitalik Buterin introduced a concept for creating synthetic assets based on options rather than traditional debt positions.
He believes this approach could eliminate forced liquidations and reduce reliance on high-speed oracles.
The developer noted that current DeFi protocol models are vulnerable during sharp market downturns. The use of secured debt leads to cascading liquidations, which create excessive pressure on the network and market prices.
The proposed scheme is based on a pair of assets (P and N) with a strike price S and an expiration date M. They are issued by splitting 1 ETH and can be redeemed at any time. At expiration, an oracle fixes the index value, and funds are distributed between the holders of P and N.
“Since the total payout always equals 1 ETH, the possibility of liquidation is absent,” Buterin emphasized.
The main advantage of this model, according to the programmer, is its resistance to manipulation. Existing protocols require real-time price updates, which pose security risks and MEV vulnerabilities. The options structure allows for the use of “slow” oracles, similar to those used in prediction markets.
To maintain stable exposure (for example, pegging to the dollar), Buterin suggested using deeply profitable options with regular automatic rebalancing through DAOs or local scripts.
Among the potential drawbacks of the concept, the developer highlighted:
- the risk of slippage during frequent rebalancing (costs could reach 2% per year);
- quadratic deviation from the target index;
- the complexity of implementing effective on-chain automation.
Buterin added that he would feel “much safer” holding algorithmic stablecoins built on such architecture rather than those based on latency-sensitive price feeds.
Recall that in early March, the Ethereum founder presented a plan for two key changes at the network execution level: transitioning to a binary state tree and a long-term replacement of the EVM.
In March, Buterin called for restructuring the ecosystem to focus on privacy and AI.
