Vitalik Buterin published a research proposal on the Ethereum Research forum on June 1, 2026, titled "Building index-tracking assets on top of options instead of debt," outlining a structural replacement for the collateralized debt position (CDP) model that underpins an estimated $54 billion in D...
"It is very easy to lose 2% per year or more from multiple rounds of slippage, and this is the largest risk by which this whole scheme might become uncompetitive." — Vitalik Buterin, Ethereum Co-Founder, EthResearch Post (June 1, 2026)
Vitalik Buterin published a research proposal on the Ethereum Research forum on June 1, 2026, titled "Building index-tracking assets on top of options instead of debt," outlining a structural replacement for the collateralized debt position (CDP) model that underpins an estimated $54 billion in DeFi lending deposits across 380+ protocols. The proposal would split assets into paired option contracts that always sum to their original value, eliminating forced liquidations entirely and allowing protocols to operate on slow, prediction-market-style oracles instead of the real-time price feeds that have been repeatedly exploited.
The timing is not incidental. DeFi has recorded $942 million in hack-related losses in the first half of 2026, according to CryptoRank. Over $3 billion in leveraged positions were liquidated between June 4-6 alone, with 84.7% of losses hitting long positions. The April KelpDAO bridge exploit ($293 million) cascaded into Aave, triggering $12 billion in deposit outflows — a 46% drop in TVL — in 48 hours. The liquidation engine, meant to protect protocols, is increasingly functioning as a transmission mechanism for contagion.
No protocol has committed to implementing the proposal. Buterin himself described it as "a design question, not a development roadmap." But by June 11, he noted on X that "the 'options thing' is already happening," indicating multiple teams had begun building on the spec. Whether this remains an academic exercise or triggers a structural migration in DeFi architecture depends on whether the rebalancing costs — which Buterin estimates at 2%+ annually — can be reduced to levels competitive with existing CDP yields of 3-8% APY.
The CDP model — where borrowers lock collateral that is automatically sold if its value drops below a defined threshold — has powered DeFi lending since MakerDAO's launch in 2017. As of mid-2026, Aave V3 leads the sector with approximately $14.6-19.4 billion in TVL. Morpho Blue holds $4.9 billion. Spark holds $6.8 billion. Compound V3 holds $2.7 billion. Together, the top ten lending protocols capture 78% of deposits.
The model works in stable markets. It fails in volatile ones. The mechanism is well-documented:
March 12, 2020 (Black Thursday): ETH fell 43% in 24 hours. MakerDAO's price oracle failed to update due to gas price spikes. When it finally refreshed, the reported price dropped 20% instantly, triggering a cascade of vault liquidations. Due to keeper bot failures, $8.32 million in ETH was liquidated for $0 — bidders won collateral auctions with zero-DAI bids. The system recorded 5.67 million DAI in bad debt. MakerDAO governance voted 65% against compensating affected users. A class-action lawsuit followed, settling at $1.16 million.
March 10, 2026: An internal configuration error in Aave's Collateral Asset Price Oracle (CAPO) systematically undervalued wrapped staked Ethereum (wstETH), making healthy positions appear undercollateralized. $27 million in positions were force-liquidated in 24 hours due to a protocol safety mechanism malfunction.
June 4-6, 2026: Over $3 billion in leveraged positions were liquidated across derivatives markets. Cascading liquidations totaling $1.75-1.84 billion were recorded within 24 hours, with longs accounting for 84.7% of losses.
The Bank of Canada published a staff analytical paper in April 2026 examining Aave V3 transaction-level data. Its findings: "liquidations occur in concentrated waves" and "many users engage in recursive leverage despite overcollateralization requirements." The liquidation model amplifies exactly the conditions it is designed to protect against.
Buterin's proposal defines two assets — P (positive) and N (negative) — with a strike price S and maturity date M. The mathematical core: P + N always equals 1 ETH (or 1 unit of the base asset), regardless of price movements. There is no collateral to seize because no debt exists.
At maturity M, an oracle resolves the index price. If ETH trades above the strike price S, asset P captures the upside; if below, asset N captures the downside. The payoffs sum to 1 ETH by construction. Solvency is guaranteed mathematically, not enforced through liquidation.
The contrast with CDPs:
| Feature | CDP Model (Aave, Maker) | Options Model (Buterin Proposal) | |---------|------------------------|--------------------------------| | Collateral | Required, seizeable | None — paired contracts | | Liquidation | Automatic at threshold | Does not exist | | Oracle dependency | Real-time, continuous | Slow, at maturity/intervals | | Solvency guarantee | Economic (liquidation proceeds) | Mathematical (P + N = 1) | | Position changes | Hair-trigger, forced | Gradual, user-directed |
The design eliminates the cascading failure mode. When ETH drops 40% in 24 hours, a CDP protocol force-sells collateral into a falling market. Buterin's model does nothing — positions adjust at maturity or through user-initiated rebalancing.
Real-time oracles are a persistent attack surface. In 2026, oracle manipulation contributed to several of the 83 DeFi exploits recorded in Q2, part of $776 million in losses. The April KelpDAO bridge exploit — the second-largest of the year at $293 million — used forged cross-chain attestations to drain assets, a failure mode linked to real-time verification requirements.
Buterin's model replaces Chainlink-style sub-second price feeds with prediction-market-style oracles that update slowly and incorporate dispute resolution periods. The tolerance for price errors increases from near-zero (in CDPs, a 1% mispricing can trigger unwarranted liquidations) to 1-4% annual drift.
The tradeoff is explicit: slower oracles cannot support use cases requiring real-time price accuracy. Perpetual futures, options pricing on existing protocols, and arbitrage bots all require sub-second feeds. The proposal targets index-tracking assets — synthetic exposure to ETH, CPI, commodities, or custom baskets — not the full range of DeFi products.
Buterin's September 2025 blog post contextualized this: low-risk DeFi primitives are "essential infrastructure for Ethereum's economic sustainability," analogous to how advertising revenue sustains Google's product portfolio. The options model is positioned as foundational infrastructure, not a replacement for all existing DeFi.
Buterin's proposal is unusual for a founder-level publication in that it explicitly identifies its own primary weakness. Options-based positions tied to indices require periodic rebalancing. Every rebalance executes a trade. Every trade incurs slippage.
Buterin estimates this cost at 2% per year or more "from multiple rounds of slippage," calling it "the largest risk by which this whole scheme might become uncompetitive." For context, stablecoin supply yields across major lending protocols currently range from 3-8% APY. If rebalancing costs consume 2%+ of returns, the remaining yield may not justify the complexity.
The proposed mitigation exploits a structural advantage: rebalancing users have "very low time preference" — they do not need to execute immediately. Unlike liquidation bots that must act within seconds, rebalancers can wait days for favorable conditions. Buterin envisions "an ideal market structure that minimizes slippage far more than traditional AMMs do," though the specifics remain unspecified.
On Ethereum mainnet, gas costs compound the problem. A rebalancing trade on a small position ($1,000-$10,000) could lose a material percentage to gas alone, before slippage. Layer-2 deployments — where gas costs are 10-100x lower — would likely be required for the model to be practical for retail users.
The proposal extends to algorithmic stablecoins, but with a significant departure: instead of tracking $1 USD, synthetic assets would track personalized baskets — a mix of USD, EUR, CPI, commodities, or other indices — maintaining purchasing power against a user-defined reference.
This creates a definitional problem. A stablecoin that drifts 1-4% annually from its reference basket is acceptable for purchasing-power preservation. It is not acceptable for accounting, invoicing, or settlement — use cases that require exact dollar equivalence. Buterin acknowledged this: the system is "not designed to replicate the US dollar precisely."
In a stablecoin market that exceeded $320 billion in mid-2026, with Tether (USDT) serving 534 million users, demand for exact-dollar-pegged instruments remains dominant. The custom-basket model may find adoption in hedging and portfolio construction, but it does not address the primary stablecoin use case.
The proposal arrives in a DeFi lending market undergoing simultaneous consolidation and institutional entry.
Morpho's $175 million raise (June 9, 2026), co-led by Paradigm, a16z crypto, and Ribbit Capital at a $2 billion valuation, signals institutional confidence in the existing CDP-adjacent model. Morpho's modular lending architecture — which allows third-party risk curators to set parameters — has reached $6.6 billion in TVL and is closing the gap on Aave. Apollo Global Management and VanEck participated, marking direct asset-manager equity exposure to DeFi credit infrastructure.
An options-based migration would structurally disadvantage:
Potential beneficiaries include:
As of June 28, 2026, the proposal remains at the research stage. No named protocol has committed to building on the spec.
However, Buterin's June 11 post on X — "the 'options thing' is already happening" — indicates that unnamed teams have moved from theory to code. He added a warning: deployments reaching mainnet quickly "should be formally verified first."
The proposal credits feedback from Lighter founder Vladimir Novakovski and developers at Curve, suggesting collaboration at the protocol-design level. Whether this translates into deployed contracts depends on two unresolved questions:
The more likely path is new protocols building on the spec from scratch rather than migration of existing ones — a pattern consistent with DeFi's historical preference for forking over upgrading.
The DeFi lending stack processes tens of billions in deposits through a liquidation model that has produced repeated cascading failures — from MakerDAO's Black Thursday in 2020 to Aave's $12 billion TVL drop following the KelpDAO exploit in April 2026. Buterin's proposal addresses a structural problem with a structural solution: eliminate debt, eliminate liquidations.
The economic obstacles are quantifiable. Rebalancing costs of 2%+ annually compete directly with CDP yields. Institutional capital — represented by Morpho's $175 million raise — is flowing into existing CDP infrastructure, not options-based experiments. Governance incentives at Aave and Sky favor the status quo.
But the liquidation model's failure modes are also quantifiable: $3 billion liquidated in three days, $12 billion in deposit outflows from a single exploit cascade, $8.32 million seized for $0. The question is not whether the options model is theoretically superior — Buterin's math eliminates liquidation risk by construction. The question is whether 2% annual rebalancing drag is a lower cost than the tail risk of cascading liquidation failures. For most of DeFi's current $54 billion in lending deposits, that calculation has not yet been made.