Toward Finance Without "Forced Liquidations": Decoding Buterin's DeFi Redesign Without the Jargon

What Is a "Forced Liquidation," Anyway?

In DeFi, you can deposit a crypto asset (say, ETH) as "collateral" and borrow money (such as a synthetic-dollar stablecoin) up to a fraction of its value—much like borrowing against your home. The problem comes when the collateral's price falls. If its value drops below a set line relative to the loan, the system automatically sells the collateral to recover the lender's funds. That is a "forced liquidation"—you are sold out mechanically, regardless of your intentions.

Why Forced Liquidations Shake the Whole Market

In calm markets, this works fine. The danger is in a sharp drop. When prices fall suddenly, many users' collateral breaches the liquidation line at once. A wave of forced selling follows, which pushes prices down further and triggers the next round of liquidations—a domino effect. Indeed, in the first week of June 2026, roughly US$10 billion was liquidated as BTC and ETH each fell more than 30% over 30 days. That slide was reportedly set off by US$2.7 billion flowing out of US BTC and ETH ETFs in May.

A second weak point is the "oracle." A blockchain cannot see outside prices on its own, so it relies on a mechanism (an oracle) to report them. Liquidations fire automatically based on the real-time price the oracle provides. If that oracle reports a wrong or manipulated value even for a moment, unnecessary liquidations can cascade. Buterin has long warned that real-time oracles, which depend on a handful of sources, are exactly the "soft spot" attackers target.

Buterin's Proposal: Build on Options, Not Debt

The core of the new design is simple. One ETH is split into two coupons, "P" and "N," whose values shift with price: P benefits when the price rises, N benefits when it falls—and the two always add back to exactly one ETH. Picture slicing a single pie into two pieces by a rule fixed in advance, where the actual cut happens only at maturity.

With this design, no matter how prices move, the total claims never exceed the original one ETH—so a "shortfall, i.e., insolvency" cannot occur by construction. Because there is no insolvency, the liquidation engine itself becomes unnecessary. And since settlement happens just once at maturity, the system can run on a "slow oracle" that allows time for dispute resolution, rather than depending on high-frequency real-time price feeds. That is the heart of Buterin's argument.

Why Make This Case Now?

Recent events underscore the urgency. In April 2026, lending giant Aave saw US$8.45 billion flow out within 48 hours after a KelpDAO-related exploit (about US$292 million). Unbacked tokens had been used as collateral, and liquidating them normally risked destabilizing the whole system, forcing Aave into emergency measures that froze the affected markets. As of June 11, Aave's deposits remained at roughly half their pre-hack level. The fragility of "works in calm, breaks in crisis" is precisely what makes the proposal feel necessary.

It is not a cure-all, however. Without liquidations, a position left untouched gradually drifts from its target, so users must periodically rebalance—and the associated fees and costs become a new point of debate. The proposal is still at the research-and-testing stage and has not been adopted by major protocols such as Aave.

[Business Development Insights]

1. Systems should be designed around "stress conditions," not "normal conditions." A mechanism that runs smoothly day-to-day but collapses only in a crisis is itself a structural flaw. Across any business or product, starting the design from worst-case behavior under maximum load—prioritizing robustness in emergencies over efficiency in calm times—is what builds durable trust.
2. Dependence on a "single data source or single trigger condition" becomes the system's vulnerable point. Highly interlinked designs that rely on one input—like real-time price data or automated execution that fires en masse once a threshold is crossed—collapse in a chain reaction when that point is attacked. Building in redundancy, time buffers, and room for dispute creates a structure that does not topple in one blow.
3. Innovation often comes from "repurposing" mature existing frameworks rather than building from scratch. This proposal applies long-standing options concepts from finance, and the mechanics of already-operating prediction markets, to a different domain. Reusing established mechanisms lets you share infrastructure and raise both security and development speed at once. In new ventures, the habit of asking "can I bring in a proven technology from an adjacent field?" is valuable.

[Sources]

• NADA NEWS, "Vitalik Buterin proposes a new plan to eliminate forced liquidations in DeFi" (June 12, 2026)
• CoinDesk, "Ethereum's Vitalik Buterin is rethinking how DeFi handles market crashes" (June 1, 2026)
• Cryptopolitan, "Options over debt: Vitalik proposes abandoning CDPs and liquidations in DeFi"
• Crypto Times, "Vitalik's Options-Based DeFi Moves From Idea to Testnet" (June 11, 2026)
• Memeburn / Crypto Economy (the P + N = 1 ETH design, slow oracles, the rebalancing issue)
• KuCoin News, "Vitalik Proposes a New DeFi Risk Model to Replace Flash Liquidations"
• CoinMarketCap (~US$10 billion liquidated in the first week of June); DeFiLlama (Aave TVL trend)