Balancer, a decentralized exchange and automated market maker, has proposed winding down the protocol after a $128 million exploit in November 2025 and a subsequent restructuring failed to restore revenue. The proposal, authored by Balancer Labs CEO Marcus Hardt and published on the Balancer governance forum on Monday, calls for an orderly shutdown and distribution of the protocol's remaining treasury, valued at more than $9 million, to BAL tokenholders.
The plan follows Balancer Labs' shutdown in March, when executives chose to continue operating the protocol under a leaner structure. Hardt said the restructuring reduced costs and delivered products promised to tokenholders, but revenue did not recover sufficiently. He said the protocol's profitability problem reflected challenges faced by several DeFi protocols this year.
Hardt said the missing element was sufficient revenue, with most protocol revenue still coming from v2 while v3 revenue had not grown enough to replace it. He said the product worked but did not sell enough, and added that he underestimated how much the November exploit would continue to limit adoption.
DefiLlama data show Balancer's monthly protocol revenue fell to $371,000 in November from $1.13 million in October after an exploit affecting composable stable pools on the legacy v2 protocol. Revenue continued to decline into 2026, reaching $56,781 in August. Hardt said the November 2025 exploit hit legacy v2 pools and that, although v3 uses a different architecture, the incident followed the protocol's name into subsequent conversations and made traction harder to build.
Under the proposal, Balancer would begin a phased shutdown next month. New business development would end, and liquidity providers would have until Oct. 30 to prepare to exit. Pools that can be paused would move to withdrawal-only, while pools that cannot be paused would continue operating with the protocol fee set to zero where contracts allow. From Nov. 1, Balancer would operate only the minimal infrastructure needed to support withdrawals, and the DAO would be wound down, with a small team managing the transition. The proposal sets aside up to $400,000 for the wind-down process.
BAL holders would receive the remaining treasury on a pro-rata basis. The first distribution is scheduled for May 2027, when holders would burn BAL in exchange for their share of treasury assets. A second distribution would return unspent wind-down funds and unclaimed assets from the first distribution, followed by a final sweep six months later.
Hardt said delaying the wind-down would consume the treasury without changing the eventual outcome. He said continuing on the current path would spend the treasury to arrive at the same place later, and that the treasury belongs to BAL holders. The question, he said, is whether remaining assets reach holders while they are still substantial or are spent on a path that has already been tried.
The wind-down requires approval from BAL holders, with a snapshot vote scheduled for Sept. 25 to 29. If rejected, Balancer's existing operating framework would remain in place.
