Cumulative token buybacks across major DeFi protocols have surpassed $1 billion, led by Hyperliquid's $920M+ assistance fund and Sky Protocol's $106M in SKY repurchases. In the past 30 days alone, Uniswap expanded its fee switch to eight new chains (est. $27M annualized uplift), Aave locked in a ...
"We're not speculating on token price—we're securing access to DeFi infrastructure and governance." — Analyst commentary on BlackRock's UNI purchase, per The Block
Cumulative token buybacks across major DeFi protocols have surpassed $1 billion, led by Hyperliquid's $920M+ assistance fund and Sky Protocol's $106M in SKY repurchases. In the past 30 days alone, Uniswap expanded its fee switch to eight new chains (est. $27M annualized uplift), Aave locked in a permanent $50M/year buyback budget, dYdX escalated its allocation to 75% of net protocol fees, and Ethena activated its sENA fee switch with yields of 4.5–15%. At the same time, BlackRock, Citadel Securities, and Apollo Global Management disclosed DeFi governance token purchases—marking the first institutional wave of direct participation in protocol governance through token ownership.
The report below catalogs eight protocols that have activated or expanded value-return mechanisms since Q4 2025, compares their structural approaches (burn vs. stake vs. treasury reserve), and assesses who actually captures the value: token holders, corporate shareholders, or foundation treasuries. The data shows a clear shift from governance-only tokens toward revenue-linked instruments, but the structural gap between token holder rights and corporate equity remains wide.
Development activity around buyback and fee-switch infrastructure is accelerating, though the tooling remains immature relative to the capital at stake.
Autonomous Buyback Contracts (Starknet/Cairo). The repo loothero/autonomous_buyback shipped a v2 buyback component in January 2026, integrating with Ekubo's TWAMM (time-weighted AMM) for programmatic token repurchases on Starknet. The latest commits include StreamComponent with ERC20 burn support and 18 unit tests covering burn mechanics, allowance patterns, and event emissions. This represents an early attempt to make buybacks composable at the smart contract layer rather than relying on multisig-executed market orders.
Uniswap Fee Switch Analysis. OmarOlek/uniswap-fee-switch-analysis (updated Feb 19, 2026) provides wallet-level LP behavior analysis after Uniswap's December 2025 fee switch activation, built on Dune Analytics. While the repo has zero stars, it signals that independent researchers are already building measurement tools around fee switch impacts—a prerequisite for data-driven governance decisions.
LIT Buyback Dashboard. trevorflipper/lit-buyback-dashboard was updated on Feb 27, 2026, providing an auto-updating dashboard tracking LIT token buybacks and revenue. This pattern—protocol-specific buyback transparency tooling—is replicating across projects.
Jubilee Protocol Treasury AI. Jubilee-Protocol/Openclaw-Skill-Jubilee (updated Feb 23, 2026) enables AI agents to manage yield-bearing vaults (jBTCi and jUSDi on Base). The convergence of AI agent frameworks with treasury management is an emerging trend, though production readiness remains unproven.
Broader trend. GitHub searches for "token buyback," "fee switch," and "value accrual" return significantly more results updated in 2026 than in prior years. However, star counts remain low (0–1 range for most repos), indicating early-stage development rather than production-grade infrastructure.
Eight protocols have active buyback programs with disclosed parameters. The table below summarizes their mechanisms and scale.
| Protocol | Mechanism | Revenue Allocation | Cumulative Buybacks | Token Destination | |----------|-----------|-------------------|---------------------|-------------------| | Hyperliquid | Auto buy-and-burn | 97% of trading fees | $1B+ (37M+ HYPE) | Burned / Assistance Fund | | Sky (MakerDAO) | Revenue-funded buyback | Portion of $611.5M est. 2026 revenue | $106M+ | Treasury / stakers | | Aave | Permanent buyback program | $50M/year from protocol revenue | 94,000+ AAVE retired | Stakers via Safety Module | | Uniswap | Fee switch + burn | $34M annualized (expanding to ~$61M) | $5.5M+ UNI burned | Burned (supply reduction) | | dYdX | Revenue-to-buyback | 75% of net protocol fees | 2.87M DYDX ($1.88M) | Staked (100%) | | Jupiter | Revenue-funded buyback | 50% of revenue ($10–20M/yr) | ~$70M spent (2025) | Under review—may halt | | Ethena | Fee switch to stakers | Direct to sENA holders | Active (fee switch live) | Stakers (sENA) | | Treehouse | 50% of MEY fees | 50% of tETH yield fees | First round complete | DAO treasury reserves |
Hyperliquid leads on absolute scale. According to DL News, the Hyper Foundation is proposing to permanently burn approximately $920M in HYPE from the Assistance Fund, which would remove roughly 13% of circulating supply. On February 5 alone, 160,750 HYPE were repurchased from $5.25M in daily revenue, per The Defiant. The mechanism is automated: 97% of trading fees route directly to buyback and burn with no governance vote required per transaction.
Sky Protocol projects $611.5M in gross protocol revenue for 2026 (81% YoY growth) and $157.8M in protocol profits (198% YoY), according to Global Fintech Series. Cumulative buybacks have surpassed $106M. The USDS circulating supply is projected to nearly double to $20.6B, which would expand the revenue base for further buybacks.
Jupiter's cautionary tale. Despite spending approximately $70M on buybacks throughout 2025, JUP declined 89% from peak levels, per crypto.news. Founder Siong publicly questioned the strategy, noting that monthly unlocks of 53M JUP through June 2026 increased circulating supply by roughly 150% since launch—overwhelming the buyback flow. Jupiter is now considering halting buybacks entirely and redirecting funds to user incentives, per Lookonchain. This case demonstrates that buybacks without supply discipline are capital destruction.
The "fee switch" model—routing protocol revenue directly to token holders rather than through buybacks—gained significant momentum in late 2025 and early 2026.
Uniswap's UNIfication. The UNIfication proposal, passed in late 2025, activated Uniswap's long-debated fee switch, per Blockworks. The mechanism burns UNI rather than distributing fees, with a retroactive burn of 100M UNI from the treasury to compensate for years of zero value accrual. Early annualized data shows $34M in protocol fees now routing to UNI burns, per Coin Metrics. A governance vote ending February 23, 2026 proposes expanding the fee switch across eight additional chains and all v3 pools, which could lift annualized revenue to approximately $61M, according to CoinDesk. UNI jumped 15% on the news.
Ethena's sENA activation. Ethena's fee switch is now live, directing protocol revenue to sENA stakers at estimated yields of 4.5–15% annualized, based on current monthly fees of $50–60M, per Blockworks. The mechanism was proposed by Wintermute in November 2024 and required meeting specific benchmarks before activation. A significant caveat: 333M ENA tokens unlock on March 2, 2026, which could dilute staking yields and create selling pressure, per CryptoRank.
Pendle's vePENDLE-to-sPENDLE transition. In January 2026, Pendle replaced its vote-escrow model (vePENDLE) with liquid staking (sPENDLE), per Pendle's Medium blog. The new system requires only a 14-day withdrawal period (or instant exit for a 5% fee) and commits up to 80% of protocol revenue to PENDLE buybacks distributed as governance rewards. Despite generating over $37M in 2025, the previous vePENDLE system concentrated rewards among a small fraction of sophisticated users. The sPENDLE model broadens access—at the cost of reducing lock-up duration and, by extension, the capital commitment that underpinned governance security. Existing vePENDLE holders receive a transitional boost (up to 4x) that decays over two years.
Beyond the blue chips, several smaller protocols are implementing revenue-sharing models that merit attention.
Treehouse launched its TREE token buyback program in late 2025, committing 50% of Market Efficiency Yield (MEY) fees from its tETH fixed-income product to open-market TREE purchases, per PR Newswire. The program passed governance with 99.59% approval. With $294M in deposits, Treehouse is a mid-tier protocol by TVL, but its approach is notable: acquired TREE tokens are held in a DAO-controlled multisig as strategic reserves rather than burned. This preserves future optionality (incentive programs, liquidity mining) but does not reduce circulating supply. According to DL News, the buyback mandate covers retrospective revenue since tETH's inception, and future governance proposals may extend it to additional revenue sources.
Maple Finance (SYRUP) ended its staking rewards program in October 2025, with 99% of stSYRUP voting in favor, per The Defiant. The protocol now allocates 25% of lending revenue to SYRUP buybacks. Looking ahead, Maple's 2026 roadmap includes "Builder Codes"—permissionless integrations allowing partners to embed syrupUSDC and syrupUSDT into mainstream platforms with configurable revenue share, per Maple Finance. The shift from staking rewards to buybacks reflects a broader pattern: protocols preferring deflationary mechanics over inflationary yield.
dYdX escalated its buyback allocation from an initial 25% to 75% of net protocol fees in November 2025, per CoinDesk. Since launch, the program has acquired 2.87M DYDX ($1.88M), with 100% of purchased tokens staked. The remaining 25% of fees split between the Treasury SubDAO (5%), MegaVault (5%), and operational costs. For 2026, dYdX is introducing trading fee discounts linked directly to DYDX staking, creating a utility loop beyond pure value return. However, the $1.88M cumulative spend remains modest relative to protocol scale.
February 2026 marks the first coordinated entry of major traditional finance firms into DeFi governance through direct token ownership.
BlackRock made its $2.2B tokenized Treasury fund (BUIDL) tradable on Uniswap and purchased an undisclosed amount of UNI governance tokens, per Fortune. UNI surged 25% on the announcement, per CoinDesk. The purchase is strategic, not speculative: BlackRock now has governance influence over the protocol through which its institutional fund trades.
Apollo Global Management secured the ability to purchase up to 90M MORPHO tokens (9% of total supply) over four years, per The Block. Morpho's governance-minimized design—permissionless lending with vault curators—represents the opposite end of the governance spectrum from Uniswap's active DAO. Apollo's interest suggests TradFi values the protocol's infrastructure regardless of governance structure.
Citadel Securities supported the launch of LayerZero's "Zero" blockchain and received ZRO tokens, per The Block.
The corporate structure implication is significant. These firms are buying governance tokens—not equity in the companies that build protocols. Value accrual for these institutional participants depends entirely on the token-level economics (fee switches, buybacks, burns), not on corporate revenue or dividends. This aligns institutional incentives with retail token holders for the first time in a meaningful way, though voting power concentration is an emerging concern.
World Liberty Financial (WLFI) introduced a governance staking proposal requiring 180-day token lockups for voting rights, with a 2% annual return from the treasury for active voters and a minimum 50M WLFI stake for "Super Node" status, per HOKANEWS. The proposal enters a seven-day vote requiring participation from 1B eligible tokens. This tiered governance model is unusual—most protocols use one-token-one-vote without minimum thresholds.
Two events this month illustrate the spectrum of treasury management outcomes.
Step Finance ceased all operations following a $27M treasury hack on January 31, 2026, per CoinDesk. Attackers compromised executive devices and extracted 261,854 SOL from treasury and fee wallets. The STEP token collapsed 96% to $0.00057. Step Finance, SolanaFloor, and Remora Markets all shut down, per Defi Planet. A post-mortem buyback program for STEP holders using a pre-hack snapshot has been committed, but the protocol is functionally dead. This case underscores that treasury concentration in hot wallets managed by a small team remains an existential risk for DAO-controlled funds.
The Ethereum Foundation took the opposite approach, announcing a strategy to stake 70,000 ETH (~$128M) from its treasury, starting with an initial deposit of 2,016 ETH, per CoinDesk. At an estimated 2.8% yield, this generates approximately $3.6M annually. The foundation targets annual spending of 15% of its treasury while maintaining a 2.5-year operational runway. Infrastructure uses minority clients spread across multiple countries to support client diversity, per CoinDesk. This is not a value-return to ETH holders—it is a foundation sustainability play. The distinction matters: ETH holders do not benefit directly from foundation staking yields.
The central question: where does the money go?
Direct to token holders (high alignment):
Partial alignment (buyback with structural ambiguity):
Low alignment (token holders secondary):
No alignment:
The data supports a clear thesis: DeFi protocols are converging on buyback and fee-switch mechanisms as the standard model for token holder value return. Cumulative buybacks have crossed $1B, fee switch activations are accelerating, and institutional capital is now entering through governance tokens rather than corporate equity. This represents a structural maturation of the sector.
However, the gap between token holder value and corporate shareholder value remains unresolved. Uniswap Labs, Aave Labs, and dYdX Trading Inc. all maintain independent revenue streams that do not flow through the token. Buybacks improve the token's economic profile but do not confer dividend rights, board seats, or legal claims on protocol assets. TradFi firms entering through governance tokens—not equity—is notable precisely because it suggests these firms see protocol-level governance as more valuable than corporate ownership in certain contexts.
The Jupiter case should temper enthusiasm: $70M in buybacks are worthless when supply inflation exceeds demand. Protocols that pair buybacks with emission discipline (Hyperliquid's 97% auto-burn, Uniswap's retroactive 100M UNI burn) are structurally better positioned than those running buybacks against aggressive unlock schedules. March 2026's $6B+ in token unlocks will test this thesis directly.