Crypto protocols spent over $1.4 billion on token buybacks in 2025, according to CoinGecko data. Hyperliquid alone accounted for $645 million — nearly half the total. In 2026, that figure is accelerating: Hyperliquid's Assistance Fund now runs at approximately $1.3 billion annualized, Aave's DAO ...
"Looking ahead to 2026, experts broadly expect buybacks to stick around in a more disciplined form, as more protocols reach product-market fit and generate revenue." — The Block Research
Crypto protocols spent over $1.4 billion on token buybacks in 2025, according to CoinGecko data. Hyperliquid alone accounted for $645 million — nearly half the total. In 2026, that figure is accelerating: Hyperliquid's Assistance Fund now runs at approximately $1.3 billion annualized, Aave's DAO has made its $50 million annual buyback permanent, and deBridge has committed 100% of protocol revenue to open-market DBR purchases. Meanwhile, Pendle abandoned its vote-escrowed lock model entirely, replacing vePENDLE with liquid sPENDLE staking — a structural shift in how governance tokens distribute fees.
Yet the data on price impact is mixed. Per CoinGecko, the average token subject to a buyback program in 2025 saw no statistically meaningful outperformance. Aave's token declined 27% relative to its average buyback price. Jupiter spent over $70 million on buybacks and is now rethinking the approach. The structural question is not whether buybacks work in theory — it is whether the entities executing them are aligned with token holders, or whether the buyback mechanism obscures a corporate structure that primarily benefits equity shareholders and insiders.
Development activity around buyback infrastructure is measurable but fragmented. A GitHub search for "token buyback" repos sorted by update recency surfaces several projects:
On the governance infrastructure side, m0-foundation/ttg (Two Token Governance, 11 stars) implements a dual-token voting system in Solidity for maintaining lists and managing communal property. The last substantive commits date to May 2024, suggesting the governance framework is stable but not actively evolving.
The most actively maintained treasury tooling repo is BobTheBuidler/dao-treasury, a financial reporting suite for DAOs with Grafana integration, updated as recently as June 13, 2026. This reflects a broader trend: DAO treasury management is professionalizing, with monitoring and reporting infrastructure catching up to the billions in assets under governance.
The pattern across GitHub is clear — buyback tooling is proliferating at the infrastructure layer, but most repos are low-star, early-stage projects. The production-grade buyback systems (Hyperliquid, Aave, Uniswap) are proprietary smart contracts, not open-source libraries.
According to CoinGecko research, 28 notable projects allocated meaningful spending toward token buybacks in 2025, totaling over $1.4 billion. Monthly average spending was approximately $145.93 million, with an 85% month-over-month surge in July 2025.
The concentration is extreme. Hyperliquid's $645 million represented 46% of all buyback spending. LayerZero executed a one-time $150 million ZRO buyback. Metaplex spent $13.78 million (half of protocol revenue) to repurchase 6.5% of total MPLX supply. The remaining 25 protocols split what was left.
Per DWF Labs research, the 2025 buyback trend stemmed from a convergence of profitability, governance maturity, and market psychology. But the effectiveness data tells a different story: Aave's token declined 27% against its average buyback price. Jupiter spent $70 million with inconclusive results. Helium paused its program entirely after observing no market impact.
The parallel to traditional equity markets is instructive but imperfect. In TradFi, share buybacks reduce outstanding shares and mechanically increase earnings per share. In crypto, buybacks typically purchase tokens that are either burned (permanent supply reduction) or held in treasury (temporary supply removal). The governance and legal frameworks surrounding these mechanisms remain underdeveloped. Token holders have no legal claim on cash flows equivalent to shareholder rights — buybacks are discretionary and can be paused or redirected by governance vote or foundation decision at any time.
Hyperliquid's Assistance Fund deploys 97-99% of protocol trading fees into continuous HYPE purchases. Per Tokenomics.com, the fund generates approximately $65 million monthly in holder revenue, with $825 million cumulative since launch. The annualized buyback rate of approximately 7% of market cap is 4-5x that of Ethereum or BNB.
Key structural features: no venture capital allocation, 31% genesis airdrop (immediately liquid), and core contributor tokens vesting through 2028. The absence of VC investors means no private shareholders extracting value ahead of token holders. However, the Hyper Foundation controls 6% of supply and the 38.88% community rewards pool remains unclaimed — a potential overhang. The Assistance Fund has spent over $1.3 billion on buybacks by May 2026, holding roughly 28.5 million tokens, per crypto.news.
The Aave DAO approved a permanent $50 million annual buyback program, per The Defiant. The proposal, authored by the Aave Chan Initiative (ACI) and passed with 100% approval, authorizes the Aave Finance Committee to purchase $250,000 to $1.75 million in AAVE weekly. A pilot phase between May and November 2025 bought over 94,000 AAVE for $22 million.
Aave's annualized revenue runs at approximately $152 million based on $12.7 million in trailing 30-day revenue. The $50 million buyback commitment represents roughly one-third of protocol revenue. The Collector contract has aggregated $190 million in protocol revenue through Q1 2026. Aave's structure features a Swiss-based foundation (Aave Companies, now Avara) that retains equity ownership of certain IP and products, creating a dual-track value accrual: token holders receive buybacks, equity holders retain corporate upside.
After years of governance debate, the Uniswap DAO activated the "UNIfication" fee switch in late December 2025, per The Defiant. The mechanism routes trading fees on eight chains into UNI purchases and burns, estimated at $27 million in annualized revenue. Simultaneously, 100 million UNI tokens were burned.
The DAO also approved a 20 million UNI annual growth budget starting January 2026. Uniswap Labs, the corporate entity behind the protocol, separately earns revenue through its front-end fee (0.15% on swaps via the Uniswap interface). This creates a structural split: protocol-level fees go to token holders via the fee switch, while interface-level fees accrue to Uniswap Labs shareholders. The Uniswap Foundation, with a $4.8 billion treasury (the largest in DeFi), operates as a distinct governance-funded entity.
Pendle executed one of the most significant governance model transitions in DeFi in January 2026. Per CoinDesk, the protocol abandoned the vote-escrowed (ve) model — where token holders locked PENDLE for up to two years — in favor of sPENDLE, a liquid staking token with a 14-day withdrawal period (or instant exit at a 5% fee).
The economic model is material: over 80% of protocol revenue flows to sPENDLE holders through PENDLE buybacks and fee distribution. The protocol collects a 3% fee on all yield accrued by YT holders and 80% of swap fees allocated to pool voters. The transition includes an algorithmic emissions model expected to cut overall token emissions by 30%.
For existing vePENDLE holders, a snapshot on January 29, 2026, determined eligibility for virtual sPENDLE credits with up to 4x multipliers based on remaining lock period. This represents a rare case of a protocol upgrading its governance model while compensating locked participants rather than forcing exit at a loss.
The deBridge Foundation launched a Reserve Fund that channels 100% of protocol revenue into open-market DBR purchases, per The Block. Since launching in June 2025, the fund has acquired 1.3% of DBR's total 10 billion supply, worth approximately $3 million. The broader treasury holds $30.1 million across DBR, USDC, SOL, and ETH.
Notably, deBridge implements yield-bearing strategies on idle assets: USDC deposited into Aave on Ethereum, ETH staked via Lido, and SOL and USDC deployed on Kamino Finance on Solana. A public dashboard tracks every purchase and balance in real-time. The "hold rather than burn" model means the foundation retains optionality — it can deploy these tokens for grants, liquidity incentives, or additional yield. Token holders should monitor whether the Reserve Fund's DBR holdings are ever sold or redistributed.
Ethena approved fee switch parameters in Q1 2026, directing 10-20% of protocol revenue to sENA stakers, per CryptoRank. The protocol generates $50-60 million monthly in fees from its USDe synthetic dollar operations, implying 4.5-15% annualized yield on $750 million in staked ENA. Two rounds of buybacks in 2025 totaled $890 million in committed capital ($360M in July, $530M in September), with an additional $500 million unlocked for future purchases.
The scale of Ethena's buyback commitment — $890 million already committed plus $500 million in reserve — dwarfs most DeFi protocols. However, the 10-20% revenue share to sENA stakers is conservative relative to Pendle's 80% or Hyperliquid's 97-99%. The majority of protocol revenue remains with the Ethena Foundation and Ethena Labs, the corporate entities behind the protocol.
Fluid's DAO initiated a buyback program in October 2025, directing 100% of Ethereum mainnet revenue to FLUID repurchases at an estimated $1.3-1.5 million monthly. However, the program was paused in March 2026 after Fluid had to cover $70 million in bad debt from the Resolv Protocol exploit, per MEXC. The protocol had already repurchased approximately 1.3% of total supply before suspension. FLUID token emissions were also significantly reduced or eliminated. The case illustrates a key risk: buyback programs funded by protocol revenue are subordinate to solvency requirements.
DAO treasuries collectively hold over $26 billion as of Q1 2026, per PatentPC data. The five largest: Uniswap ($4.8B), Sky/MakerDAO ($3.9B), Optimism ($2.1B), Arbitrum ($1.7B), and Lido ($1.4B).
The structural tension between corporate entities and token holders is sharpening. According to Pantera Capital, multiple high-profile cases in 2025-2026 exposed misalignment — including acquisitions and restructurings where token holders received no direct compensation (Tensor, Axelar). Tokens confer governance rights and sometimes economic claims, but they do not confer the legal protections of equity.
Buyback programs operate in this ambiguous zone. When Aave Labs (Avara) generates revenue from its consumer products while the Aave DAO funds buybacks from protocol revenue, value accrues to different constituencies through different channels. When the Uniswap Foundation deploys a $4.8 billion treasury while Uniswap Labs charges a separate front-end fee, the protocol's economic output is split between token-governed and equity-governed entities.
June 2026 token unlocks add supply pressure. Over $634 million in tokens unlock in the second week of June alone, per BeInCrypto. Jupiter unlocks approximately 53.47 million JUP monthly (~1.68-1.78% supply increase). Sahara AI faces a 9.58% circulating supply unlock on June 26. These unlocks primarily benefit insiders, team members, and early investors — diluting existing token holders even as buyback programs attempt to offset sell pressure.
Lido presents a particularly stark case: $288 million in cumulative protocol revenue, zero distributed to LDO holders to date. The governance community is actively discussing revenue sharing, but as of mid-2026, LDO remains a governance-only token while Lido's revenue accrues to the protocol treasury.
The protocols analyzed fall into three tiers of token holder value return:
Tier 1 — Direct, Sustained Revenue Return:
Tier 2 — Partial Revenue Return with Corporate Leakage:
Tier 3 — Announced but Unproven or Paused:
The corporate structure angle is determinative. Protocols without separate equity-backed corporate entities (Hyperliquid, deBridge) return a higher share of revenue to token holders. Protocols with dual-entity structures (Uniswap Labs/Foundation, Aave/Avara) split value across constituencies, with token holders receiving a fraction of total economic output.
The crypto industry's adoption of buybacks mirrors TradFi's trajectory with a critical structural difference: token holders have no legal claim on the cash flows funding these programs. Buybacks are governance-approved, foundation-discretionary, and can be paused overnight. Hyperliquid's model — no VC allocation, 97-99% fee return, automated execution — is the closest the sector has come to genuine shareholder-equivalent value accrual, and its $1.3 billion annualized pace reflects it. But Hyperliquid is the exception. Most protocols operate dual-entity structures where corporate labs capture interface fees, IP monetization, and equity upside while DAO treasuries fund buybacks from protocol revenue. Until token holders obtain enforceable economic rights — through regulation, legal precedent, or binding smart contract guarantees — buyback programs remain voluntary acts of alignment, not structural obligations. The data shows the market has not yet rewarded them accordingly.