DeFi protocols have now collectively spent over $2 billion on token buybacks since late 2024. Calendar year 2025 alone accounted for $1.4 billion in buyback spending, according to [The Defiant](https://thedefiant.io/news/research-and-opinion/crypto-token-buybacks-surge-fivefold-as-projects-return...
"BLabs, as a corporate entity, has become a liability rather than an asset to the protocol's future." — Fernando Martinelli, Co-founder, Balancer — DAO governance post, March 2026
DeFi protocols have now collectively spent over $2 billion on token buybacks since late 2024. Calendar year 2025 alone accounted for $1.4 billion in buyback spending, according to The Defiant and Bitget. Q1 2026 has seen the trend accelerate: Lido proposed a $20M one-off buyback, Optimism committed 50% of sequencer revenue (~$8M annualized) to OP repurchases, Aave adjusted its annual buyback budget from $50M to $30M, Balancer put forward a $3.6M buyback-and-burn, and Uniswap activated its fee switch with an annualized burn rate of approximately $34M. The largest single program by far is Hyperliquid, whose Assistance Fund has repurchased $890M of HYPE tokens and proposed burning 13% of circulating supply.
The critical question is whether buybacks accrue value to token holders or primarily serve as price support while corporate insiders and vesting-schedule beneficiaries exit positions. Jupiter's $70M buyback against an 89% price decline offers a data point. So does Balancer Labs' corporate dissolution, which eliminates the legal entity while shifting 100% of protocol revenue — and 100% of legal ambiguity — to the DAO. The evidence assembled here suggests that buyback effectiveness is entirely a function of net supply dynamics, not nominal spend.
Developer activity around buyback infrastructure provides a leading signal on adoption trajectory. The following repositories indicate that buyback tooling is moving from ad-hoc multisig operations to commoditized, deployable infrastructure.
| Repository | Description | Last Updated | Signal | |---|---|---|---| | m0-foundation/ttg | Two Token Governance — dual-token voting for communal property management | Dec 2025 | Governance innovation separating economic and voting rights (11 stars, Solidity) | | loothero/autonomous_buyback | Cairo library for autonomous token buybacks via Ekubo TWAMM on Starknet | Jan 2026 | Cross-chain buyback automation; latest commit "feat: v2 buyback component with enhanced features" co-authored with Claude Opus 4.5 | | Alex000115/treasury-tax-harvester | Professional treasury management module integrating Uniswap V3 for automated fee-to-token conversion | Mar 2026 | Production-grade TaxHarvester.sol contract for protocol-level treasury automation | | esskslifetech/BagsAI-Agent-Forge | Framework for creating AI agents on Solana with automatic fee splitting and token buybacks for holders | Apr 2026 | AI-agent-mediated buyback distribution on Solana |
Two patterns are worth noting. First, the autonomous_buyback repository's v2 commit was co-authored with an AI model, indicating that AI-assisted DeFi development has reached buyback infrastructure. Second, the treasury-tax-harvester contract provides a turnkey module for any protocol to automate Uniswap V3-based buybacks — the kind of tooling that converts a one-off governance proposal into a persistent, programmatic treasury operation. The emergence of buyback-specific tooling on GitHub signals that this infrastructure is commoditizing. Any protocol can now deploy automated buyback smart contracts without building from scratch.
The following table summarizes active and proposed buyback programs with disclosed data as of April 26, 2026.
| Protocol | Mechanism | Scale | Supply Impact | Status | |---|---|---|---|---| | Hyperliquid (HYPE) | Assistance Fund repurchases | $890M spent; 40M HYPE (~4% total supply) | Proposed HIP-3 burn of 13% circulating supply | Active; $1.7M/week as of Jan 2026, accelerating 26% WoW | | Raydium (RAY) | Cumulative buybacks | ~$196M; ~71M RAY | 26.4% of circulating supply retired | Active | | LayerZero (ZRO) | Treasury repurchases | $150M; 5% of total supply | Supply reduction | Active | | Aave (AAVE) | Adaptive weekly purchases ($250K-$1.75M) | $50M/yr approved, adjusted to $30M/yr (Mar 2026) | ~$15.7M spent by Aug 2025 | Active; budget under review | | Uniswap (UNI) | Fee switch burn-to-withdraw | ~$34M annualized burn rate; $5.5M+ burned to date | 100M UNI retroactive treasury burn (~$596M) | Active since Mar 2026 on-chain vote | | Lido (LDO) | One-off treasury buyback | $20M via 10,000 stETH | Could retire ~8% circulating supply | Proposed Mar 2026 | | Optimism (OP) | 50% Superchain sequencer revenue | ~$8M annualized; 12-month pilot | Tokens to Collective Treasury | Approved Jan 2026 (84.4% vote) | | Balancer (BAL) | Buyback and burn | $3.6M (~35% DAO Treasury); ~22.7M BAL at NAV (~$0.16) | ~35% circulating supply | Proposed; opens 12 months post-snapshot |
Hyperliquid dominates by volume, accounting for approximately 45% of the cumulative $2B total. Its Assistance Fund model routes trading fees into systematic HYPE repurchases, with weekly spending of $1.7M as of January 2026 and a 26% week-over-week acceleration rate, per DL News. The proposed HIP-3 would burn 13% of circulating supply, according to The Defiant.
Aave's program, initially approved at $50M annually and executed by the Finance Committee and TokenLogic with adaptive weekly purchases of $250K to $1.75M, was reduced to $30M in a March 2026 governance proposal, per the Aave Governance Forum. By August 2025, the protocol had spent approximately $15.7M, per The Defiant.
Lido's proposal targets a $20M buyback using 10,000 stETH, executed in batches of 1,000 stETH via CEX and market makers with a 3-day objection period per batch. At current prices, LDO trades at a 70% discount to its 2-year median and 95% below its all-time high, per CoinDesk. The buyback could retire approximately 8% of circulating supply.
For context: S&P 500 companies spent $982B on buybacks in 2024. DeFi's $2B is marginal in absolute terms but represents 5-10% of participating protocol treasuries — a materially higher proportion than corporate America's typical buyback-to-cash ratios.
The buyback wave runs parallel to a distinct but related mechanism: fee switches that route protocol revenue directly to token holders through burns or staking distributions.
Uniswap. The fee switch was approved in December 2025 with 99.9% governance support; on-chain votes concluded March 4, 2026, per The Defiant and DL News. The mechanism takes approximately one-sixth of swap fees from LPs and directs them to protocol revenue. Results to date: over $5.5M of UNI burned, an annualized pace of approximately $34M at current volumes. A retroactive burn of 100M UNI from the treasury (~$596M) was executed concurrently. Q1 2026 gross profit reached approximately $3.12M — effectively the protocol's first profitable quarter. Uniswap generated $1.05B in trading fees across 2025, per Blockworks and KuCoin reporting.
Ethena. The fee switch activated after September 2025 benchmarks were met, routing revenue to sENA stakers, per Blockworks. A concurrent $890M token buyback program (DAT) launched in late 2025. Q1 2026 gross protocol revenue totaled $65.06M, down 32% quarter-over-quarter. TVL declined approximately $130M since early March 2026, per Bankless and Cryptopolitan reporting. The combination of fee switch and buyback creates a dual accrual mechanism, though declining revenue raises questions about sustainability at current scale.
Pendle. In January 2026, Pendle migrated from vePENDLE (requiring 2-year locks) to sPENDLE (14-day withdrawal period), per The Block. Up to 80% of protocol revenue is now directed to PENDLE buybacks distributed to sPENDLE holders. Emissions were cut approximately 30% via algorithmic allocation. The structural redesign is sound, but the revenue base has contracted sharply: monthly revenue fell 87.6%, from $4.44M in August 2025 to $552K in March 2026, per ainvest. The saving metric is that buybacks currently exceed emissions by a 4.4x ratio. Pendle is diversifying revenue through Boros, a rates trading venue, to offset yield compression in its core market.
Jupiter spent $70M buying back JUP using approximately 50% of fee revenue in 2025. The token declined 89% from its peak, per crypto.news.
The arithmetic explains the outcome. Against the $70M in repurchases, Jupiter's token unlock schedule injected approximately $1.2B in new supply. Through June 2026, roughly 53M JUP unlock monthly. Circulating supply increased approximately 150% since launch. The buyback absorbed less than 6% of new issuance — a rounding error in supply terms.
Governance responses have included delaying a 400M JUP distribution until May 2026 and cutting a planned airdrop from 700M to 200M JUP. Solana co-founder Anatoly Yakovenko publicly suggested a lockup staking model as an alternative, per BeInCrypto.
The lesson is structural: buybacks without addressing supply inflation are cosmetic. When the token unlock schedule exceeds buyback capacity by an order of magnitude, repurchases function as exit liquidity for vesting-schedule beneficiaries rather than as value accrual for open-market holders. Jupiter's case is not an anomaly — it is the default outcome when buyback programs are designed without reference to net supply dynamics.
The Balancer case is the most consequential structural development in the current buyback wave, not because of the $3.6M buyback itself, but because of the corporate dissolution it accompanies.
On November 3, 2025, a v2 exploit drained $110-128M across six chains via a rounding flaw in swap logic, per CoinDesk. In March 2026, Balancer Labs announced its shutdown. Fernando Martinelli stated the corporate entity had become "a liability rather than an asset to the protocol's future." Core staff are transitioning to "Balancer OpCo," subject to a governance vote.
The linked proposals restructure protocol economics entirely. The DAO treasury captures 100% of revenue, up from 17.5%. Token emissions are eliminated. LPs receive 75% of swap fees. The $3.6M buyback-and-burn would retire approximately 22.7M BAL (~35% of circulating supply) at NAV of roughly $0.16, with redemption opening 12 months post-snapshot. A $500K compensation package was proposed for veBAL holders losing economic rights under the transition, per The Block.
This represents a potential template. The "progressive decentralization" narrative reaches its logical terminus when a corporate entity carries liability — in this case, from exploit lawsuits — and dissolution shifts risk to a legally ambiguous DAO. Token holders gain the economic upside of 100% revenue capture but inherit the legal and operational exposure previously borne by the corporate shell. The unresolved question: when there is no corporate entity, who do hack victims sue? The DAO has no legal personality in most jurisdictions. Dissolution does not eliminate liability; it disperses it into a vacuum.
Buyback programs can be assessed along two axes: direct value return (treasury-funded repurchases and burns) and structural value accrual (fee distribution to stakers via buyback mechanics).
Direct value return programs. Hyperliquid ($890M), Aave ($30-50M/yr), Uniswap ($34M/yr annualized burn), Optimism ($8M/yr), Lido ($20M one-off), and Balancer ($3.6M) represent the range. Hyperliquid is an outlier by an order of magnitude. The remainder cluster between $8M and $50M annually.
Fee distribution via buyback. Ethena routes revenue to sENA stakers. Pendle routes up to 80% of revenue to sPENDLE holders via buyback intermediation. Fluid employs an algorithmic buyback mechanism that adjusts based on FDV, targeting $30M in annualized revenue, with 100% of initial Ethereum mainnet revenue ($1.3-1.5M/month) directed to buybacks, per DeFiLlama and CoinMarketCap data.
The corporate extraction question. Buyback programs do not eliminate corporate value extraction; they operate alongside it. Uniswap Labs collects frontend fees separately from protocol fees captured by the fee switch. Lido's corporate structure (Lido DAO plus node operators) means buybacks do not prevent value leakage through operational layers. Optimism Foundation retains 50% of sequencer revenue — the other 50% funds the buyback. In each case, the buyback coexists with corporate or foundation revenue streams that are not subject to token holder governance.
Key insight. Buybacks are necessary but not sufficient for genuine value accrual. Jupiter demonstrates that without controlling the supply variable — unlocks, emissions, airdrops — buybacks transfer value from the protocol treasury to selling insiders. The programs that work (Hyperliquid, Uniswap) share two characteristics: they are funded by recurring fee revenue rather than treasury drawdowns, and they operate against a controlled or declining supply schedule.
Cumulative DeFi buybacks have crossed $2B, with Hyperliquid alone accounting for approximately 45% ($890M) of the total.
Fee switch activation is now a default expectation. Uniswap, Ethena, and Pendle all activated fee switches in late 2025 or early 2026. Protocols without a revenue-sharing mechanism face a governance credibility deficit.
Buyback effectiveness depends entirely on net supply dynamics. Jupiter's $70M program failed against $1.2B in token unlocks. The ratio of buyback spending to new supply issuance is the only metric that matters.
Corporate dissolution may become a template. Balancer's shift of 100% revenue to the DAO while eliminating the corporate entity establishes a precedent that other protocols carrying exploit or regulatory liability may follow.
Aave's budget reduction ($50M to $30M) signals buyback fatigue or treasury conservation, depending on interpretation. Declining lending fees drove the adjustment.
The sPENDLE model (short lockups plus buyback distribution) may replace traditional vote-escrowed tokenomics. Pendle's 14-day withdrawal period versus the prior 2-year lock reduces friction while maintaining a 4.4x buyback-to-emissions ratio.
GitHub shows buyback infrastructure commoditizing. Automated TWAMM and tax harvester contracts are proliferating, lowering the implementation barrier for any protocol considering a buyback program.
Circular value creation. Buybacks funded from treasuries that hold the protocol's own token create circular value — the equivalent of buying your own stock with your own stock. Only buybacks funded by external fee revenue (denominated in ETH, stablecoins, or other assets) represent genuine demand.
Regulatory risk. The SEC has signaled that buyback-and-burn mechanisms may constitute securities characteristics. Direct revenue distribution to token holders via fee switches further narrows the gap between governance tokens and securities.
Exploit risk. Balancer's $110M exploit destroyed more value in a single incident than years of buybacks could create. Smart contract risk remains the dominant downside exposure for any protocol, and buyback programs do not hedge against it.
Token unlock schedules. As the Jupiter case demonstrates, unlock schedules frequently overwhelm buyback programs. Protocols with large pending unlocks relative to buyback budgets are running a negative-sum program for open-market holders.
Revenue cyclicality. Pendle's 87.6% revenue decline and Ethena's 32% quarter-over-quarter contraction demonstrate how rapidly buyback capacity can erode. Programs sized for peak revenue conditions become undersized when volumes normalize.
Legal ambiguity of corporate dissolution. Balancer's model shifts liability to a legally ambiguous DAO structure. Hack victims and creditors may have no corporate entity to pursue. This is a feature from the protocol's perspective and a risk from the ecosystem's perspective.
The buyback wave represents DeFi's maturation toward TradFi-style capital return, but the data shows a wide spectrum of effectiveness. Hyperliquid and Uniswap demonstrate that buybacks paired with genuine fee generation and supply control can create sustainable value accrual. Jupiter's $70M failure demonstrates the opposite: when token unlocks dwarf repurchases, buybacks become exit liquidity for insiders.
Balancer's corporate dissolution raises the most consequential structural question of this cycle. As protocols shed their corporate shells, token holders gain economic upside — 100% revenue capture, elimination of corporate overhead — but inherit legal and operational risk previously borne by corporate entities. The dissolution does not remove liability from the system; it relocates it to a structure that most legal frameworks do not recognize. This trade-off will define the next phase of DAO governance design.
The protocols that will define whether buybacks survive as a credible value accrual mechanism are those where buyback spending exceeds token emissions by a meaningful multiple. Pendle's 4.4x buyback-to-emissions ratio is the current benchmark. Programs that cannot demonstrate a comparable ratio are, by the arithmetic, providing exit liquidity. The $2B spent to date has purchased that clarity, if nothing else.