Crypto projects have poured roughly 640 million USD into buying back their own tokens so far in 2026, an increase of about 17 percent from the same period last year and a staggering jump from the 366,000 USD spent across all of 2024. Hyperliquid and Pump.fun alone account for nearly 90 percent of that spending, and the trend is forcing a difficult conversation across the industry about whether buybacks build lasting value or simply make tokens look more valuable than the businesses underneath them.
The mechanics are straightforward. A protocol takes the revenue it generates, uses it to buy its own token on the open market, and then either burns those tokens permanently or holds them in the treasury. Buybacks create ongoing demand for the token while burns reduce circulating supply, and the combination can put upward pressure on price. More importantly for an industry that has spent a decade struggling to connect token prices with protocol performance, it gives holders a tangible link to real economic activity.
Orest Gavryliak, chief legal officer at decentralized exchange aggregator 1inch, says projects implementing revenue-funded buybacks and burns typically have one of two objectives in mind: either decreasing the token supply in circulation or demonstrating the rationale for investing in protocol revenues. Telling users a project has bought and burned tokens, he argues, is far more straightforward than explaining how governance rights work, how fees are set, or how the protocol is actually used.
A sea change for token value accrual
Max Shannon, senior research associate at Bitwise Europe, believes buybacks and burns remain an effective way to accrue value to tokenholders because they create a continuous bid in the open market for the token, directly tethering token success to platform adoption. That marks a genuine shift for an industry that spent the last cycle chasing narratives and memecoin mania. Users who bought Fartcoin or Peanut the Squirrel were not evaluating sound economic models, and projects that survived the hangover are now eager to prove their tokens have a fundamental floor.
The scale of commitment varies dramatically. Hyperliquid has directed 99 percent of its revenue into buying back and burning HYPE, making it the flagship case study for the model. Pump.fun commits 50 percent of revenue to the same purpose, and more than 446 million USD worth of PUMP has already been removed from circulation. DeFi infrastructure protocol Spark takes a different route, acquiring over 143 million SPK through open-market buybacks funded by protocol surplus, but rather than burning them, the tokens remain in the treasury to reward long-term ecosystem participants.
Spark co-founder and chief executive Sam MacPherson says the point is not simply to reduce supply. Tokenholders should participate in the long-term economic success of the protocol rather than receive a distribution every time it generates revenue, he argues, and buybacks allow Spark to create that alignment while retaining flexibility over how and when the acquired SPK is deployed. There is also a practical tax angle: buybacks are a highly tax-efficient way to return revenue to holders, since users avoid the hefty tax bill that dividends or rewards can trigger.
Is buying your own token the best use of the money?
The bigger question is whether repurchasing tokens is genuinely the best allocation of a project’s capital. MacPherson frames it bluntly: the real question is what the highest-value use of the next dollar of surplus is. If a protocol can reinvest capital at attractive returns, that can be far more valuable than distributing revenue as it arrives. Every dollar spent buying the token is a dollar not spent hiring developers, expanding the business, strengthening the balance sheet, or shipping product.
The price data offers a sobering reality check. Pump.fun has been aggressively buying and burning PUMP since July 2025, yet the token still hovers roughly 50 percent below its September 2025 all-time high. UNI has given back around half of the gains it made after Uniswap unveiled its UNIfication proposal in November 2025. Shannon cautions that many factors contributed to those moves, so they do not prove buybacks failed, but they have prompted investors to debate whether startup-like projects would be better served by reducing the share of revenue committed to buybacks and reinvesting more in the team and product.
MacPherson draws the essential distinction: a buyback does not make an unsustainable protocol sustainable. Investors should separate a buyback scheme that pumps prices from a successful business model. A healthy protocol generating genuine surplus may decide repurchases are the best use of some capital, but a project limping along might simply be buying its own token to move the price.
Regulatory storm clouds ahead
As crypto imitates TradFi, regulators are watching. The draft Digital Asset Market Clarity Act of 2025 is not settled law, but Gavryliak notes its proposed framework highlights the key question of where a token’s value comes from. If value stems from the functionality of the network itself, the asset looks like a commodity. If it is based on the efforts of the team in shipping, marketing, or providing returns to holders, it is already a security. His warning is blunt: do not put the clothes of a stock on a token and expect it to be treated as a commodity.
Tokens also remain legally distinct from shares. A shareholder owns part of a company with voting rights, dividends, or a claim on residual assets. Tokenholders generally have none of those enforceable rights, and buyback participation is a market mechanism, not a legal entitlement. MacPherson describes SPK as a form of pseudo-equity that recreates the economic characteristics of ownership, governance participation and long-term alignment, without the corporate structure.
The ultimate test, according to Gavryliak, is simple: if the buybacks stopped, would there still be a reason to hold the token? If the answer is no, the problem runs deeper than tokenomics.
hyperliquid and pump.fun being 90 percent of all buyback spend tells you everything. two protocols with real revenue, everyone else doing theater
agree with the split but hyperliquid buying back with fee revenue is legit accrual. pump.fun is just recycling casino volume into its own token
640 million from basically zero in 2024. the gavryliak quote nails it, telling holders you bought and burned tokens is way easier than explaining governance
from 366k across all of 2024 to 640 million in eight months. hyperliquid and pump.fun are 90% of it, so really two treasuries are doing all the talking here
you say that like pump.fun buying its own token is a flex. their holders are the exit liquidity for that treasury, been obvious since the airdrop allocations dropped
treasury holding instead of burning is the tell. greta holm said it above, its market making with extra steps and nobody marks that bag to market
mark to market is the part nobody wants. audit those treasury bags at spot instead of cost and half the 640 million starts looking like papered over sell pressure
even cost basis disclosure would be a start. most buyback tweets dont say if they bought at 2 dollars or at the top
and every ‘held’ buyback quietly becomes sell-side liquidity the moment price dumps. no lockup, no schedule, no disclosure of when they can dump it back
mark to market rules on treasury bags would end the buyback bragging overnight. expect a fierce lobby against that disclosure
Buybacks only mean something if the revenue is real. Holding bought tokens in the treasury instead of burning them is just market making with extra steps, and nobody audits any of it
burn vs hold matters less than disclosure. listed firms report buyback size and price every quarter, most protocols just tweet a burn address and call it transparency
disclosure point is underrated. hyperliquid at least publishes the buyback wallet, half the protocols doing burns cant show you a source of funds statement
hyperliquid publishing the buyback wallet set the bar. no public address means a treasury buying its own bags in the dark
640 million with two names doing 90 percent of it isnt a trend its a duopoly discovering capital allocation. call me when the long tail starts
financial engineering is when the buyback replaces a dividend the token never had. look at what the business earns net of incentives and it gets ugly fast
Pump.fun buying back its token with meme launch fees is just routing retail losses through a burn address. at least hyperliquid earns real trading revenue first
Recycling launch fees into your own token is a treasury meme. At least the revenue side of that comparison exists without new victims walking in.
17 percent up on last year means nothing once you strip hyperliquid out. the other few hundred projects spent couch cushion money and still issued a press release about it
exactly. 366k across all of 2024 vs 640M now is mostly hyperliquid discovering the treasury button. one protocol’s program dressed up as an industry trend