Author: StarPlatinum
Translation: Shan Oppa, Golden Finance
The crypto industry has spent many years creating a number of tokens that have nothing to do with the business behind them. The protocol made $100 million, and token holders got governance rights - "Great tech geek, I have voting rights" (sorry, this sounds boring).
Fortunately, this model is changing.
More and more protocols are beginning to use real income to buy back their own tokens on the market. A protocol spending $10 million to buy its own tokens and destroy them is completely different from "announcement of a buyback proposal that may be implemented in the future, provided that the governance vote passes."
As "income" has once again become one of the biggest narratives in the crypto industry, I have been sorting out those projects where the chain between "product usage → income → token demand" has truly begun to take shape. Here are 12 that I think are worth figuring out.
1. $HYPE Hyperliquid
Probably the most obvious one.
Hyperliquid has built one of the largest on-chain perpetual contract exchanges and established a very simple relationship between "someone is trading here" and HYPE.
Transactions incur handling fees.
About 99% of the handling fees are used to buy HYPE through the Assistance Fund, and the purchased tokens are withdrawn from circulation according to the current mechanism.
This is basically the ideal form of the token model.
Someone trades BTC, ETH, or some unknown perpetual contract at three in the morning, Hyperliquid makes money from it, and a good chunk of that money ends up being demand for HYPE.
The key metrics for me are Hyperliquid's trading volume and revenue.
If this exchange continues to gain significant market share, the buyer behind HYPE will continue to have ammunition. If trading activity shrinks significantly, so does the ammunition.
It's that simple.
2. $PUMP PumpFun
PumpFun is a "dirtier" version of the same logic.
Its business relies on people continuing to issue and trade meme coins, so revenue is extremely sensitive to attention cycles.
But when the heat is on, Pump can make ridiculous amounts of money.
A portion of the proceeds was used to buy PUMP from the open market, with data at some points showing cumulative purchases reaching nine figures.
Every spam coin issued by a teenager posting an animal picture may indirectly turn into a bid for the platform token.
What I will focus on here is: how much of PUMP has actually withdrawn from the circulation supply.
Buying a token into a treasury and permanently destroying it are two completely different economic events.
And Pump also has to fight a far bigger enemy than its competitors: people will one day get tired of issuing garbage coins on it.
3. $RAY Raydium
RAY is probably the least exciting name on this list.
And that's exactly why I find it interesting.
Raydium has been around for years, survived multiple Solana cycles, and underpins a lot of trading infrastructure that people use but never think about.
Historically, a portion of the protocol transaction fees were used to buy RAY and then destroy it.
So when the Solana activity breaks out, Raydium benefits not only as a DEX: the activities of currency issuance, exchange and LaunchLab can also feed the economic system surrounding RAY.
We recently saw another example: StonkFun plugged into Raydium LaunchLab.
STONK takes away attention, new coins take away speculation.
4. $AAVE Aave
Aave makes money through the lending market.
The DAO then uses a portion of the protocol revenue to buy AAVE from the market.
The original plan was designed to be $50 million per year; however, as revenue conditions change, discussions have considered lowering that to about $30 million in 2026.
Compared to HYPE or classic “buy and burn”, there is a big difference here:
The AAVE purchased by DAO enters the Ecosystem Reserve.
In other words, the DAO is accumulating its own assets, and these tokens can eventually be used for incentives, staking, grants, or other purposes.
I actually like this distinction because it forces you to look at the entire balance sheet instead of going long as soon as you see the word "repurchase."
5. $SKY Sky
It’s the protocol originally called MakerDAO.
Surplus will be generated by USDS, mortgage lending, treasury assets and the wider Sky ecosystem.
A portion of the surplus can then be allocated to buy SKY, with its capital allocation system allowing tokens to be moved out of circulation.
This makes SKY's revenue stream completely different from the likes of PUMP.
Pump requires meme coin activity.
What Sky needs is for its balance sheet and stablecoin system to continue to generate surpluses.
6. $RLB Rollbit
RLB is probably the weirdest business on this list.
Rollbit combines crypto trading, gambling, and other speculative products and has historically used a portion of the money earned from these activities to buy and burn RLB.
It's acting almost like a private internet company that's buying back its own stock.
When users lose money in trading or gambling, Rollbit earns revenue, part of the revenue is bought back into RLB, and RLB is destroyed.
There are a few obvious reasons that lead me to give a larger risk discount here than Aave or Raydium: high business concentration, large regulatory exposure, and less transparency into some of the underlying funding flows.
7. $SYRUP Maple Finance
What’s interesting about Maple is that it’s been breaking away from crypto industry tropes.
The old formula is roughly: print tokens → send tokens to stakers → call this “revenue”.
Maple approved a change: use the protocol revenue to buy SYRUP.
8. $ETHFI EtherFi
EtherFi’s buyback system does not focus on permanently destroying ETHFI, but rather directs value to those who truly participate in the token economy.
Part of the income from eETH redemption fees is used to buy ETHFI weekly, and part of the income from various products in the Ether.fi ecosystem can also support additional monthly purchases.
The purchased ETHFI is mainly distributed to sETHFI holders.
So this looks closer to a token-denominated dividend.
EtherFi makes money, buy ETHFI, and those who lock ETHFI share a portion of the value.
Of course, the tokens still exist and the recipient can eventually sell them.
But I think this raises an interesting question for every buyback model: Would you rather destroy the asset permanently, or use the income to make "holding it and locking it" economically useful?
9. $ENA Ethena
ENA probably has the most potential buyback facility on this list.
The word "potential" carries a lot of weight here.
Ethena has discussed a framework in which a very high percentage of the protocol’s net proceeds—up to 95% under the proposed structure—could eventually be used to buy into ENA once USDe reaches required milestones.
If this mechanism were fully launched at scale, the numbers could be huge.
USDe is already a real financial product, with revenue derived from its collateral, hedging structure and broader ecosystem.
But Ethena’s economic model changes drastically with market conditions. The crypto industry has ruined too many beautiful spreadsheets.
10. $LDO Lido
What’s fascinating about Lido is that it controls a huge economic engine, the value of which LDOs have historically struggled to directly capture.
People stake ETH, Lido generates protocol revenue, stETH becomes one of the most important assets in DeFi, and LDO holders basically just govern things.
This gap has been one of the biggest points of criticism for the coin over the years.
The proposed buyback framework seeks to change this by allowing protocol revenue to fund the purchase of LDOs upon meeting certain economic conditions, including sufficient annualized revenue.
And this is where I distinguish between "implemented repurchases" and "possibility of repurchases." The question is how much of that business will ultimately be channeled to LDO.
11. $STONK StoneFun
I've been talking about this project a lot lately.
StonkFun has quickly become one of the weirdest launch pads on Solana because it gives people much more room to build markets outside of SOL. The basic gameplay is this: Find an asset, build a market around it, and then somehow crypto Twitter will trade it.
The cumulative income was US$22.9 million, of which US$8.3 million came from within seven days; and the project’s widely cited model is to use 60% of the income to buy STONK and destroy it.
This project needs more scrutiny, because many of the above figures come from project bulletin boards and ecological party calibers, and there are currently a lot of doubts surrounding reward tokens and selling pressure.
But it also makes it clear “what to keep an eye on”: revenue, the actual capital used to buy STONK, and the actual tokens sent to the burning address. If these three continue to move in the same direction, then there is a real economic mechanism underlying the speculation.
12. $NET NetNet Capital
NET is different from almost every other project here in that its buybacks have a reference price.
NetNet is building an on-chain treasury on Robinhood Chain. The treasury holds USDG and other assets, deploys capital into products like Morpho, and increasingly accumulates tokenized shares.
Each NET is calculated by the protocol’s Net Asset Value (NAV) and represents an equity in that balance sheet.
Under NAV, the protocol can buy NET at a price of approximately "NAV minus 1.5%" and destroy the purchased part.
This means that treasuries are not blindly buying NET every day regardless of price. Buybacks only make sense if the market prices NETs at a discount to their counterpart assets.
Above certain NAV multiples, this mechanism actually works in reverse: NetNet can issue new NET via bonds and add the proceeds to the treasury.
Buying below NAV and issuing at a high premium to NAV allows assets sandwiched between these two sets of mechanisms to grow.
NET has traded at a sizeable premium to NAV recently, so buybacks are not a reason for me to look at it today. The real decisive question is whether the treasury can compound interest faster than the holders are diluted.
in conclusion
I think “buyback” is about to become one of the most overused words in the crypto industry.
And what I want to know is: where did the money go.
I would look at "annual repurchase amount ÷ market capitalization". If a protocol with a market cap of $500 million can continue to spend $50 million a year on its own tokens, that deserves my attention. If a $10 billion market cap token announces a $5 million buyback while unlocking $500 million in supply, then I really don’t care.
The crypto industry finally figured out how to make protocols generate revenue. Now what I want to see is which protocols actually make these revenues meaningful for the token.