The sources of repurchase funds vary across different DeFi protocols, with the destruction affected by trading popularity; maintaining a high annual destruction rate is difficult, and the quality of the repurchase depends on the stability of underlying cash flow.

Author: Conflux

On September 1, the daily trading volume of Robinhood Chain, the public chain launched in July this year, surged to $1.43 billion, setting a historical high. This surge in trading volume has started to directly reflect on Uniswap's balance sheet.

The more transactions on Robinhood Chain, the more protocol fees Uniswap receives; once the protocol fees enter Uniswap's repurchase and destruction mechanism, they will be converted into the purchase and destruction of UNI. Trading volume, protocol revenue, and Token supply are interconnected by the same mechanism. This is a change worth paying more attention to in this round of UNI's market trends.

Where does the money for UNI's rise come from

The UNIfication mechanism launched by Uniswap in December 2025 includes two contracts: TokenJar and Firepit. The former is responsible for collecting protocol fees, while the latter is responsible for destroying UNI. Once the protocol fees enter TokenJar, unlocking requires destroying an equivalent amount of UNI, which effectively binds protocol revenue to the demand for UNI.

The previous repurchase scale of this mechanism was not large, until the trading volume on Robinhood Chain began to rise.

It is reported that from July 27 to August 12, the average daily protocol revenue of Uniswap soared from $99,800 over the previous 17 days to $244,000; as of the first seven days before August 12, Uniswap's total protocol revenue was about $1.55 million, of which $925,000 came from Robinhood Chain, accounting for about 60%.

Subsequently, the trading scale of Robinhood Chain continued to expand, and the destruction speed of Uniswap also increased. On August 21, Uniswap destroyed approximately 150,000 UNI in a single day, setting a record at the time. According to multiple statistics, the cumulative destruction since the launch of UNIfication has exceeded 100 million UNI.

However, there is an area in this set of figures that requires vigilance.

According to Geoff Kendrick, Global Head of Digital Assets Research at Standard Chartered Bank, if calculated based on the destruction rate in mid-August, the annualized destruction amount of UNI corresponds to about 4% of the circulating supply. He believes this speed is "clearly unsustainable"; even based on his previously given target price of $6.50 by the end of 2026, the corresponding annualized destruction rate would still be about 2.2%.

The reason is not complicated: the destruction speed ultimately depends on protocol revenue, which in turn depends on trading activity.

If Robinhood Chain continues to scale up, there will be room for UNI's repurchase and destruction to expand; if trading popularity declines, this new buying pressure will also shrink simultaneously.

This also shifts the question from UNI to the entire DeFi industry: although repurchases are similar, the underlying revenue sources, fund scales, and sustainability can vary greatly.

Where are the differences in repurchase layouts

According to statistics from the Token Rights section of the data platform DefiLlama, as of August 27, 2026, among the 106 protocols it recorded, 55 were marked as Active Buybacks.

However, there is a significant gap between "having a repurchase mechanism" and "the scale of repurchase being sufficient to impact the Token."

Data from blockchain data agency Allium Labs illustrates this point: since 2026, the total repurchase expenditure across the industry has been about $638 million, with Hyperliquid and Pump.fun contributing nearly 90%. This means that although dozens of protocols have opened repurchase mechanisms, only a few have actually formed large-scale actions.

Hyperliquid is the most extreme case. Approximately 99% of the platform's perpetual and spot trading fees enter the Assistance Fund, used to buy and destroy HYPE, with an estimated annualized repurchase scale of about $714 million based on current income levels. The AQAv2 launched on August 26 further includes platform USDC reserve earnings as a source of repurchase funds, expected to add about $135 million to $200 million annually. For HYPE, the more profitable the protocol, the more sufficient the repurchase funds, making the relationship between the two very direct.

The funding sources for Sky are different. SKY's repurchase relies on protocol surpluses, not trading fees. These surpluses mainly come from stablecoins and RWA asset allocations. According to Allium data, since 2026, Sky has invested about $26 million in repurchasing SKY; the official dashboard shows that since the mechanism was initiated in February 2025, the cumulative repurchase amount has exceeded $100 million.

Spark follows a similar logic but on a significantly smaller scale. According to governance proposal SAEP-09, Spark allocates a portion of the protocol surplus for repurchase each month, with approximately $2 million repurchased in the first half of 2026. The mechanism has been established, and the execution rhythm is relatively transparent, but the current amount is still insufficient to create a strong market impact on SPK.

Aave has experienced another situation.

After the rsETH incident in April 2026, Aave DAO temporarily paused repurchases to prioritize funds for risk management; Aavenomics 3.0, launched on June 27, re-integrated repurchase into an automated, immutable execution mechanism. According to the adjusted repurchase schedule, the DAO expects to buy about 292 AAVE per day. This means Aave's repurchase is no longer just a governance proposal but is written into the protocol's automated execution framework.

When looking at these several cases together, the differences become quite intuitive: Hyperliquid relies on trading fees, Sky and Spark depend on protocol surpluses, while Aave re-established automated repurchases after addressing risk management.

Can repurchase withstand cycles

Looking at these projects together reveals a clear dividing line: Hyperliquid, Uniswap, Sky, and Aave have linked some economic value generated by the protocol to the Token, but the income sources they rely on and their repurchase mechanisms differ; Spark, on the other hand, is still at a stage where the mechanism has been established, but the scale remains small.

For UNI, this incremental increase from Robinhood Chain is indeed real and can be sustainably tracked, but it is highly dependent on the trading popularity of a new chain that has only been live for two months—while the trading activity on this chain largely comes from the explosion of new assets and new applications. Standard Chartered's warning of "4% annualized destruction is unsustainable" precisely speaks to this issue: repurchase is merely a result; what truly determines its value is whether the underlying revenue can be sustained.

The real question to ask is: when "repurchase" becomes a universal mechanism used by various projects, whose repurchase is backed by stable business cash flow, and whose repurchase merely expands passively with a wave of trading enthusiasm?