Coinbase serves as the official treasury deployer. It manages the large reserve pool where the majority of USDC sits, deploying those funds into yield-bearing instruments—primarily short-term US Treasury bills .
Hyperliquid receives approximately 90% of the cost-adjusted reserve yield generated from the USDC sitting on its Layer 1 blockchain—income that previously flowed entirely to Circle and Coinbase .
The USDC on Hyperliquid is algorithmically split at a fixed 1:9 ratio between a contract execution layer (the active trading float) and a treasury reserve layer (the idle reserves that earn yield) . The yield from the reserve layer is paid into Hyperliquid's Assistance Fund, the protocol-controlled pool that executes automatic HYPE buybacks from the open market
.
Projections vary based on the total USDC balance on Hyperliquid and prevailing T-bill rates, which sat around 3.8–4.3% in mid-2026 , but the magnitudes across sources converge on a clear range:
For perspective, the previous native stablecoin USDH had a supply of roughly $100 million, generating only about $1.9 million in annual yield for buybacks . AQAv2 represents a roughly 70 to 100x expansion in the buyback revenue base
.
This new revenue stream stacks on top of Hyperliquid's existing buyback flows. The protocol already directs 97% of trading fee revenue into the Assistance Fund for automated HYPE purchases . With AQAv2, the total annual buyback mechanism pushes toward $900 million or more, combining trading fees with USDC reserve yield
.
The rollout followed a clear schedule:
AQAv2 is novel not just for its scale but for its structure, which inverts a long-standing power dynamic in crypto.
1. The first large-scale on-chain yield recapture. Stablecoin reserve yield—the interest earned on the fiat backing USDC or USDT—has historically been a centralized revenue stream captured entirely by issuers like Circle and their distribution partners like Coinbase. With AQAv2, a DeFi protocol has successfully negotiated to claim the majority share (~90%) of that yield and route it on-chain to its own token holders . It is a structural shift from yield flowing outward to Wall Street toward yield flowing inward to the protocol.
2. Skin in the game through staking. Both Circle and Coinbase were required to stake HYPE tokens as a condition of participation: Circle staked 500,000 HYPE, and Coinbase increased its existing staked position . This directly aligns the financial interests of the centralized issuers with the protocol's token price. If HYPE underperforms, these partners lose alongside holders. This "skin in the game" requirement is a new pattern for how DeFi protocols can condition access to their ecosystems.
3. Consolidation of quote assets. Before AQAv2, Hyperliquid's stablecoin landscape was fragmented between USDC and the protocol's native USDH—and potentially future entrants. AQAv2 eliminates that fragmentation by establishing a single "aligned" quote asset, deepening liquidity and simplifying market structure .
4. A potential precedent for the industry. If the model proves durable, other major DeFi derivatives protocols—dYdX, GMX, Synthetix—have a clear template for negotiating similar yield-sharing agreements with Circle and Coinbase . Each would likely involve staking commitments that tie issuers into the protocol's success, creating positive-sum feedback loops. The implication is that the massive T-bill interest generated by stablecoin reserves may no longer be a private revenue stream for a few centralized firms—it could become a distributed layer of value accrual across DeFi.
Sources note that the arrangement is not without friction. USDH must be gradually phased out, creating migration complexity for users . The framework concentrates Hyperliquid's quote asset risk on a single centralized issuer (Circle and its bank partners), which could pose regulatory or operational concentration risk. And while projections are large, they depend on sustained USDC balances and relatively stable T-bill yields—variables that can shift with Fed policy or market sentiment.
Comments
0 comments