渠道向发行方收税，Hyperliquid如何撬开Circle的钱袋子？

Original Author: Xiaobing, TechFlow

On May 14, Coinbase and Circle jointly announced they would re-enter Hyperliquid under the "AQAv2" framework, with Coinbase becoming the treasury deployer for USDC, returning the majority of the yield generated from USDC reserves back to the Hyperliquid protocol. Native Markets' USDH agreed to have its brand assets acquired by Coinbase and will be gradually phased out.

Sounds like an ordinary cooperation announcement? Not quite.

Here are the specific numbers: The USDC volume on Hyperliquid is approximately $5 billion. At current Treasury yields, this generates roughly $200 million in reserve income annually. According to leaked cooperation details, after deducting "costs," about 90% of the reserve yield will flow back into the Hyperliquid ecosystem, potentially boosting protocol revenue by 22%–26%.

This is the largest concession an issuer has ever made to a single channel in the history of the stablecoin industry. Previously, only Coinbase (as a co-issuer, receiving over half of Circle's distribution revenue), Binance, and a handful of undisclosed partners could secure revenue sharing from Circle.

And Hyperliquid is a decentralized protocol with no equity ties, no history of co-issuance, and an ambiguous legal entity.

What gives it the leverage?

The Standoff

To understand this deal, we need to go back to September 2025.

At that time, Hyperliquid was still using bridged USDC as its primary collateral asset. USDC volume had nearly touched $6 billion, accounting for 7.5% of USDC's total circulating supply. At prevailing interest rates, this $6 billion contributed approximately $220 million annually to Circle's reserve income, while Hyperliquid received nothing.

One KOL commented: "Hyperliquid holds $5.5 billion USDC, generating $220 million annually for Circle. After USDH launches, it can capture $110 million within the protocol. No new products, no new users needed—just redirecting reserve income from Circle shareholders to HYPE holders."

So the Hyperliquid team did something incredibly clever: instead of issuing their own stablecoin, they **put the "USDH" ticker up for public tender**. Paxos, Ethena, Frax, Sky, Agora, Native Markets—half the stablecoin ecosystem jumped in to bid. All bidding conditions revolved around "how much reserve yield can you return to the Hyperliquid ecosystem," with almost every bidder offering 95%–100% revenue share ratios.

Ultimately, the community awarded the ticker to Native Markets, a team tailor-made for Hyperliquid founded by Mary-Catherine Lader (former COO of Uniswap Labs), among others. The allocation was set at 50% for HYPE buybacks and 50% for ecosystem incentives.

The true power of this move wasn't that USDH could replace USDC—in fact, eight months after launch, USDH's scale remained far below USDC's. Instead, it held a knife to Circle and Coinbase:

Either you accept this "protocol sovereignty" rulebook and give up the yield, or we slowly replace you.

Coinbase's reaction was telling. Instead of forming a "united front" with Circle and fighting back, it directly acquired USDH's brand assets and "copied" the entire AQA framework into the USDC system. On the surface, Coinbase stepped in to preserve USDC's home turf. In essence, Coinbase admitted: **The rules of the game have changed—concessions are necessary.**

Native Markets co-founder Mary-Catherine Lader tweeted on the day of Coinbase's announcement: "When we won USDH 8 months ago, our thesis was simple: people care about stablecoins that pass value back to the network and users. Today, that thesis was validated."

She was being too polite. This was a meticulously designed, textbook example of power transfer across the industry chain.

What Has This Changed?

Layer 1: The "Channel Revenue-Sharing Era" for USDC Reserve Yields Has Officially Begun

For the past decade, the business model for stablecoin issuers has been brutally simple: users mint stablecoins → issuers buy U.S. Treasuries with the dollars → all yield goes to the issuer. In 2025, Circle generated $2.6 billion in reserve income this way, supporting a $30 billion IPO valuation.

This model was built on an assumption: issuers are scarce, channels are abundant. As the two deepest liquidity stablecoins, USDT and USDC were assets that CEXs and DEXs had to beg to list.

Hyperliquid proved that when a channel becomes large enough (accounting for 7.5% of USDC's circulation), and **it has the capability to issue its own stablecoin to replace you at any time**, the power dynamic flips. **The issuer becomes the supplier vying for scarce resources.**

What happens next? Just look at Circle's newly filed Q1 2026 financial report: reserve income of $2.637 billion, the absolute pillar of its revenue. If future giants like Binance, OKX, Bybit, Phantom on Solana, or even major Ethereum L2s come to the table with this "AQA script," Circle's profit margins will be systematically shaved down.

The market priced in this anxiety, with CRCL shares dropping 7.6% from intraday peak to close at $122.34 on May 14. The market cast a clear vote with real money: a short-term positive (USDC expansion on Hyperliquid's turf), but a long-term negative (institutionalization of the revenue-sharing model).

Layer 2: HYPE Has Gained a True "Cash Flow Anchor"

Many haven't realized that this deal represents a structural upgrade to HYPE's valuation logic.

Previously, HYPE's value story was: trading fees → aid fund → buyback and burn. This model depends on trading volume, which is cyclical and highly volatile.

Now there's an additional leg: **Treasury yield → protocol revenue → HYPE buybacks**. This leg doesn't depend on market sentiment or trading activity—it relies on just one thing: **how many dollars are locked on Hyperliquid.**

This is a very different type of cash flow. Its nature is closer to a bank's net interest margin than an exchange's fee income. The latter fluctuates wildly with bull and bear markets; the former provides a steady stream as long as interest rates don't hit zero and the deposit base doesn't vanish.

A simple calculation based on current scale: $5 billion × ~4% Treasury yield × 90% share ≈ $180 million in new annual protocol revenue. If all of this is used for HYPE buybacks and the aid fund, for a token with a circulating market cap of around $15 billion, it means an additional ~1% "passive deflation" annually—and this pool is doubling year-over-year.

On the news day, HYPE rose 14%. The market reaction was correct. But what's more noteworthy isn't the single-day gain, but the migration of HYPE's valuation model **from 'exchange token' to 'sovereign stablecoin's treasury yield distribution certificate.'**

The latter is a completely different asset class—one the market hasn't yet developed a pricing framework for.

Layer 3: The "Neutrality" of USDC Begins to Crumble

This is the most easily overlooked, yet potentially most profound layer.

Stablecoins can serve as the settlement layer for the crypto world because of neutrality. USDC theoretically treats all chains, all exchanges, and all applications equally. This is where it differs from banks: banks have customer tiering, while stablecoins historically haven't.

But the AQAv2 agreement offers Hyperliquid treatment that differs from USDC's treatment on Ethereum mainnet, Solana, or Arbitrum. Hyperliquid gets a 90% reserve yield share, and Circle and Coinbase must also stake HYPE as validators. This is a highly customized, deeply entrenched relationship.

So the question arises: **When USDC offers different economic terms to different networks, can it still be considered a 'neutral' settlement layer?**

Every channel with bargaining power will start demanding its own "special terms." Wouldn't Solana want them? Base? Arbitrum? USDC could ultimately become a highly fragmented "revenue-sharing network" pieced together by dozens of bilateral agreements.

This is the true legacy USDH leaves behind. It didn't lose to USDC; it forced USDC to become USDH.

The real meaning is hidden in the Native Markets co-founder's words: "USDH may be disappearing, but its core innovation already won because Coinbase is adopting the underlying economics."

TechFlow's Take

From a trader's perspective, the most interesting part isn't that HYPE rose 14% or CRCL fell 7%. It's that: **Every time in financial history a 'channel prices upstream in reverse,' the endings look very similar.**

Visa and Mastercard continue to capture the fattest profits in the card network ecosystem because they are the channels. Commercial banks eventually agreed to share revenue with Walmart and Costco for co-branded credit cards because no terminal means no transaction. Apple's 30% App Store commission was, at its core, a tax from the channel on developers.

But the other side of the story is: **When a channel grows to a certain critical mass, it begins to eat into upstream profits.** Costco's private label Kirkland has become a top-of-mind brand for consumers; Spotify forced record labels to accept subscription models; Steam compelled publishers to accept a 30% cut while also ceding refund rights.

Stablecoins in the crypto world previously remained in the "upstream calls the shots" stage. **What Hyperliquid did has abruptly pushed this industry into the next phase: 'the channel calls the shots.'**

In the short term, this is just one deal. In the medium term, it's the beginning of a structural breakdown in Circle's business model. In the long term, it's an inflection point where stablecoins move from "issuer sovereignty" to "network sovereignty." Stablecoins no longer belong solely to the company that issues them; they begin to belong to the network where they settle.

Those who think this was just Hyperliquid winning a single round haven't realized the entire table has been flipped.

Who will make the next move? My bet is on Solana, and it won't take long.