Billions of dollars are up for grabs, but the ultimate return is only about four cents. That's the annual return generated by every dollar of U.S. Treasury bonds. For nearly a decade, decentralized finance (DeFi) protocols have relied on USDT and USDC as the backbone of their products, while allowing Tether and Circle to capture the returns on their reserves. These companies have earned billions of dollars in profits through the world's simplest way to generate returns. But now, DeFi protocols want to capture some of these returns themselves. Stablecoin leader Tether currently holds over $100 billion in reserves and generates over $4 billion in interest income. This is more than the total profit of $3.761 billion that Starbucks earned from selling coffee worldwide in the previous fiscal year. And the USDT issuer achieved this simply by investing its reserves in U.S. Treasuries. Circle adopted the same approach when it went public last year, emphasizing its float as its core revenue source. Currently, the total value of stablecoins in circulation exceeds $290 billion, generating approximately $12 billion in revenue annually. This is a significant amount of capital that cannot be ignored. This has sparked a new war in DeFi, where protocols are no longer content to let issuers capture these returns. They now want to own the products and their infrastructure. Earlier this month, Hyperliquid launched a bidding war for its native stablecoin, USDH, requiring the winner to return the returns. Native Markets, Paxos, Frax, Agora, and Ethena all participated in the bidding. The eventual winner, Native Markets, pledged to return 100% of USDH's treasury revenue to the blockchain: half for HYPE token buybacks and the other half for ecosystem funding. Currently, $6 billion in USDC is held on Hyperliquid's Layer-1, potentially generating $240 million in revenue. This revenue, previously held by Circle, may now be redirected to token burns and developer incentives. For reference, Hyperliquid generated $260 million in net revenue from transaction fees in June, July, and August. Ethena is growing faster and at a much larger scale. In just two months, the circulating supply of its synthetic stablecoin, USDe, jumped from $5 billion to nearly $14 billion, surpassing Maker's DAI to become the third-largest dollar-pegged stablecoin after USDT and USDC. In August, Ethena's revenue reached $54 million, a record high so far in 2025. Now, with the official launch of its long-awaited fee conversion mechanism, up to $500 million per year can be redirected to ENA buybacks, tightly tying the fate of the ENA token to the cash flow generated by the system. Ethena's model involves going long on spot cryptocurrencies, shorting perpetual swaps, and distributing returns between Treasury bonds and staking. As a result, sUSDe stakers enjoyed an annualized yield (APY) exceeding 5% in most months. Maker, a veteran of the industry, was one of the first companies to utilize US Treasury bonds as a stablecoin reserve. At one point, it held over $1 billion in short-term Treasury bills, enabling it to offer an 8% savings rate on DAI, briefly exceeding the average yield on US junk bonds. Excess funds were funneled into its surplus buffer and then used for buybacks, destroying tens of millions of MKR tokens. For token holders, this transforms MKR from a mere governance badge into a claim to actual income. Frax, on the other hand, is smaller but more focused. Its supply hovers below $500 million, a fraction of Tether's $110 billion, but it's still a money-making machine. Founder Sam Kazemian designed FRAX to reinvest every dollar of reserve revenue back into the system. Some of this revenue is burned, some is shared with stakers, and the rest is deposited into sFRAX, a vault that tracks the Federal Reserve's interest rate. Even at its current scale, the system generates tens of millions of dollars in revenue annually. Aave's GHO stablecoin was built with verticalization in mind. The stablecoin, launched in 2023, currently has $350 million in circulation. The principle is simple: each borrower pays interest directly to the DAO, rather than to an external lender. With a borrowing rate of 6-7%, this will generate approximately $20 million in revenue, half of which will be shared with AAVE stakers, with the remainder going into the treasury. The new sGHO module will offer depositors an annual interest rate of up to 10% (subsidized by the reserve), further enhancing the deal's appeal. In effect, the DAO is willing to use its own funds, making its stablecoin resemble a savings account. Other networks are using stablecoin revenue as primitive infrastructure. MegaETH's USDm is backed by tokenized treasuries, but rather than being paid to holders, its revenue is used to pay rollup sequencer fees. At scale, this could mean millions of dollars annually in gas fees, effectively transforming treasury coupons into a public good. The common thread in all these initiatives is verticalization. Each protocol is no longer content to rely on someone else's dollar rails. They are minting their own currencies, capturing the interest that would otherwise belong to the issuer and repurposing it for buybacks, treasury bonds, user incentives, and even subsidizing blockchain construction. While the yield on treasury bonds may seem mundane, in DeFi, it has become the spark for building self-sustaining ecosystems. When you compare these models, you'll find that each protocol is setting different valves to tap into this 4% yield stream: buybacks, DAOs, sequencers, and users. The yield is passive income. It makes everyone reckless. Each model has its own bottlenecks. Ethena's peg relies on perpetual funding to remain positive. Maker has experienced real-world loan defaults and had to cover losses. After Terra's collapse, Frax withdrew funding and scaled back its issuance to prove it wouldn't be next. All of these institutions rely on one thing: US Treasuries held by custodians like BlackRock. These are decentralized wrappers around highly centralized assets. And centralization brings the risk of collapse. Meanwhile, new regulations present challenges. The US's GENIUS Act outright bans interest-bearing stablecoins. Europe's MiCA legislation imposes restrictions and licensing requirements. DeFi has found workarounds by labeling returns as "buybacks" or "sequencer subsidies," but the economics remain the same. If regulators choose to take action, they are fully capable of doing so. Yet, this approach helps build sustainable business models—something the crypto space has long struggled with. The sheer number of models in play demonstrates the enormous potential that DeFi protocols currently possess. The battle for the world's most boring yield is fierce. However, the risks are high. Hyperliquid ties it to token burns, Ethena to savings accounts and buybacks, Maker to central bank-style buffers, and MegaETH to rollup operating costs. I wonder if this movement will erode the market share of the giants, draining liquidity from USDC and USDT. If not, it will certainly expand the market, creating a layer of yield-generating stablecoins to rival zero-yield stablecoins. No one knows yet. But the battle has begun, and the battlefield is vast: a stream of interest is flowing from US government debt, through protocols, to tokens, DAOs, and blockchains. The quarter-cent profit that once belonged to the issuer is now driving the latest developments in DeFi.
Catherine
Weatherly
Kikyo
Alex
Anais