On-Chain Stablecoins vs. U.S. Treasuries: Higher Risks, Are They Matched by Sufficient Yield Premiums?

Stablecoin yields do not always outperform U.S. Treasuries, but they offer more stable returns compared to volatile assets.

Written by: Cooper Duschang, Tanay Ved, Coin Metrics

Compiled by: Chopper, Foresight News

The U.S. Treasury market sees daily trading volumes reaching up to $1.2 trillion. From U.S. Treasuries to junk bonds, various fixed-income products serve as effective tools for earning interest on idle funds. Today, the stablecoin market has surpassed $290 billion, allowing holders to deposit stablecoins into various on-chain platforms, with mainstream options including lending protocols like Aave and Morpho, thereby earning yields similar to traditional financial instruments.

The emergence of on-chain tokenized Treasuries, corporate bonds, and money market funds has further expanded opportunities for asset interest accrual and collateralized lending. These assets can provide competitive yield levels compared to stablecoins in lending protocols. As market products continue to diversify, investors need to reassess the risks and returns of different income-generating products.

In this article, we dissect the additional risks associated with on-chain yields, the yield differentiation brought by stablecoins and various lending protocols, and compare the lending yields of stablecoins with those of high-volatility assets.

The Boundaries Between Traditional and On-Chain Yields Are Blurring

For a long time, investors have allocated short-term U.S. Treasuries and money market funds to earn returns with minimal risk. The return levels of these investments are relatively low, but due to their short duration and ample liquidity, they are considered the safest assets.

The on-chain yield ecosystem now includes various forms: stablecoins with inherent yield properties, interest from stablecoin deposits within DeFi protocols, and tokenized Treasuries that benchmark against traditional fixed-income products. Both designs aim to provide users with low-risk, sustainable yields backed by authoritative institutions and reserve assets.

Stablecoin lending yields vs. U.S. money market benchmark yields, data source: Talos CM Protocol Metrics, FRED

On-chain investments feature real-time settlement, programmability, and composability—advantages that traditional markets find hard to replicate. However, investors must also bear additional risks. For instance, tokenized Treasuries like Franklin Templeton Money Market Fund (BENJI) and BlackRock U.S. Dollar Institutional Digital Liquidity Fund (BUIDL) face smart contract risks, redemption risks, and liquidity risks.

The yield trajectory of tokenized Treasuries anchors to the effective federal funds rate while bearing higher potential risks. This also presents new opportunities for stablecoins and their DeFi composite products to create an on-chain yield system that can compete with traditional financial products.

DeFi introduces more complex variables and potential risks; theoretically, stablecoins should offer higher yields as risk compensation. However, such high yields are not always achievable.

Significant Yield Differentiation for the Same Stablecoin Across Different Lending Protocols

The two major lending protocols, Morpho and Aave v3, have accumulated over $8.6 billion in stablecoin deposits. Following the KelpDAO security incident in April 2026, many analysts raised concerns: the lending pools harbor various potential risks, yet the yields offered to investors do not match the corresponding risk premiums. Stablecoin lenders face risks such as stablecoin de-pegging, oracle manipulation, and smart contract theft. This aligns with the logic of traditional fixed-income products: higher risks should yield higher returns.

The yield levels of the same stablecoin across different protocols are not uniform. From January 2026 to the present, the average yield of USDC deposited in Aave and Morpho differs by 1.59%. The underlying design differences of the protocols create differentiated risk-adjusted yields. Morpho establishes independent lending markets for each collateral-lending transaction, resulting in differentiated borrowing and deposit rates; Aave adopts a shared liquidity pool model, using a unified borrowing rate for the same asset within a single market.

Aave USDC annualized yield vs. U.S. Treasury yield, data source: Talos CM Protocol Metrics, FRED

The core market for Aave shows significant fluctuations in USDC lending yields, contrasting sharply with the stable yield characteristics of traditional fixed-income products. The average yield of USDC on Aave is 31 basis points lower than the 1-year U.S. Treasury yield. In 2026, during 78% of the time intervals, the annualized yield of USDC on the Aave platform was lower than that of the 1-year U.S. Treasury.

Morpho USDC annualized yield vs. U.S. Treasury yield, data source: Talos CM Protocol Metrics, FRED

The median average yield of USDC in the Morpho v2 vault exceeds the 1-year U.S. Treasury yield by 65 basis points, but the annual yield volatility is approximately 3.3 times that of U.S. Treasuries. Lending yields are driven by the supply-demand relationship of funds and can change at any time; U.S. Treasury yields are determined by Federal Reserve policy, resulting in smoother fluctuations.

Different Stablecoins Yield Differently Within the Same Lending Protocol

The generation logic of lending rates relies on the deposit volume of the liquidity pool and the borrowing demand. Therefore, even if the reserve mechanisms are similar, yield levels for different stablecoins within the same lending protocol still exhibit disparities.

Aave different market deposit annualized yield comparison, data source: Talos CM Protocol Metrics

On the Aave platform, the two leading stablecoins by deposit volume, USDC and USDT, have an average yield difference of 90 basis points, rooted in differences in fund utilization rates. Over the past 90 days, the average fund utilization rate for USDC has been 10% higher. When a large number of lenders withdraw liquidity, the deposit rate for PYUSD briefly spiked; the increased fund utilization rate raised the rates for all borrowers and lenders.

Morpho v2 various asset head vault deposit annualized yields, data source: Talos CM Protocol Metrics

The Morpho ecosystem also exhibits similar yield differentiation. Funds continue to shift from the v1 vault to the v2 vault, causing fluctuations in market deposits and remaining liquidity, resulting in stronger yield volatility for the Morpho v1 vault. In the Morpho v2 vault, the average yield difference between USDC and USDT reaches 126 basis points. All four stablecoins mentioned in the report provide reserve proofs, with assets allocated to cash and short-term U.S. Treasuries, and support multiple chains. However, the degree of integration within the DeFi ecosystem, regulatory environment, market demand, and subtle differences in protocol mechanisms ultimately lead to vastly different yield expectations.

Vault Managers and Yield-Generating Stablecoins Broaden Yield Strategy Boundaries

On-chain yields are not solely determined by protocol architecture or the assets themselves. Governance mechanisms of protocols and vault managers also influence the final yields. The Aave governance community can adjust lending rate curves based on asset risk assessments; the yield performance of Morpho is determined by the allocation decisions of vault managers. Vault managers act as investment managers, allocating deposits to different lending markets to earn yields.

Comparison of Morpho platform USDC vault yields, data source: Talos CM Protocol Metrics

Managers configure different markets based on their own risk preferences; even with identical underlying lending assets, diverse trading strategies can evolve. Over the past 90 days, the median yield for all USDC vaults was 4.79%; some high-yield targets raised the overall average, pushing the vault average yield up to 5.31%. This means that without modifying the lending protocol rules, vault managers can enhance user yields.

Comparison of GHO and USDS circulating supply, data source: Talos CM Network Data Pro

Stablecoin issuers can also directly provide incentives to attract users to hold assets, no longer relying on intermediary lending markets. Users can directly stake stablecoins like USDS issued by Sky and GHO issued by Aave to receive rewards set by governance rules. Over 66% of circulating USDS is staked and converted to sUSDS, currently yielding approximately 3.52%; staking GHO can be exchanged for sGHO, earning a fixed 4.25% yield.

Comparing Lending Yields of Stablecoins with Volatile Crypto Assets

At the same yield levels, stablecoins bear higher risks, and their yield performance can sometimes be difficult to benchmark against similar assets; however, reviewing historical data, stablecoins can provide more stable yields compared to high-volatility crypto assets in lending.

Comparison of net asset growth of $10,000 initial investment in Aave v3, data source: Talos CM Protocol Metrics

If investors start depositing assets in Aave from 2024, the yields from USDC will surpass those from ETH. ETH itself achieves a price increase of 8.9%, but the interest earned in Aave is about $940 less than that from USDC. Of the yield from WBTC, only 0.1% comes from lending interest, while the remaining 86.9% of returns are entirely from price appreciation, with WBTC being used more as collateral.

If one starts allocating Morpho vault products from 2025, the final yield of the Steakhouse USDT vault will be 4.5%, while the Steakhouse ETH vault records a loss of 1.9%. Excluding the gains and losses from price fluctuations, the yield of Steakhouse USDT remains 2% higher than that of Steakhouse ETH.

Stablecoins can continuously generate stable yields. Compared to traditional financial products, the yields of stablecoins do not always hold an advantage; however, when benchmarked against other crypto assets, the yield advantage of stablecoins sufficiently explains why over $8.6 billion in funds choose to capture on-chain yields through stablecoins.