Why does $1 earn more in banks? Analyzing the structural dilemma of DeFi lending

Author: Silvio

Translation: Saoirse, Foresight News

The $1 in bank deposits generates ten times the revenue for the bank compared to an equivalent amount of USDC on Aave. While this phenomenon may seem unfavorable for the DeFi lending sector, it actually reflects more the current structural characteristics of the cryptocurrency market rather than the long-term potential of on-chain credit.

Net interest margin is an indicator of deposit profitability. FIDC’s affiliated banks, Aave under Blockworks.

This article will explore the following questions: the actual application methods of current lending protocols, why their profit margins are structurally lower than banks, and how this situation might change as lending activities gradually detach from the native leverage cycles of crypto.

The Role of On-Chain Credit

My first job involved analyzing bank books and assessing borrower qualifications. Banks lend credit funds to real-world enterprises, with profit margins directly related to macroeconomic conditions. Similarly, analyzing borrower situations in decentralized finance protocols also helps understand the role of credit in the on-chain economy.

Aave Unpaid Loan Data Chart

Aave’s outstanding loan amount has surpassed $20 billion, a remarkable achievement — but why do people borrow on-chain?

The Actual Use Cases of Aave Borrowers

Borrowers’ strategies can be divided into four categories:

1. Using interest-bearing ETH as collateral to borrow WETH: The yield on staked ETH is usually higher than WETH, creating a structural basis trade (“borrowing WETH while earning yield”). Currently, these trades account for 45% of total outstanding loans, mostly from a few “whale” wallets. These wallets are often associated with ETH staking issuers (like EtherFi platform) and other “circular stakers.” The risk of this strategy is that WETH lending costs may suddenly spike, quickly causing collateral health to fall below liquidation thresholds.

Estimated WETH Lending Rate Chart: If the rate stays below 2.5%, the basis trade remains profitable

2. Stablecoin and PT circular stakers: Yield-bearing assets (like USDe) can also form similar basis trades, with yields potentially higher than USDC lending costs. Before October 11, this type of position was very popular. Although structurally attractive, these strategies are highly sensitive to changes in funding rates and protocol incentives — which explains why their scale shrinks rapidly when market conditions change.

3. Volatility collateral + stablecoin debt: This is the most popular strategy among users, serving two main needs: one, leveraging to increase crypto holdings; two, reinvesting borrowed stablecoins into high-yield “liquidity mining” for basis trading. This strategy is directly related to mining yield opportunities and is a primary driver of stablecoin lending demand.

4. Other residual types: including “stable collateral + volatility debt” (used for shorting assets) and “volatility collateral + volatility debt” (used for currency pair trading).

1. Distribution of Aave wallet borrowing strategies; 2) Distribution of wallets by strategy

Collateral health chart weighted by loan amount

For each of these strategies, there exists a value chain composed of multiple protocols: these protocols leverage Aave to streamline trading processes and distribute yields to retail users. Today, this integration capability is the core competitive barrier in the crypto lending market.

Among these, the “volatility collateral + stablecoin debt” strategy contributes the most marginal interest income (over 50% of lending revenue from USDC and USDT).

Interest income share chart by asset type

While some enterprises or individuals do use crypto loans to finance operations or daily expenses, compared to “using on-chain leverage/arbitrage of yield differentials,” the scale of such real-world uses is very limited.

Three core factors driving the growth of lending protocols:

1. On-chain yield opportunities: such as new project launches, liquidity mining (e.g., Plasma platform’s mining activities);
2. Deep liquidity structural basis trades: such as ETH/wstETH trading pairs and stablecoin-related trades;
3. Collaborations with major issuers: these partnerships help develop new markets (e.g., the combination of pyUSD stablecoin and RWA).

The lending market mechanism is directly linked to “crypto GDP” (correlated with Beta), much like how banks are essentially a barometer of “real-world GDP.” When crypto prices rise, yield opportunities increase, the scale of interest-bearing stablecoins expands, and issuers adopt more aggressive strategies — ultimately driving revenue growth for lending protocols, token buybacks, and an increase in Aave’s token price.

Correlation chart of lending market valuation and revenue: lending market valuation is directly related to revenue

Comparison Between Banks and On-Chain Lending Markets

As mentioned earlier, the $1 profit efficiency in banks is ten times that of $1 USDC on Aave. Some see this as a bearish signal for on-chain lending, but I believe this is fundamentally a market structure inevitability, for three reasons:

1. Higher financing costs in crypto: banks’ financing costs are based on the Federal Reserve’s benchmark rate (lower than treasury yields), while USDC deposits on Aave usually have yields slightly above treasury yields;
2. Traditional commercial banks’ risk transformation activities are more complex and should command higher premiums: large banks manage hundreds of millions of dollars in unsecured loans to enterprises (e.g., financing data center construction), with risk management far more complex than “ETH circular staking collateral management,” thus justifying higher returns;
3. Regulatory environment and market dominance: banking is an oligopoly with high customer switching costs and significant industry entry barriers.

Moving Lending Away from the “Cycle Binding” of Crypto

Successful crypto sectors are gradually detaching from the crypto market’s own boom-bust cycles. For example, the open interest in prediction markets continues to grow despite price volatility; stablecoin supply is similarly less volatile than other crypto assets.

To better align with the broader credit market operation model, lending protocols are gradually incorporating new risk types and collateral, such as:

• Tokenized RWA and stocks;
• On-chain credit originating from off-chain institutions;
• Using stocks or real-world assets as collateral;
• Structured underwriting via crypto-native credit scoring.

Asset tokenization creates conditions for lending to become the “natural endpoint” of crypto. When credit activities decouple from price cycles, their profit margins and valuations will also shed their cyclical constraints. I expect this transition to begin manifesting around 2026.