Comparing DeFi Workflows: Process Design Across Lending Protocols

Understanding DeFi Lending Workflows: A Process Design Primer

Decentralized finance lending protocols have transformed how individuals and institutions access credit and earn yield on digital assets. Unlike traditional banking, these systems operate without intermediaries, relying on smart contracts to automate lending, borrowing, and liquidation. This guide compares the workflow design of three leading protocols—Aave, Compound, and MakerDAO—focusing on process differences that affect user experience, risk, and efficiency. We explore how each protocol handles collateralization, interest rate determination, and liquidation, providing actionable insights for developers, investors, and power users. Understanding these workflows is crucial for optimizing capital efficiency and avoiding costly mistakes. This overview reflects widely shared professional practices as of April 2026.

Core Components of a Lending Workflow

Every DeFi lending protocol shares a common set of components: a liquidity pool, an oracle for price feeds, a collateralization ratio, and a liquidation mechanism. However, the way these components interact differs significantly. For example, Aave uses a pooled lending model where lenders deposit assets into a common pool, and borrowers can choose between stable and variable interest rates. Compound employs a similar pool structure but uses a utilization rate to dynamically adjust interest rates. MakerDAO, on the other hand, uses a vault-based system where each collateral type is isolated, and users mint DAI against their collateral. These differences create distinct user experiences and risk profiles.

Why Workflow Design Matters

The workflow design directly impacts capital efficiency, gas costs, and liquidation risk. For instance, Aave's flash loans allow users to borrow without collateral as long as the loan is repaid within the same transaction, enabling complex arbitrage strategies. Compound's cTokens represent a user's deposit and accrue interest continuously, simplifying yield tracking. MakerDAO's vault system requires users to manage collateralization ratios manually, which can lead to liquidation if the ratio drops below the threshold. Choosing the right protocol depends on the user's goals: maximizing yield, minimizing risk, or executing advanced strategies. Practitioners often report that understanding these workflow nuances can reduce liquidation events by up to 30%.

Common Pitfalls in DeFi Lending Workflows

One frequent mistake is underestimating the impact of gas fees during liquidation events. On Ethereum, network congestion can cause transactions to fail, leading to unnecessary liquidations. Another is failing to account for oracle manipulation risks; some protocols use decentralized oracles like Chainlink, while others rely on centralized price feeds. Users should also be aware of the differences in interest rate models: Aave's stable rate provides predictability, while Compound's variable rate can spike during high utilization. These pitfalls highlight the need for careful process design when interacting with DeFi lending protocols. Always consult official documentation and consider using simulators before committing significant capital.

Comparing User Journeys: Aave, Compound, and MakerDAO

The user journey from depositing collateral to borrowing assets varies across protocols in terms of steps, complexity, and required interactions. This section compares the typical workflows for a user looking to borrow USDC against ETH collateral on each platform. We examine the number of transactions, gas costs, and decision points involved.

Aave Workflow: Streamlined but Feature-Rich

Aave's user journey begins with depositing ETH into the lending pool via a single transaction. Once deposited, the user receives aTokens that represent their deposit and accrue interest. To borrow USDC, the user specifies the amount and chooses between stable and variable interest rates. Aave then checks the collateralization ratio (typically 125% for ETH) and executes the borrow transaction. The entire process involves two transactions (deposit and borrow) and can be completed in under a minute on Layer 2 solutions like Arbitrum. Aave also offers features like swap-to-credit and debt tokenization, which add flexibility but increase complexity. One team I read about used Aave's flash loans to refinance debt across multiple protocols, saving 20% on interest costs.

Compound Workflow: Simplicity with Dynamic Rates

Compound's workflow is similar but with key differences. Users deposit ETH and receive cETH, which automatically earns interest based on the protocol's utilization rate. To borrow USDC, users must first enable the asset as collateral (if not already enabled) and then specify the borrow amount. Compound uses a single interest rate model that adjusts continuously based on supply and demand. The process involves at least two transactions: deposit and borrow. However, users can also use Compound's native token, COMP, to participate in governance, adding a layer of engagement. A common complaint is that the variable rate can spike unexpectedly during high utilization, leading to higher borrowing costs. For example, during the May 2021 market crash, Compound's borrow rate for USDC surged to 30% APY.

MakerDAO Workflow: Vault-Based Flexibility

MakerDAO's vault workflow is more manual but offers greater control. Users create a vault by depositing ETH (or other approved collateral) and selecting the amount of DAI to generate. The collateralization ratio must be maintained above 150% (for ETH-A). Users can adjust the ratio by adding more collateral or repaying DAI at any time. Unlike Aave and Compound, MakerDAO does not have a variable interest rate; instead, the stability fee is set by governance and applies to all DAI borrows. The process requires multiple transactions: approve collateral, open vault, generate DAI, and manage the position. This manual oversight can be a drawback for passive users but offers flexibility for active traders. For instance, a user might open multiple vaults to isolate risk across different assets.

Key Differences and Trade-offs

The choice between these protocols depends on the user's risk tolerance and activity level. Aave is best for users who want feature-rich tools like flash loans and stable rates. Compound is ideal for those who prefer simplicity and automated interest accrual. MakerDAO suits users who want to mint a stablecoin (DAI) and manage their debt actively. In terms of gas costs, MakerDAO's vault creation is typically more expensive due to multiple contract interactions, while Aave and Compound are comparable. Liquidation processes also differ: Aave uses a discount mechanism where liquidators can purchase collateral at a discount; Compound uses a similar system; MakerDAO auctions collateral through a decentralized auction mechanism. These nuances affect the likelihood and cost of liquidation.

Step-by-Step: A Typical Borrow Workflow on Aave

This section provides a detailed, actionable walkthrough of borrowing USDC against ETH on Aave, highlighting decision points and best practices. The steps are designed to be followed by an intermediate DeFi user.

Step 1: Connect Wallet and Approve ETH

First, connect your wallet (e.g., MetaMask) to the Aave app. Navigate to the 'Deposit' tab and select ETH. Approve the Aave contract to spend your ETH by signing an ERC-20 approval transaction. This is a one-time setup per asset. Gas costs vary but expect around $5-$15 on Ethereum mainnet. For lower fees, consider using Polygon or Arbitrum versions of Aave.

Step 2: Deposit ETH and Receive aETH

After approval, deposit the desired amount of ETH. The transaction will mint aETH tokens in your wallet, which represent your deposit and earn interest. The amount of aETH received is determined by the current exchange rate (which increases over time as interest accrues). For example, depositing 10 ETH might yield 10.025 aETH after a week, depending on the supply APY.

Step 3: Configure Borrow Parameters

Go to the 'Borrow' tab and select USDC. You will see two options: stable rate and variable rate. Stable rate offers predictable monthly payments but is typically higher than variable rate. Variable rate fluctuates with utilization but can be lower during periods of low demand. Choose based on your risk preference. If you expect rates to rise, lock in a stable rate; if you anticipate stable or falling rates, choose variable. Most practitioners recommend starting with variable and monitoring rates weekly.

Step 4: Execute Borrow and Manage Position

Enter the amount of USDC to borrow, ensuring your collateralization ratio remains above the liquidation threshold (typically 80% for ETH). Aave will display the current ratio; aim for at least 150% to avoid liquidation in volatile markets. Confirm the transaction. Once borrowed, USDC will appear in your wallet. You can repay at any time by navigating to the 'Repay' tab. Regularly check your health factor (a metric that combines collateral and debt value) and consider setting up alerts for when it drops below 1.5.

Step 5: Monitoring and Adjusting

After borrowing, monitor your position daily. Use tools like DeBank or Zapper to track your health factor. If the health factor approaches 1, you may need to add more collateral or repay part of the loan to avoid liquidation. Aave's liquidation penalty is 5-10% of the borrowed amount, which can be significant. Consider using automated strategies like Gelato's keepers to maintain a safe ratio. For example, one composite scenario involved a user who set up a keeper to add collateral when the health factor dropped below 1.2, preventing liquidation during a flash crash.

Risk Management in Lending Workflows: Liquidation and Collateralization

Risk management is central to DeFi lending, with liquidation being the most critical event. This section examines how different protocols handle liquidation thresholds, penalties, and processes, and offers strategies to mitigate risk.

Liquidation Thresholds and Penalties

Each protocol defines a liquidation threshold (the collateralization ratio at which liquidation is triggered) and a penalty fee. Aave typically uses an 80% threshold (e.g., for ETH, liquidation occurs when debt exceeds 80% of collateral value) with a 5-10% penalty. Compound uses a 75% threshold for ETH, with a 5% liquidation bonus for liquidators. MakerDAO's liquidation ratio is 150% for ETH-A, and the penalty is a 13% stability fee plus a liquidation penalty that can be up to 20% during auctions. These numbers are subject to governance changes. Users should check current parameters on each protocol's dashboard before borrowing.

Liquidation Process Comparison

Aave and Compound use a 'partial liquidation' model where liquidators can repay a portion of the debt (up to 50% on Aave) in exchange for the equivalent collateral plus a bonus. MakerDAO uses a 'collateral auction' where liquidators bid for the collateral, and the protocol sells it to cover the debt. The auction mechanism can result in better prices for the borrower if competition is high, but it also adds complexity. In practice, Aave's partial liquidation is faster and more predictable, while MakerDAO's auction can lead to worse outcomes during low-liquidity periods. One composite scenario: during a market downturn, a MakerDAO vault with 151% collateralization was liquidated at a 15% discount because only one bidder participated.

Strategies to Avoid Liquidation

To minimize liquidation risk, maintain a comfortable buffer above the threshold. A practical rule is to keep the health factor above 2 for volatile assets. Use stablecoins as collateral if possible, as they have lower volatility. Monitor oracle prices closely; some protocols use price feeds that lag during rapid moves. Consider diversifying collateral across multiple vaults or protocols to isolate risk. For advanced users, flash loans can be used to repay debt temporarily and adjust positions, but this requires technical expertise. Always simulate worst-case scenarios using tools like Gauntlet or RiskDAO's dashboards. Remember that liquidation is not a failure if managed proactively; it's a safety mechanism that protects lenders.

Interest Rate Models: How Protocols Determine Borrow Costs

Interest rate models are the engine of DeFi lending, determining the cost of borrowing and the return on lending. This section compares the models used by Aave, Compound, and MakerDAO, explaining how they work and how to predict rate changes.

Aave's Dual-Rate Model

Aave offers two interest rate options: stable and variable. The stable rate is fixed for a user for a set period (e.g., 30 days) and is determined by a formula that considers the overall utilization and the user's borrowing history. The variable rate is based on a utilization curve: as the pool's utilization increases, the rate rises exponentially. Aave's model uses a kink point (typically 80% utilization) where the slope increases sharply to discourage further borrowing. This dual-rate system gives users flexibility: stable rate for predictability, variable rate for potential savings.

Compound's Utilization-Based Model

Compound uses a single utilization-based model with a kink at around 90% utilization. The interest rate is calculated as a linear function below the kink and an exponential function above it. This means rates remain relatively stable until the pool is nearly full, then spike quickly. Compound's model is simpler than Aave's but offers less flexibility. The advantage is that lenders always earn the same rate as borrowers pay (minus a reserve factor), simplifying yield calculations. However, borrowers face the risk of sudden rate increases during high demand.

MakerDAO's Governance-Based Stability Fee

MakerDAO's interest rate is called the stability fee and is set by MKR token holders through governance. It is a flat fee applied annually to all DAI borrows, regardless of the vault's collateralization ratio. The stability fee is adjusted periodically based on market conditions and the demand for DAI. This model is transparent but subject to governance delays, which can lead to mispricing during fast-moving markets. For example, during the March 2020 crash, the stability fee was too low relative to market rates, leading to a surge in DAI demand and a deviation from its peg.

Predicting Rate Changes

To predict rate changes, monitor on-chain metrics like utilization rate, total supply, and governance proposals. For Aave and Compound, watch for utilization approaching the kink point, which signals an impending rate increase. For MakerDAO, follow governance forums for stability fee proposals. Tools like LoanScan and DeFi Rate provide historical rate data. Practitioners often set up alerts when utilization crosses 70% for Aave or 80% for Compound, giving them time to adjust positions. Understanding these models allows users to optimize borrowing costs and lending yields.

Oracle Design and Price Feed Workflows

Oracles are critical for DeFi lending, as they provide the price data that determines collateralization ratios and triggers liquidations. This section compares how Aave, Compound, and MakerDAO source and update price feeds, and the risks involved.

Aave's Oracle Integration

Aave uses a price oracle contract that aggregates data from multiple sources, primarily Chainlink. The oracle updates prices every few minutes or when significant price changes occur. Aave's oracle is designed to be resistant to manipulation by using a median of multiple feeds and a time-weighted average price (TWAP) for some assets. However, during extreme volatility, the oracle may lag, leading to stale prices. Aave also allows governance to add new oracles for new assets.

Compound's Oracle Model

Compound initially used its own oracle system (Open Oracle) but has since migrated to Chainlink for most assets. The Compound oracle returns the price from a trusted source (e.g., Chainlink) with a configurable deviation threshold. If the price deviates beyond a set percentage, the oracle updates immediately. Compound's oracle is considered reliable but has faced criticism for its reliance on a single source for some assets. In 2020, a price manipulation attack on Compound's oracle resulted in a temporary mispricing, highlighting the need for redundancy.

MakerDAO's Oracle Security Framework

MakerDAO uses a decentralized oracle system called the MakerDAO Oracle Module, which relies on multiple independent price feeds submitted by whitelisted oracles. The median of these feeds is taken, and the system updates the price when a threshold deviation is reached. MakerDAO's oracle is known for its security but is slower to update than Chainlink. The protocol also uses a 'price feed delay' to prevent flash loan attacks. For example, the ETH/USD feed updates every hour unless the price moves more than 1%, in which case it updates immediately. This design balances accuracy with gas efficiency.

Risks and Best Practices

Oracle manipulation is a top risk in DeFi lending. Attackers can use flash loans to manipulate prices on decentralized exchanges, causing false price feeds. To mitigate this, protocols use TWAP oracles and circuit breakers. Users should be aware of the oracle types used by their chosen protocol and consider diversifying across protocols with different oracle designs. For high-value positions, consider using protocols with a proven track record of oracle security, such as MakerDAO. Always monitor for oracle-related governance proposals that might change price feed parameters.

Governance and Upgrade Workflows in Lending Protocols

DeFi lending protocols are not static; they evolve through governance processes that can change parameters, add assets, or upgrade contracts. This section compares the governance workflows of Aave, Compound, and MakerDAO, focusing on how decisions are made and implemented.

Aave Governance: Aavegotchi and AIPs

Aave's governance is managed through Aave Improvement Proposals (AIPs) and the Aavegotchi token (AAVE). Anyone can submit an AIP, which is then debated in the community forum. If it passes a temperature check, it moves to an on-chain vote where AAVE holders can vote. The voting period is 3 days, and a quorum of 2% of total AAVE supply is required. Once approved, the proposal is executed by a smart contract that implements the change. Aave's governance is relatively fast, with parameter changes (like interest rate curves) being implemented within a week.

Compound Governance: COMP and Timelocks

Compound's governance uses the COMP token and a timelock contract. Proposals are submitted on-chain and require a minimum of 1% of COMP supply to propose. Voting lasts 3 days, and a majority (50%+1) is needed for approval. Once approved, the proposal enters a timelock of 2 days before execution, allowing users to exit if they disagree. Compound's governance is known for being decentralized, but the timelock period can delay critical changes, such as adding a new asset during a market opportunity.

MakerDAO Governance: MKR and Executive Votes

MakerDAO has a two-tier governance system: governance polls (for signaling) and executive votes (for implementation). MKR holders vote on proposals that change risk parameters, add collateral types, or adjust the stability fee. Executive votes are continuous; a proposal is executed when it has the most votes at the end of a voting period (typically 3 days). However, MakerDAO's governance is slower due to the need for multiple polls and community discussion. This can be a disadvantage during fast-moving markets but ensures thorough deliberation.

Impact on Users

Governance changes can affect users directly, such as when a protocol changes the collateralization ratio or adds a new asset. Users should participate in governance to have a say, or at least monitor changes that could affect their positions. For example, if MakerDAO proposes to increase the stability fee, borrowers may want to repay DAI before the change takes effect. Tools like Boardroom and Sybil provide governance dashboards that track proposals across protocols. Staying informed is part of effective risk management.

Advanced Workflows: Flash Loans and Credit Delegation

Beyond basic lending and borrowing, DeFi protocols offer advanced features like flash loans and credit delegation that enable complex strategies. This section explains these workflows and how they differ across protocols.

Flash Loans on Aave

Aave pioneered flash loans, allowing users to borrow any amount of an asset without collateral, provided the loan is repaid within the same transaction. The workflow involves: 1) initiate a flash loan from the Aave pool, 2) execute a series of actions (e.g., arbitrage, swapping), and 3) repay the loan with interest (0.09% fee). If the loan is not repaid, the transaction reverts. Flash loans are used for arbitrage, liquidations, and collateral swaps. For example, a user can flash loan USDC, buy ETH on a DEX, deposit ETH as collateral on Aave, borrow more USDC, and repay the flash loan—all in one transaction. This requires advanced Solidity knowledge and careful gas optimization.

Credit Delegation on Aave

Aave also offers credit delegation, where a lender can delegate borrowing power to a specific borrower, allowing them to borrow without posting collateral. The workflow involves a credit delegation agreement that specifies the terms (maximum amount, interest rate, duration). The borrower can then draw funds from the pool up to the delegated amount. This feature is still nascent but has potential for institutional lending, where a trusted entity can borrow on behalf of a client. The risk is that the borrower defaults, and the lender loses the delegated funds.

Compound's cToken Wrapping

Compound does not offer flash loans natively, but its cTokens can be used in combination with other protocols to create similar effects. For example, a user can deposit ETH into Compound, receive cETH, and then use cETH as collateral in a separate protocol like MakerDAO to mint DAI. This cross-protocol strategy is common but adds complexity and gas costs. Compound's focus is on simplicity, so advanced users often combine it with other protocols for more sophisticated workflows.