Compound Finance Borrowing Review

Introduction

Compound Finance pioneered algorithmic money markets in 2018, allowing anyone with a Web3 wallet to borrow crypto against deposited collateral without an intermediary. Seven-plus years on, it remains one of the most battle-tested lending platforms in DeFi, even though it is now roughly a twelfth of Aave's size. If you hold ETH, WBTC, or stablecoins and need liquidity without triggering a taxable sale, Compound lets you borrow against those holdings at variable rates that currently sit around 4-6% APY for stablecoins.

What makes Compound worth considering over centralised alternatives? Three things stand out. First, the interest rate model is fully algorithmic — Compound III accrues interest every second using the block timestamp, at a rate driven by the ratio of borrowed funds to supplied funds, so you always see transparent, market-driven pricing. Second, the protocol is non-custodial: your collateral sits in audited smart contracts, not on a corporate balance sheet that could freeze withdrawals. Third, every parameter — collateral factors, rate slopes, the kink — is readable straight from the contract, so you can verify the terms instead of trusting a marketing page.

What is no longer a reason to pick Compound is COMP: the borrower subsidy has shrunk to roughly 0.1% APY on the main Ethereum markets and to nothing on several others.

Compound III, called Comet, is the version you actually use. It redesigned the architecture around isolated markets — each market has one borrowable base asset (Ethereum runs six: USDC, USDT, USDS, WETH, WBTC and wstETH) with its own accepted collateral set, and it issues no cTokens. V2, the original cToken model, was deprecated on-chain in December 2025: new supply and borrow are paused across every V2 market and reserve factors were raised to 100%, so the only actions left there are repay and withdraw. The money has followed — roughly $1.12 billion in V3 against $89 million of stranded V2 balances in August 2026. Check which V3 market suits your collateral mix before depositing.

This review covers Compound's borrowing mechanics in detail: how the three per-asset collateral parameters work, what rates you can realistically expect, why a V3 liquidation takes the whole position, and where Compound falls short compared to Aave. Whether you are evaluating your first DeFi borrow or comparing protocols for a larger position, the sections below give you the concrete numbers and trade-offs you need.

Compound Protocol Overview

If you are looking for a battle-tested DeFi lending protocol, Compound Finance deserves your attention. Launched in 2018 by Compound Labs, it has facilitated billions of dollars in lending and borrowing transactions across seven-plus years.

When you borrow through Compound, your assets remain under your control at all times. The protocol operates on Ethereum and several Layer 2 networks, using algorithmic interest rate models that automatically adjust based on supply and demand.

Compound catalysed DeFi Summer 2020 by pioneering liquidity mining — distributing COMP tokens to both suppliers and borrowers proportionally to interest accrued. That mechanism became the template for hundreds of subsequent token launches. Beyond its historical influence, the protocol's architecture prioritises simplicity: in V3 you supply an accepted collateral asset, then borrow that market's base asset. There is no cToken to mint, no enable-as-collateral toggle, and no complex position management screen — just standard token approvals.

Key Statistics

• Total Value Locked: around $1.2 billion across V3 and V2 (DeFiLlama, August 2026)
• Supported Assets: six V3 base markets on Ethereum (USDC, USDT, USDS, WETH, WBTC, wstETH), each with its own collateral set
• Outstanding borrows: around $560 million across V3 (DeFiLlama, August 2026)
• Governance Token: COMP with 10 million total supply
• Security Audits: Multiple audits by Trail of Bits, OpenZeppelin, and others

Governance Model

COMP holders govern the protocol directly. Governance moved off Governor Bravo in February 2025 to a new Compound Governor contract, which keeps the same numbers: a proposal threshold of 25,000 COMP of delegated voting power and a quorum of 400,000 votes in favour, with roughly two days of voting delay and under three days of voting. In practice, only well-capitalised entities or coalitions can propose changes — which keeps frivolous proposals out but concentrates power amongst large holders. Recent proposals have gone the other way from expansion: zeroing COMP emissions outside Ethereum, cutting rsETH and USDe collateral factors to zero, retiring several Layer 2 deployments, and funding a $52 million Compound V4 development programme in May 2026.

For most borrowers, governance matters primarily when it affects the assets you use. A proposal could change your collateral factor, adjust the liquidation penalty, or wind down the market you are sitting in — several were retired during 2026. You can monitor active proposals at compound.finance/governance and delegate your COMP voting power even if you do not meet the proposal threshold yourself.

Historical Milestones

• 2018: Protocol launch with initial 5 assets
• 2020: COMP token distribution begins, pioneering "liquidity mining"
• 2021: All-time peak TVL of roughly $12.3B on 9 November (DeFiLlama)
• 2022: Compound III (Comet) launch with improved capital efficiency
• 2023-2025: Expansion to Layer 2 networks (Polygon, Arbitrum, Base)
• 2025-2026: V2 deprecated on-chain in December 2025; COMP emissions zeroed outside Ethereum and the Scroll, Linea, Mantle, Polygon and Unichain markets retired

If you are new to DeFi lending, Compound's documentation at docs.compound.finance walks through every concept above with interactive examples. For experienced users, the open-source contracts on GitHub allow you to verify parameters and build integrations directly.

Key Borrowing Features

Algorithmic Interest Rates

When you borrow on Compound, your interest rate adjusts automatically based on supply and demand. As borrowing demand increases, rates rise to incentivise more supply, so you always receive market-driven pricing without manual intervention.

COMP Token Rewards

As a borrower on some markets you still accrue COMP, distributed in proportion to interest accrued. Treat it as a governance vote rather than a discount. Three markets emit anything at all, all on Ethereum: USDC 55 COMP per day, USDT 30 and WETH 20, split between the supply and borrow sides. That is worth roughly 0.1% APY against outstanding borrows in August 2026. The Ethereum USDS, WBTC and wstETH markets emit nothing, and neither does any Layer 2 deployment — governance zeroed them through 2026.

Supported Assets

• Borrowable base assets on Ethereum: USDC, USDT, USDS, WETH, WBTC, wstETH
• Collateral in the Ethereum USDC market: WETH, wstETH, WBTC, cbBTC, tBTC, weETH, LINK, UNI, COMP
• Borrow collateral factors in that market: 50% on COMP up to 82.5% on WETH
• Four further assets (rsETH, USDe, deUSD, sdeUSD) are still registered in that market but carry a zero borrow factor and a zero supply cap — they give you no borrowing power, and deUSD and sdeUSD are being wound down
• DAI is not a Compound III base market or collateral type on Ethereum

cToken System (V2) vs Comet Balances (V3)

In Compound V2, supplying assets gave you cTokens (cUSDC, cETH) — ERC-20 tokens whose exchange rate rose over time as interest accrued. Supply 1,000 USDC, receive 50 cUSDC at a 20:1 exchange rate, and those 50 cUSDC became redeemable for 1,050 USDC, 1,100 USDC and so on. That model is now history rather than a live choice. The December 2025 deprecation set every V2 reserve factor to 100%, which means the supply rate on the major cTokens reads zero: cUSDC, cDAI, cETH, cUSDT and cWBTC2 holders no longer earn anything, and mint and borrow are both paused. If you still hold cTokens, the only useful action is to redeem them.

Compound V3 (Comet) eliminated cTokens entirely. Instead, each market tracks your supply balance internally and accrues interest directly to your account. V3 markets are structured around a single borrowable base asset — the USDC market on Ethereum, for example, accepts WETH, wstETH, WBTC, cbBTC, tBTC, weETH, LINK, UNI and COMP as collateral but only lets you borrow USDC. This isolated-market design means a problem with one collateral type cannot cascade into losses for the entire protocol, unlike V2 where all assets shared a common risk pool. If you are starting fresh, V3 is the only option: V2's roughly $89 million is stranded collateral awaiting withdrawal, not active lending.

Collateral Management

Each collateral asset carries three separate percentages, and running them together is the classic sizing error. The borrow collateral factor sets your borrowing power. The higher liquidation collateral factor is the level at which the position becomes seizable. The third, called liquidationFactor in the contract, is not a threshold at all: it is the share of seized collateral credited back to you, so one minus it is the penalty. On the Ethereum USDC market:

• WETH: 82.5% borrow, 88% liquidation, 7% penalty — borrow up to $8,250 per $10,000 supplied
• WBTC: 80% borrow, 85% liquidation, 10% penalty; wstETH: 82% / 86% / 9%
• LINK: 73% borrow, 79% liquidation, 17% penalty; UNI: 68% / 74% / 17%
• COMP: 50% borrow, 70% liquidation, 25% penalty — the protocol discounts its own token heavily

You can supply several collateral assets in the same market simultaneously, and your borrowing power is the sum of each collateral value multiplied by its own borrow factor. The gap between the borrow and liquidation collateral factors is your buffer. Check the market you are actually in, because the same asset is priced differently across Comets: WETH takes an 82.5% borrow factor in the Ethereum USDC market and 83% in the USDT and USDS markets.

Multi-Chain Deployment

Compound operates on multiple networks, each with different characteristics:

• Ethereum Mainnet: deepest liquidity (about $1.0bn of the $1.12bn V3 total) and six base markets; base fees sat near 0.1 gwei in August 2026, so a supply or borrow cost cents rather than dollars
• Arbitrum: second-largest deployment at roughly $65m, with USDC, USDT and WETH markets
• Base: roughly $20m across USDC, USDS, WETH and AERO markets
• Optimism: roughly $8m; Ronin: under $0.2m — small but not scheduled for closure
• Being retired: governance has executed deprecations of Scroll (February 2026), Linea and Mantle (July 2026) and Polygon and Unichain (July 2026); their collateral caps are zero and the supply rate curve is set to zero, so suppliers earn nothing there

Choose the network on liquidity rather than gas. With mainnet fees near zero through mid-2026, the old cost case for Layer 2 has largely gone, and what remains is depth: outside Ethereum and Arbitrum, a large borrow can push utilisation past the kink and move your own rate against you. Do not open a new position on a deprecated market — you would be lending at 0% while borrowers there still pay for the reserve. Gas is volatile, though, and has spiked by orders of magnitude in past congestion.

Interest Rates & Fees

Current Borrowing Rates (Indicative)

• USDC: 4.1% borrow APY, 3.3% supply APY, at 89% utilisation
• USDT: 4.0% borrow APY, 3.2% supply APY, at 87% utilisation
• USDS: 6.2% borrow APY, 5.2% supply APY, at 91% utilisation — above the kink
• WETH: 1.7% borrow APY, 1.3% supply APY, at 66% utilisation
• WBTC: 1.3% borrow APY, 0.3% supply APY, at 22% utilisation
• wstETH: 1.2% borrow APY, 0.3% supply APY, at 14% utilisation

These are spot readings from the six Ethereum Comet contracts on 4 August 2026 and move continuously with utilisation. Compound III accrues interest every second using the block timestamp. COMP rewards now shave only about 0.1% off the USDC, USDT and WETH cost. Always verify current rates at app.compound.finance before committing.

How the Utilisation Curve Works

Compound uses a kinked interest rate model, and its parameters are readable straight from the contract. The Ethereum USDC market puts its kink at 90% utilisation, with a 1.5% base rate and a low slope of 2.78%, so borrow cost climbs gently from about 2.6% APR at 40% utilisation to 4.0% APR at the kink itself. Past 90% the high slope of 360% takes over: each further percentage point of utilisation adds 3.6 points of borrow rate, which puts the market near 22% APR by 95%. That cliff is what keeps liquidity available for withdrawals.

Do not carry those parameters to another Comet. The kink is 90% on the Ethereum USDC, USDT, USDS, WETH and wstETH markets but 85% on WBTC, and the high slope is 360% a year on the three stablecoin markets against 30% on WETH and wstETH and 115% on WBTC — the same mechanism, a spike an order of magnitude gentler. The eight rate parameters are governance-settable and readable on each market's contract, so you can check yours rather than assume. USDC utilisation sat at 89% in early August 2026, just under its kink, and close enough that a large borrow could tip the market into the steep zone and take your own rate with it.

Fee Structure and Liquidation Mechanics

• Borrowing fees: None beyond variable interest
• Liquidation penalty: set per collateral asset, not per protocol. On the Ethereum USDC market: 7% on WETH, 9% on wstETH, 10% on WBTC, cbBTC, tBTC and weETH, 17% on LINK and UNI, 25% on COMP
• Close factor: none at all — Compound V3 has no partial liquidation, as the worked example below sets out
• Gas costs: cents on Ethereum mainnet at the sub-1 gwei base fees of mid-2026, and lower still on the Layer 2 deployments
• Withdrawal fees: None

The two thresholds work as set out above. You may borrow up to your collateral times its borrow collateral factor, but you only become liquidatable once your debt exceeds your collateral times the higher liquidation collateral factor. Supply $10,000 of WETH into the USDC market and you can draw $8,250 at the 82.5% borrow factor, yet liquidation only bites when 88% of your collateral falls below your debt. On a $7,000 borrow that means a WETH fall of roughly 20.5%, to $7,955.

At that point any liquidator can call absorb, and V3 does not take a slice. It seizes every collateral asset in the account, credits you only the liquidationFactor share of each — 93% for WETH, so a 7% penalty — repays the debt in full out of that value, and returns the surplus as a withdrawable USDC balance. On $7,955 of WETH the penalty costs about $557, clears the $7,000 debt, and leaves roughly $400. It therefore scales with your entire collateral balance, not with the amount repaid.

Rate Comparison with Aave

On the readings taken for this review the two protocols price closely, with no consistent winner. USDC borrowed at 4.06% APY on Compound V3 against 4.23% on Aave V3; USDT at 3.99% against 3.69%; WETH at 1.75% against 2.14%. The old rule of thumb — that Compound charges a premium which COMP rewards then repay — no longer holds in either direction. The spread is now asset-specific and small, and the COMP subsidy is far too thin to swing it. Compare live numbers on the asset you actually want rather than assuming a protocol-level gap.

How to Borrow on Compound

Step-by-Step Guide

• Connect Wallet: Visit app.compound.finance and connect MetaMask or WalletConnect-compatible wallet
• Choose Market: pick the base asset you intend to borrow — USDC, USDT, USDS, WETH, WBTC or wstETH on Ethereum. Each market has its own collateral list
• Supply Collateral: deposit an accepted collateral asset. V3 issues no cTokens and has no enable-as-collateral toggle, and collateral earns no interest — only the base asset does
• Borrow: max borrow = the sum of each collateral value × its own borrow factor (50-82.5% on the Ethereum USDC market), subject to that market's floor — 100 USDC on Ethereum USDC
• Monitor Your Buffer: the gap between the borrow collateral factor and the higher liquidation collateral factor is all the room you have
• Repay: repay anytime to free collateral. Interest accrues every second

Example Scenario

If you supply $10,000 of WETH into the Ethereum USDC market, your maximum borrow is $8,250 at the 82.5% borrow factor. A safer operating limit is $6,200-6,600, which leaves room for a 25-30% WETH fall before the 88% liquidation collateral factor is breached. COMP accrues on the borrow side only, since V3 collateral earns neither interest nor rewards, and at August 2026 emissions and prices that comes to roughly 0.1% APY — a rounding error rather than a subsidy. On a $6,500 borrow at the current 4.1% USDC rate, your annual interest is about $267 before any COMP offset.

Best Practices

• Draw no more than 70-75% of your borrowing power, so you keep a buffer between the borrow factor and the higher liquidation collateral factor
• Monitor the liquidation collateral factor of every asset you have posted, not just the largest
• Claim any accrued COMP before closing a position, but do not size the position around it
• Use stablecoins as collateral for lower volatility risk

Advanced Strategies

Recursive Borrowing: not available inside a V3 market. A market's base asset cannot be posted as collateral in that same market, so the classic supply-borrow-resupply loop needs a second protocol or a leverage router. Risk: higher liquidation exposure and an extra contract in the chain.

Yield Farming: borrowing USDC at roughly 4.1% only pays if the destination genuinely clears that after fees, and mainstream stablecoin venues no longer do — Curve's 3pool was yielding effectively nothing in August 2026. Risk: smart contract exposure across multiple protocols for a spread that may not exist.

Hedging: Supply ETH, borrow USDC to lock in USD value while maintaining ETH exposure. Useful for tax planning or temporary liquidity needs without selling.

Common Mistakes to Avoid

• Borrowing at maximum capacity (leaves no buffer for price volatility)
• Ignoring gas costs (can eat into profits on small positions)
• Treating COMP as a subsidy — at roughly 0.1% APY it will not rescue a bad rate
• Using volatile assets as collateral without monitoring
• Forgetting to repay before major market moves

Security & Track Record

Security Audits

• Trail of Bits: Comprehensive smart contract audit (2019, 2020)
• OpenZeppelin: security partner since December 2021, reviewing governance proposals continuously — its 2025 renewal reported six critical vulnerabilities found in a single year, which is a better safety anchor than "no hacks"
• Certora: Formal verification of protocol logic
• Bug Bounty: Immunefi programme paying $50,000 to $1,000,000 for critical smart contract bugs — 10% of the funds at risk, capped at $1m

Track Record

The honest version is two clauses. No exploit has ever drained user deposits from Compound's lending contracts in seven-plus years across V1, V2 and V3 — which is a genuinely rare record. But the protocol has had serious incidents, none of them an outside attack on deposits.

In September 2021 a Comptroller upgrade shipped with a comparison operator the wrong way round and over-distributed protocol-owned COMP; published estimates of the loss range from roughly $50m to $90m, and some recipients returned tokens voluntarily. Depositors lost nothing; the treasury did. In August 2022 an oracle upgrade returned empty data for cETH and froze Compound V2's largest market for about a week, because the fix had to clear the governance timelock. Governance latency is itself a risk surface.

Two more in July 2024. A delegate bloc pushed a proposal to move about $24m of DAO COMP into their own vault; it passed before being cancelled, and a proposal guardian was added to the governor afterwards. The same month the compound.finance domain was DNS-hijacked and served a phishing front end — contracts untouched, but anyone who signed on the spoofed site was at risk. Bookmark the app rather than searching for it.

Risk Factors

• Smart contract risk: Code vulnerabilities despite audits
• Oracle risk: Price feed manipulation could trigger false liquidations
• Governance risk: Malicious proposals could change protocol parameters
• Liquidation risk: Market volatility can trigger cascading liquidations

Insurance Options

Discretionary smart contract cover for protocols like Compound does exist and is worth pricing if you hold a large position. We are not naming providers or a premium band here: this market has consolidated sharply, cover capacity moves, and we could not confirm current terms at the time of writing. Check quoted premiums and, more importantly, the exact wording of what triggers a payout directly with whichever underwriter is still listing Compound cover.

Risk Mitigation Strategies

Diversify Collateral: Don't put all collateral in one asset. Mix stablecoins with ETH/WBTC to balance volatility and yield.

Set Price Alerts: Use tools like DeFi Saver or Instadapp to monitor how much of your borrow capacity is used and receive alerts before liquidation risk.

Maintain Buffer: Never draw more than 70-75% of your borrowing power, and check the position daily during volatile markets.

Pros and Cons

Advantages

• Battle-tested: since 2018, and no exploit has ever drained user deposits from the lending contracts
• COMP tokens: a governance vote, not a subsidy — worth roughly 0.1% APY on the three markets that still emit
• Simple interface: Easy to use for DeFi beginners
• Transparent: All rates and parameters visible on-chain
• Non-custodial: Full control of your assets
• Governance rights: COMP holders vote on protocol changes
• Multiple audits: Extensive security reviews

Disadvantages

• No rate advantage: Compound and Aave now price within a few tenths of a point of each other, and the COMP subsidy no longer bridges any gap
• Narrower asset range: six base markets on Ethereum, and each one lets you borrow only its own base asset
• Whole-position liquidation: absorb seizes all your collateral at a 7-25% penalty. Aave caps a single liquidation at 50% of the debt only while your health factor is above 0.95 and both sides of the pair exceed $2,000 — below either, Aave can clear the whole position too
• Collateral earns nothing: only the base asset accrues supply interest in V3
• Thin Layer 2 liquidity: every non-Ethereum deployment holds well under $70m
• No flash loans: Missing advanced DeFi features
• Lower TVL: Compound's roughly $1.2B against the Aave family's roughly $14.5B (DeFiLlama, 4 August 2026 — verify before sizing)

Compound vs Aave

When to Choose Compound

• You want a vote on the parameters that govern your own position
• You prefer simpler interface
• You're borrowing major assets (ETH, WBTC, stablecoins)
• You value protocol maturity and track record

When to Choose Aave

• You want a chance of partial liquidation rather than a guaranteed whole-position absorb — but only if you keep both sides of the pair above $2,000
• You want more asset options
• You need flash loans
• You want lower liquidation penalties (1-6% on Aave V3 against 7-25% on Compound V3)

Read full comparison: Compound vs Aave Detailed Analysis

Advanced Borrowing Strategies and Risk Management

Leveraged Yield Farming: Worked Example

The carry trade is the most common advanced Compound strategy, and in August 2026 it mostly does not clear. Take the standard setup: you supply $20,000 of WETH into the USDC market, giving you $16,500 of borrowing power at the 82.5% borrow factor, and draw $12,000 of USDC at 4.1%, costing about $492 a year.

The classic destination was Curve's 3pool, but Curve's own API and DeFiLlama both put total yield there at effectively zero in August 2026 — CRV gauge emissions to that pool have collapsed. COMP accrues only on the borrow side, since V3 collateral earns neither interest nor rewards, and adds roughly $13 a year. The carry is negative unless you find a venue clearing well above 4.1%, and whatever does is carrying the risk that explains the spread.

The risk on the collateral side is straightforward. If ETH drops 20%, your $20,000 becomes $16,000 and your borrowing power falls to $13,200 at the 82.5% borrow factor, leaving only $1,200 of headroom against a $12,000 debt. Liquidation uses the higher 88% liquidation collateral factor and does not bite until collateral reaches $13,636, a 32% drawdown. When it does, absorb takes the whole position at a 7% penalty rather than the roughly $2,000 of repayment that would have restored the buffer. You can mitigate this by borrowing at 60-65% of capacity instead of 73%, or by borrowing WETH against stablecoin collateral in the WETH market, which removes the volatility but earns nothing while posted.

Automated Liquidation Protection

Manual monitoring works for small positions, but if you borrow above $10,000, automation tools provide meaningful safety. DeFi Saver offers "Automation" for Compound positions — you set a target collateral ratio (e.g. 200%) and minimum/maximum thresholds (e.g. 160%/250%). When ETH drops and your ratio hits 160%, DeFi Saver automatically repays part of your debt using flash loans, restoring your ratio to 200%. When ETH rises and your ratio exceeds 250%, it borrows more and adds collateral to capture the upside. The service costs 0.25% per automated transaction — a reasonable premium against a V3 absorb, which is all-or-nothing.

Instadapp provides similar functionality through its "Automation" module and also supports cross-protocol debt migration — useful if Aave rates drop significantly below Compound, allowing you to move your position without manual unwinding. Both tools work on Ethereum mainnet and Arbitrum deployments.

Cross-Protocol Rate Arbitrage

Rate differentials between Compound and Aave open up on the same asset, especially when one protocol's utilisation spikes before the other. You can borrow on whichever charges less and supply on the other, pocketing the spread minus gas and minus the risk of running exposure to two protocols instead of one. On 4 August 2026 the gap was slim — USDC borrowing at 4.06% on Compound V3 against 4.23% on Aave V3 — so treat this as an opportunistic trade rather than a standing one. Gas is no longer the constraint it was: with mainnet base fees near 0.1 gwei, a supply-and-borrow round trip on Ethereum costs well under a dollar, much as it does on Arbitrum or Base.

Watch for rate convergence: these spreads rarely persist beyond 24-48 hours, as arbitrageurs close the gap by shifting capital. If you plan to run this strategy, set up rate monitoring through DefiLlama's rates dashboard or the Compound and Aave subgraph APIs to receive alerts when spreads exceed your target threshold.

Technical Architecture and Protocol Mechanics

How Compound V3 (Comet) Differs Under the Hood

V3 restructured the protocol around isolated markets. Each Comet deployment serves one borrowable base asset, and there are now many: Ethereum alone runs USDC, USDT, USDS, WETH, WBTC and wstETH markets, with further deployments on Arbitrum, Base, Optimism, Polygon, Linea, Mantle, Scroll, Ronin and Unichain. The Ethereum USDC market accepts WETH, wstETH, WBTC, cbBTC, tBTC, weETH, LINK, UNI and COMP as collateral, each with its own borrow factor, liquidation factor and penalty set independently. If LINK crashes, only borrowers using LINK as collateral face liquidation risk; WETH-collateralised borrowers in the same market are unaffected. This isolation was the primary design motivation, addressing V2's weakness where a single bad asset could threaten the entire pool.

V3 also cut the gas cost of a supply or borrow relative to V2, because the protocol tracks balances internally rather than minting and burning cTokens, which means fewer storage writes per operation. That efficiency matters less than it used to: with Ethereum base fees near 0.1 gwei in August 2026 and a borrow consuming a few hundred thousand gas, the transaction costs a few cents. The same operation cost tens of dollars when the network ran at 30-50 gwei, and would again in a congestion episode.

Oracle and Liquidation Architecture

Compound V3 reads Chainlink price feeds directly through its getPrice function, one feed per collateral asset, returning USD prices to eight decimals. There is no time-weighted average and no automatic stale-price pause in the documented design: the protocol takes Chainlink's latest answer as the price, so oracle quality is a question of that feed's deviation threshold and heartbeat rather than anything Compound layers on top. Pausing supply, borrow or absorb is a guardian and governance action, not an automatic response to a stale feed. If you are collateralising a thinner asset, check its feed's update parameters before assuming the screen price is the contract price.

Compound V3 has no close factor — the parameter does not exist on the contract, and there is no partial-liquidation path. Once your debt exceeds the sum of your collateral valued at each asset's liquidation collateral factor, anyone can call absorb, and the protocol takes the account whole: it assumes all the debt, zeroes every collateral balance, and credits you the liquidationFactor share of each rather than market value.

Say you borrow $10,000 USDC against $15,000 of WETH. WETH falls 25% to $11,250, and 88% of that is $9,900 — under your debt, so you are absorbable. The protocol seizes the full $11,250, credits you 93% of it, or $10,462, clears the $10,000 debt and leaves $462 as a withdrawable USDC balance. The 7% penalty cost you $787.

Absorption and resale are two separate steps, and this is not a Dutch auction. After absorb, anyone holding the base asset can buy the seized WETH through buyCollateral at a fixed discount of storeFrontPriceFactor times the penalty — 0.6 x 7%, or 4.2% below the oracle price on the USDC market, with the remaining 2.8% accruing to reserves. The discount does not decay and there is no bidding. Treat the liquidation collateral factor as a cliff rather than a slope: one call ends the position, and the penalty falls on everything you posted.

Reserve Factor and Protocol Revenue

Compound keeps a spread between what borrowers pay and what suppliers earn. In V3 this is not a single reserve-factor parameter but the gap between the separate borrow and supply rate curves. On the Ethereum USDC market on 4 August 2026, borrowers paid 4.06% while suppliers earned 3.26% on a larger balance, so roughly a tenth of the interest borrowers paid accrued to protocol reserves rather than to suppliers.

Those reserves absorb bad debt when an absorbed account's collateral does not cover its borrow. They are held per market and read by calling getReserves() on each Comet — not through V2's Comptroller — and stood at about $8.2m on the USDC market, against a $20m target, and $1.2m on USDT. While reserves sit below target, buyCollateral stays open, which is the mechanism that recapitalises a market after an absorption.

Market Position and Competitive Analysis

Compound vs Aave: Where Each Protocol Wins

On 4 August 2026 Compound held roughly $1.2 billion across V3, V2 and V1, against roughly $14.5 billion for the Aave family — Aave is about twelve times larger by TVL and around twenty times larger by outstanding borrows. That gap reflects Aave's broader asset support, its presence on more networks, and features such as flash loans that Compound lacks.

What it does not reflect is price: the two protocols charged USDC borrow rates within a fifth of a percentage point of each other on that date, and Compound was the cheaper of the two. On a $50,000 USDC borrow over six months that gap came to roughly $42, and the COMP subsidy added about $28 over the same period.

Where Compound clearly wins: simplicity of interface, clean per-market risk isolation, and gas efficiency on V3. Where Aave clearly wins: asset variety, flash loan access, the possibility of partial rather than whole-position liquidation, and lower penalties — 1-6% on Aave V3 against 7-25% on Compound V3.

Read that partial-liquidation advantage carefully. Aave's 50% cap applies only while your health factor is above 0.95 and both the collateral and the debt in the liquidated pair exceed $2,000, and a partial that would leave under $1,000 on either side is converted to a full clear instead. On a small position Aave can take the lot just as Compound does. If you need long-tail collateral, or run a position large enough for the cap to engage, Aave is the more practical venue.

Newer Competitors: Morpho, Spark, and Euler V2

The lending market has expanded beyond the Compound-Aave duopoly. Morpho runs curated, isolated lending markets. Spark (the lending platform within Sky, the protocol formerly known as MakerDAO before its August 2024 rebrand) offers DAI or USDS borrowing at the Sky Savings Rate. Euler V2 provides customisable lending vaults with user-defined risk parameters. Each serves a specific niche, but none match Compound's combination of track record, liquidity depth, and per-market risk isolation. If you are managing over $100,000 in DeFi lending, splitting positions across Compound, Aave, and Morpho gives you rate optimisation and protocol diversification.

Regulatory Landscape and Compliance Considerations

Regulatory Status

Compound is permissionless — no KYC, no geographic restrictions enforced by the protocol itself. However, regulatory pressure is increasing. The EU's MiCA framework, effective from 2025, classifies DeFi front-ends as potentially regulated entities, though the underlying smart contracts remain outside direct regulatory scope. In the US, the SEC has pursued enforcement actions against centralised lending platforms but has not directly targeted non-custodial protocols like Compound. You should verify whether your jurisdiction restricts DeFi protocol usage, as some countries (e.g. China, parts of Southeast Asia) block access to DeFi front-ends even though the contracts remain accessible through direct wallet interaction.

Tax Implications for Borrowers

In most jurisdictions, receiving borrowed funds is not a taxable event — borrowing USDC against ETH collateral does not trigger capital gains on the ETH. However, COMP rewards you receive are typically taxable as income at the market value when claimed. If you receive 10 COMP at $50 each, that is $500 of taxable income regardless of whether you sell the tokens.

Subsequent price changes create a separate capital gains event when you dispose of the COMP. Track each COMP claim as a distinct tax lot using tools like CoinTracker or Koinly, which both support Compound transaction import. Liquidation events are also taxable — a forced sale of your collateral triggers capital gains or losses based on your original cost basis of the liquidated asset.

Conclusion

Compound Finance remains a strong choice for DeFi borrowing if your collateral appears on a V3 market's list and you want predictable, algorithmic rates. Do not choose it for the COMP subsidy, which has all but disappeared. The protocol's V3 Comet architecture provides isolated market risk — a vulnerability in one collateral asset cannot cascade into your position if you use a different collateral type. Seven-plus years of operation without a deposit-draining exploit, continuous OpenZeppelin review, and an Immunefi programme paying up to $1m for critical bugs give it a safety record newer lending protocols cannot match — though the incident history above is worth reading before you call it spotless.

Size the decision on the rate, not the reward. Borrow $20,000 of USDC at the August 2026 rate of 4.06% and you pay about $812 a year, while the COMP accruing against that borrow is worth roughly $22 at current emissions and prices. That is a rounding error rather than a subsidy, and it is the single biggest change from how this protocol was marketed in 2020 and 2021. What COMP still buys is a vote on parameters such as collateral factors and market deprecations, which is worth something if you hold a large position.

For UK investors specifically, Compound borrowing creates a favourable tax structure: borrowing against ETH avoids the CGT disposal that selling would trigger, whilst COMP rewards are taxable as miscellaneous income at receipt value. If you hold appreciated ETH (cost basis well below current price), the annual interest plus COMP income tax is almost always cheaper than the CGT that selling and rebuying would create — especially with the £3,000 annual CGT allowance now so low.

Practical Recommendations by Position Size

Pick the network on liquidity rather than gas. Mainnet fees near 0.1 gwei mean even a $2,000 borrow on Ethereum costs pennies to open, which removes the old reason to push small positions onto Layer 2. Arbitrum is the only sizeable non-Ethereum deployment at roughly $65m, Base is about $20m, and everything else is single-digit millions or being retired outright — thin enough that a moderate borrow can move the rate against you. For positions above $50,000, consider splitting between Compound and Aave to diversify smart contract risk — no single protocol should hold more than 60% of your DeFi lending exposure.

Regardless of position size, draw no more than 70-75% of your borrowing power. ETH-collateralised positions can experience 15-20% drawdowns within hours during market stress, and gas costs for emergency repayment spike during the same congestion periods. If your position exceeds $10,000, set up automated protection through DeFi Saver or Instadapp rather than relying on manual monitoring.

Best Use Cases

• Long-term borrowing: No fixed terms and no repayment schedule
• Stablecoin borrowing: Competitive rates on USDC and USDT
• ETH/WBTC collateral: Borrow factors of 82.5% and 80% on the USDC market
• DeFi beginners: Simple interface, clear documentation

Not Ideal For

• Borrowers needing a fixed rate (Compound has none, and Aave V3's stable rate is disabled)
• Borrowing exotic altcoins (limited asset selection)
• Flash loan strategies
• Positions you cannot monitor, since one absorb call takes the whole thing

Integration Ecosystem

Compound integrates with major DeFi tools and platforms:

• Wallets: MetaMask, Ledger, Trezor, WalletConnect
• Aggregators: Zapper, Zerion, DeBank for portfolio tracking
• Automation: DeFi Saver, Instadapp for automated position management
• Analytics: Dune Analytics, Token Terminal for protocol metrics
• Tax Tools: CoinTracker, Koinly for DeFi tax reporting

These integrations mean you rarely need to interact with Compound's contracts directly. Compound V3 publishes no health factor of its own, but Zapper and Zerion render your Comet position as a health-factor-style number alongside accrued interest and claimable COMP, in a single dashboard with your other DeFi positions.

Overall Rating: 4.0/5 — Compound delivers reliable, transparent DeFi borrowing at rates that now sit level with Aave's. It lacks Aave's asset breadth, its liquidation is all-or-nothing where Aave's is sometimes partial, and COMP no longer pays you anything meaningful for using it. What it still offers is a seven-year record with no deposit-draining exploit, V3's isolated-market architecture, and parameters you can read straight off the contract — enough to make it a sound choice for borrowing against ETH and WBTC, but no longer an automatic one.

Visit Compound Finance or read our Complete Crypto Borrowing Guide first.

Frequently Asked Questions

Is Compound Finance safe?

No exploit has ever drained user deposits from Compound's lending contracts since 2018, and OpenZeppelin reviews governance proposals continuously. The record is not spotless: a September 2021 Comptroller upgrade over-distributed protocol-owned COMP through a one-character bug, a 2022 oracle upgrade froze the V2 cETH market for about a week, and the compound.finance front end was DNS-hijacked in July 2024. Smart contract risk remains.

What is the minimum amount to borrow on Compound?

Each market sets its own floor: the Ethereum USDC market enforces a 100 USDC minimum borrow (baseBorrowMin). Gas is no longer the barrier it was, with Ethereum base fees near 0.1 gwei in August 2026 putting a supply or borrow well under $1. Check live gas before sizing, as fees spike during congestion.

How do COMP rewards work?

COMP goes to base-asset suppliers and borrowers in proportion to interest accrued, but the subsidy is now marginal. On Ethereum the USDC market emits 55 COMP per day, USDT 30 and WETH 20, worth roughly 0.1% APY against outstanding borrows at August 2026 prices; the USDS, WBTC and wstETH markets and every Layer 2 market emit nothing. What COMP buys is a governance vote, not a meaningful discount.

Can I get liquidated on Compound?

Yes, once your borrow exceeds your collateral valued at each asset's liquidation collateral factor (70-88% on the Ethereum USDC market). Compound V3 has no partial liquidation: the absorb function seizes the entire account. The penalty is set per collateral asset, from 7% on WETH to 25% on COMP.

What happens if I can't repay my loan?

Loans have no fixed term. As long as you maintain sufficient collateral, you can keep the position open indefinitely. The only risk is liquidation if the collateral value drops.

Does Compound require KYC?

No. Compound is fully decentralised and permissionless. Only requirement is a compatible Web3 wallet (MetaMask, WalletConnect, etc.).

Which networks does Compound support?

Compound III is deployed on ten networks: Ethereum, Arbitrum, Base, Optimism, Polygon, Linea, Mantle, Scroll, Ronin and Unichain. Ethereum holds around $1.0 billion of the $1.12 billion V3 total. Governance has already executed deprecations of the Scroll, Linea, Mantle, Polygon and Unichain markets, so treat those as exit-only rather than places to lend.

How often do interest rates change?

Compound III accrues interest every second using the block timestamp, and the rate itself moves with utilisation on every supply, borrow or repayment. Check current rates before borrowing.

Can I use Compound from any country?

Compound is permissionless and accessible globally. However, some countries restrict access to DeFi. Check local regulations before using.

What's the difference between Compound V2 and V3?

V3 (Comet) replaced the shared-pool cToken model with isolated markets, each built around one borrowable base asset. Ethereum runs six of them — USDC, USDT, USDS, WETH, WBTC and wstETH — and the USDC market accepts WETH, wstETH, WBTC, cbBTC, tBTC, weETH, LINK, UNI and COMP as collateral. This isolation prevents risk contagion between assets, and V3 also uses less gas per transaction. V3 holds roughly $1.12 billion of TVL against V2's roughly $89 million (DeFiLlama, August 2026). V2 was deprecated on-chain in December 2025: new supply and borrow are paused and every reserve factor was raised to 100%, so V2 is repay-and-withdraw only. V3 is the only version to open a position in.

How do I track my Compound position?

Use Zapper.fi, Zerion, or DeBank to monitor positions across wallets. Compound V3 publishes no health factor of its own, only a borrow capacity and a liquidation point, so these dashboards render your position as a health-factor-style number alongside accrued interest and COMP.

Sources & References

• Compound Official Website
• Compound Documentation
• Compound Governance
• Compound GitHub — Open-source smart contract code and security audit history
• Compound on DeFiLlama — Independent TVL tracking and historical market share data

Related Reviews

• Compound Finance Referral Guide
• Aave Borrowing Review
• Nexo Borrowing Review
• DeFi lending: the complete guide

Affiliate Disclosure

This page contains affiliate links. When you sign up through our referral links, we may earn a commission at no additional cost to you. This helps support our platform and allows us to continue providing valuable content and recommendations.