TVL — Total Value Locked — is the total dollar value of all assets currently deposited in a DeFi protocol’s smart contracts. For a lending protocol like Aave, TVL is the value of all the collateral users have deposited plus any assets in the protocol’s reserves. For a DEX like Uniswap, TVL is the value of tokens in the liquidity pools. For a staking protocol like Lido, TVL is the value of ETH that has been staked through it. The number is reported in USD terms and is tracked by DefiLlama, the dominant DeFi analytics site, which publishes TVL rankings for every major protocol across every major chain.
TVL became the canonical metric of DeFi success during the 2020-2021 DeFi summer, and it has stuck around as the default way to rank protocols despite having meaningful limitations. The underlying appeal is simple: TVL is concrete, measurable, and hard to fake at the protocol level (the deposits are visible on-chain and can be independently verified). A protocol with $10 billion in TVL is obviously bigger and more trusted than a protocol with $10 million, and the number is a useful first filter for separating the established from the experimental.
What TVL Does and Doesn’t Tell You
TVL is useful for size comparisons but less useful as an indicator of quality or profitability.
Size comparisons work. If you are trying to figure out which lending protocols are the biggest on Ethereum, TVL rankings will give you a reasonable first cut. Aave, Compound, Morpho, Spark, and the other top-ten protocols all have TVLs in the billions and are meaningfully larger than the smaller protocols below them. The ordering is roughly what you would expect from looking at trading volume, user counts, and other independent measures of activity.
TVL is not revenue. A DeFi protocol can have enormous TVL and generate almost no fees if the capital is sitting idle. Staking protocols in particular have high TVL because users deposit to earn yield, but the revenue generated per dollar of TVL for the protocol itself can be tiny. TVL growth without accompanying revenue growth is a warning sign that the capital is being parked rather than used productively.
TVL can be gamed. The most common gaming technique is offering high token-based rewards for deposits, which attracts “mercenary capital” that moves between protocols chasing the highest yields. A protocol can double its TVL overnight by launching a liquidity mining program, and it can lose all of that TVL just as quickly when the rewards run out or a competitor offers more. DeFi summer in 2020 saw exactly this pattern repeatedly, and many of the protocols that had briefly explosive TVL rankings have since faded back to near-zero.
TVL double-counts. If a protocol accepts another protocol’s deposit tokens as collateral, the same capital shows up in both protocols’ TVL figures. A user who deposits ETH into Lido gets stETH, which they can then deposit into Aave as collateral. The ETH is counted once in Lido’s TVL and again (in stETH form) in Aave’s. The total DeFi TVL number published by DefiLlama and similar sites includes some amount of this double-counting, and the “real” unique TVL is smaller than the sum of protocol TVLs suggests.
The Yearn Critique
Andre Cronje, the founder of Yearn, wrote a widely-circulated post in 2020 arguing that TVL was the wrong metric for evaluating DeFi protocols and that more useful alternatives — fees generated, profit margin, user counts — were ignored because they were harder to game for marketing purposes. His specific complaint was that a protocol with $1 billion in TVL generating $100k in annual fees was clearly less valuable than a protocol with $100 million in TVL generating $10 million in annual fees, and ranking them by TVL gave the wrong answer.
This critique has aged well. DefiLlama now publishes fee and revenue metrics alongside TVL, and the more sophisticated DeFi coverage has shifted toward evaluating protocols on multiple dimensions rather than on TVL alone. Raw TVL is still the headline number in most discussions, but anyone actually making decisions about where to deploy capital or which protocols to study has to go beyond TVL to understand what is really happening.
The Context of Cycles
TVL rises and falls with crypto prices and with DeFi interest more broadly. At its peak in late 2021, total DeFi TVL across all chains exceeded $180 billion. By the 2022 bottom, it had dropped below $40 billion — a roughly 75 percent decline, reflecting both falling asset prices and capital withdrawing from DeFi entirely. The 2024-2025 recovery brought TVL back above $100 billion, and as of early 2026 the figure is in the $130-160 billion range across all chains, with Ethereum still dominant but Base, Arbitrum, Solana, and a few others having meaningful shares.
These swings make TVL a noisy metric for comparing protocols across different times. A protocol with $5 billion TVL in 2021 and $1 billion in 2023 did not necessarily get worse; it might just be reflecting the broader market. What matters more for evaluating a protocol’s trajectory is its market share of TVL within its category — is it gaining or losing ground relative to competitors? — and its revenue, user counts, and fee generation independently of the headline TVL number. The useful framing is that TVL is a starting point for research, not an endpoint.