Fees, TVL and Volume Are Three Different Businesses. Here Is How to Read Them Together

The question

Three numbers get quoted whenever anyone describes a chain: total value locked, daily decentralised exchange volume, and fees. They are usually presented as a set, as though they measure the same underlying thing at different resolutions. They do not. TVL measures a stock of capital, volume measures a flow across that capital, and fees measure what the chain itself charges for processing the flow. Each can rise while another falls, and the combinations tell you different things about what a chain is actually for.

What each number counts

TVL is the dollar value of assets sitting in contracts on the chain: liquidity pool balances, lending collateral, staked positions. It is a balance sheet snapshot, marked at current prices. Because it is marked, it moves when asset prices move even if not one unit of anything is deposited or withdrawn. A chain whose TVL is mostly volatile tokens will show TVL swings that reflect nothing about user behaviour. A chain whose TVL is mostly stablecoins will show a much flatter line, because the denominator and the numerator are the same currency.

DEX volume is the notional value of swaps executed in a period. It is a flow, and flows are not bounded by stocks. The same dollar of liquidity can be traded against many times in a day. On Robinhood Chain, DefiLlama puts TVL at $923.6m and 24 hour DEX volume at $1.38bn, which is roughly one and a half times the entire capital base turning over in a day. That ratio is not an anomaly, it is what concentrated liquidity venues are built to do, and it is the single most useful derived statistic on the page.

Why fees are their own animal

Chain fees are what users pay the network to include and execute transactions. They are a function of transaction count, transaction complexity, and whatever pricing rule the chain uses, which on most rollups means a base fee that responds to congestion plus a component covering the cost of posting data to the settlement layer. They are not a percentage of volume. A chain can process a billion dollars of swaps in a few thousand large transactions and collect very little, or process a fraction of that in hundreds of thousands of small ones and collect a great deal.

This is why the fee line and the volume line can point in opposite directions for weeks. Robinhood Chain shows $11.4m of fees in 24 hours against $284.2m over 30 days, which averages to roughly $9.5m a day, so the most recent day sits above the monthly run rate even as DEX volume is down 8.32% on the day and 8.65% on the week, per DefiLlama. That combination, falling notional with holding fees, usually means the mix of activity shifted toward more transactions of smaller size, not that anything broke.

The four combinations worth recognising

TVL up, volume up: capital arriving and being used. The healthy case, and the rarest. TVL up, volume flat: capital parked, usually chasing an incentive or waiting for something, contributing depth but not turnover. TVL flat, volume up: the existing capital base working harder, which is what a market-making-heavy chain looks like and what the Robinhood Chain ratio implies. TVL down, volume up: liquidity leaving while the remaining pools absorb more flow, which raises slippage and tends to be a leading indicator that the volume follows the liquidity out.

The contrast that makes the point

Ink, on the same DefiLlama data, carries $163.6m of TVL against $2.2m of daily DEX volume. That is a turnover ratio of roughly one to seventy four in the other direction: the capital sits and very little of it trades on any given day. Neither profile is better in the abstract. One describes a venue where the dominant activity is trading, the other describes a chain where the dominant activity is holding, lending or waiting. But a strategy that depends on exiting a large position quickly cares enormously about which one it is standing on, and TVL alone will not tell you.

What TVL does not capture

Three things, all material. First, it excludes assets held in plain wallets, so a chain where users self-custody rather than deposit into protocols will look thinner than it is. Second, double counting: a token deposited into a lending market, borrowed against, and redeposited can appear more than once depending on how the aggregator handles it. Third, and most important for anyone sizing an exit, TVL says nothing about where the liquidity sits. A pool with concentrated liquidity clustered tightly around the current price shows the same TVL as one spread across a wide range, but the two behave completely differently when a large order arrives.

Stablecoin supply is the partial corrective. Robinhood Chain's $1.04bn of stablecoins against $923.6m of TVL means the settlement currency on the chain exceeds the total marked value locked in its protocols, which tells you the float is not an artefact of token prices. Ink's $171.5m of stablecoins against $163.6m of TVL sits in a similar relationship. In both cases the stable float is the closest thing to a hard floor under the headline figure.

What to watch

Watch the ratios, not the levels. Volume divided by TVL tells you how hard the capital is working. Fees divided by volume tells you what a dollar of notional costs to move on that chain, and a sharp move in that ratio is usually a change in transaction mix or in the cost of data availability rather than a change in demand. Stablecoin supply divided by TVL tells you how much of the balance sheet is price-insensitive. And watch venue concentration: on Robinhood Chain the top two venues, Uniswap V3 at $607.5m and Uniswap V4 at $565.8m, account for the overwhelming majority of the $1.38bn day, which means the chain's volume statistic is in practice a statistic about two contracts.

Where it breaks

All of these figures are constructed by aggregators from onchain data using judgement calls about what counts. Wash trading inflates volume without inflating anything real. Incentive programmes inflate TVL for exactly as long as the incentives run. Fee figures can include or exclude the cost the chain itself pays to a settlement layer, which is the difference between revenue and profit and is frequently elided. Peak figures are especially treacherous: Ink's all-time TVL high of $572.8m in January 2026 against $163.6m today is a drawdown of more than two thirds, and whether that represents capital flight or the mechanical unwind of an incentive schedule is not visible in the number itself. The figures are a starting point for a question, not the answer to one.