Total value locked, trading volume, and fees often appear together on DeFi dashboards. They can look like three versions of the same success measure, but they describe different things. A balance indicates what is present under a counting rule; volume describes activity over time; fees describe charges generated under a product's economics.
This guide separates those concepts and demonstrates how plausible-looking comparisons can go wrong. It uses fictional protocols and simplified arithmetic, not current market data. The goal is to make a DeFi meter explain its inputs before it invites readers to interpret a trend or compare one application with another.
Separate a balance from a flow
Think of a balance as a photograph at an observation point and a flow as a record covering an interval. If a fictional contract contains assets valued at $2 million at noon, that is a balance observation. If its exchange activity totals $500,000 between midnight and noon, that is a flow observation. The two numbers need different time labels.
A useful starting reference is DefiLlama's data definitions, which distinguish TVL, inflows, volume, fees, and revenue and explain important inclusion policies. Those definitions are provider-specific. When using another source, inspect its methodology rather than assuming identically named fields use identical rules.
Do not compare a balance change with a period total as though both describe fresh deposits. A protocol can have large activity while the balance remains stable, because the same capital can participate in multiple transactions. That is not a contradiction in the numbers.
Separate price effects from deposit changes
For a simple one-asset illustration, value equals quantity multiplied by price. A fictional pool holding 100 units at $50 has a value of $5,000. At $60 with the same quantity, it has a value of $6,000. The 20% increase comes entirely from the chosen price change.
Now imagine the pool instead holds 110 units at $60. Its ending value is $6,600. Relative to the initial $5,000, value increased by $1,600. Under one explicitly chosen decomposition, repricing the original 100 units accounts for $1,000, while the ten additional units valued at the ending price account for $600.
This is an illustrative decomposition, not a universal implementation of a provider's inflow metric. Timing matters when deposits and prices change throughout a period. Record the valuation convention used by the source before comparing an inflow estimate with your own calculation.
Read volume with its instrument and interval
Volume requires an activity definition. Does a field cover spot exchanges, derivatives notional, transfers, or another event type? Keep these categories separate unless the provider supplies a deliberate aggregation method and your research question calls for that method.
Also record whether the data covers one pool, a protocol deployment, every deployment of a protocol, or a broader market. A small pool's volume and a protocol-wide volume can both be valid observations, but the larger number does not prove better execution at the smaller pool's trading pair.
For a fictional exchange with $1 million of daily volume and a $2 million balance, the ratio of daily volume to the balance is 0.5. This is a daily turnover-style ratio under the chosen definitions. It is not a 50% yield, a deposit growth rate, or a probability of profit.
Distinguish fees from retained revenue
A fee total describes charges under the provider's definition. A revenue measure may describe only the portion attributed to the protocol or another recipient. Read the allocation rule before treating all fees as income belonging to token holders, liquidity providers, or the application operator.
Trace the fee allocation
Consider an entirely fictional system that generates $10,000 in fees during one day. Assume its specified allocation sends $8,000 to liquidity providers and $2,000 to a protocol account. Under that simplified setup, the total fee figure and the protocol's retained amount are different. Neither figure automatically describes a distribution to a token holder.
Do not infer economic rights from a ticker. If a token's relationship to fees is part of the research question, it needs its own documented mechanism. A chart of application usage cannot establish that relationship on its own.
Treat annualization as a scenario
Multiplying one day's fees by 365 produces an annualized rate under a strong assumption: that the same daily result repeats. For the fictional $10,000 day, that calculation gives $3.65 million. It does not establish actual trailing-year revenue or forecast what the following year will produce.
Use labels that preserve the distinction. “Illustrative annualization of one day's fees” is clearer than “annual revenue.” If you have a complete historical year, calculate and label that actual period separately. Do not move between a rolling window and a calendar period without explaining the change.
The APR and APY guide applies the same discipline to yield calculations. A percentage becomes more useful, not less useful, when the compounding and persistence assumptions are visible.
Keep the inclusion policy attached
Suppose a fictional application issues a receipt against deposited assets and another application accepts that receipt. Both applications may report balances relevant to their operation. Summing those balances as independent underlying capital can misrepresent their relationship.
Decide whether your research describes application-level exposure, non-overlapping capital, or some other quantity. There is no benefit in pretending these are the same measurement. A broad aggregate needs an explicit rule about related claims, borrowed assets, staking categories, and network coverage.
When the provider offers filters, save their settings. Two screenshots from the same site can disagree because different categories are enabled. Before explaining a difference as capital movement, check whether the definition changed between observations.
Read a mixed result without forcing a verdict
Imagine fictional Protocol A with a 15% increase in dollar TVL, unchanged token quantities, and a 10% decline in matching-period fees. Protocol B has a 5% decline in dollar TVL, increased token quantities, and unchanged fees. These profiles cannot be reduced to a meaningful winner merely by counting positive signs.
For Protocol A, valuation effects deserve attention. For Protocol B, the relation between quantity growth and price movement deserves attention. The fee observations add another dimension, but they do not settle questions about asset quality, costs, control permissions, or the relevance of the product to a particular user.
A useful DeFi meter describes the conflict. The core research framework keeps liquidity, activity, context, and risk as separate readings so that a strong-looking headline does not erase a weak or unknown input.
Write a measurement-first summary
Use one sentence for the measured result and another for its limits. For example: “Reported dollar TVL increased over matching observations while the available unit balance was unchanged. This suggests a valuation contribution in the simplified comparison; intra-period flows and additional assets were not assessed.”
This format is deliberately narrower than a market prediction. It lets another reader reproduce the arithmetic and identify the missing evidence. For a repeatable workflow, continue with the daily DeFi research routine.
Which denominator belongs in a ratio?
Use the denominator that answers the stated question, and label it explicitly. Period volume divided by an ending balance differs from period volume divided by an average balance. In an illustrative series with a rapidly changing inventory, those choices can produce noticeably different values without either calculation containing an arithmetic error.
Do not compare a ratio based on an average denominator with one based on a single snapshot unless you explain the mismatch. Save the underlying observations as well as the ratio. This allows a reviewer to reconstruct the result rather than accept an unlabeled efficiency score.
Conclusion: define first, calculate second
TVL, volume, fees, and revenue become useful together only when their differences remain intact. Name the scope, distinguish balances from flows, separate price effects, and show allocation and inclusion rules. When the numbers disagree, explain the disagreement. A clear measurement note is more informative than a confident label built from incompatible inputs.
