Total value locked, or TVL, is meant to describe the assets deposited in a financial protocol. In Bitcoin finance, it is often read as a measure of backing or economic security. Those interpretations deserve separate questions. A coin can exist without being available to meet a withdrawal, and an asset can back a token without being committed to staking.
What interests me as a researcher is the statement behind the number. Before designing a proof, I want to know exactly what a verifier is supposed to learn.
| Claim | What needs to be established |
|---|---|
| Reserves | The identified assets exist, and the claimed party controls them under the relevant spending conditions. |
| Solvency | Those assets are sufficient to cover the liabilities included in the statement. |
| Staking | The coins satisfy the protocol’s actual commitment, withdrawal, and penalty rules. |
| Aggregate TVL | Receipts and the assets behind them are counted according to an explicit rule, without silently treating both as new capital. |
There is a substantial literature behind the first two claims. Dagher et al.’s Provisions (2015) shows how a Bitcoin exchange can prove that its assets cover customer liabilities while preserving the privacy of individual balances and addresses. The liabilities are essential: a large wallet balance, by itself, cannot establish solvency. A solvency proof also does not establish that the same assets are locked into a particular staking protocol.
Aggregation introduces a different problem. Suppose bitcoin backs a receipt token, and that token is then deposited into another application. Counting both layers can make one pool of capital appear several times. Luo et al. (2024) study this issue in Piercing the Veil of TVL and propose total value redeemable, or TVR, to distinguish the underlying assets from the chain of claims built on them.
Bitcoin adds an unusually explicit object to inspect: an unspent transaction output, or UTXO. Each output has an amount and spending conditions. A signed transaction involving it is evidence of a particular authorization. Whether that transaction was confirmed, whether a protocol can enforce penalties, and whether an owner can still spend through another path require further checks.
The Nubit–Nebra technical report Proof of TVL: Regaining Trust in BTCFi (2025) applies this perspective to Bitcoin staking. It describes tracing deposits into staking transactions, inspecting their locking conditions, and comparing the backing with the supply of a liquid staking token, a receipt that represents a claim on staked assets. It also outlines a zero-knowledge extension, rather than presenting that extension as an already completed formal proof.
For me, the useful connection between these works is their treatment of what must be counted and what must be proved. Provisions includes liabilities. TVR examines overlapping economic claims. Proof of TVL asks whether the Bitcoin behind a staking claim is committed in the required way. One check cannot stand in for all three.
These questions also shaped my public criticism of SolvBTC’s reported Bitcoin totals. The discussion included evidence that needed correcting: a questioned asset movement was identified as routine SolvBTC.CORE staking, and unstaking could explain some changing balances. That correction belongs alongside the criticism. It does not remove the need for an explicit accounting rule.
A verifiable Bitcoin TVL figure should therefore identify its assets, the conditions under which they qualify, the liabilities or receipts being compared, and the block at which the claim holds. The result should be a statement another researcher can reproduce and examine, with its assumptions visible.