In Q2 2026, the total value of proof-of-reserves (PoR) disclosures from crypto exchanges stood at roughly $192.6 billion, a 5.3% drop from the previous month, according to CoinMarketCap. But the metric, which shows on-chain assets backing customer balances, has well-known shortcomings that make it a partial picture rather than a full solvency audit.
What Proof of Reserves Shows — and Doesn't
Proof of Reserves is a disclosure that lists on-chain assets held by an exchange to back customer balances at a specific point in time. It often uses a Merkle tree to aggregate liabilities. But the snapshot has limits: it doesn't capture off-chain liabilities like fiat or loans, can't prove exclusive control of keys, and can be gamed by moving funds just before the check. A July 2026 primer from The Block laid out these gaps, calling PoR a point-in-time snapshot that is necessary but not sufficient for a full solvency audit.
The Q2 2026 Snapshot
CoinMarketCap's data for the quarter showed USDT as the largest reserve asset at roughly $57.6 billion, followed by BTC at $55.5 billion. The month-over-month decline of 5.3% came as some exchanges adjusted their disclosures or saw outflows. The numbers reflect only what exchanges chose to report on-chain, leaving off-chain positions invisible to users.
A Playbook for Evaluating PoR
Industry observers have developed a checklist for assessing PoR disclosures. It includes checking the frequency and scope of reports, verifying the snapshot date, analyzing per-asset reserve ratios, confirming on-chain addresses and signed messages, testing Merkle proof inclusion, scanning for timing games, and cross-checking with independent data. The idea is to combine PoR with other checks and risk limits, since no single snapshot can guarantee solvency.
Lessons from the Field
In an editor's note, the author of the Q2 2026 analysis said they started tagging venue risk and observed exchanges with strong per-asset ratios but increasing stablecoin weights. The biggest win, they noted, was catching timing quirks around snapshots — a reminder that even well-structured PoR can be misleading if the timing is off. The author didn't name specific exchanges, but the pattern suggests that stablecoin-heavy reserves may mask risk if those tokens aren't fully backed or easily liquidated.
The challenge for users remains: PoR is a necessary signal, but it's not a full audit. Without combining it with other checks and risk limits, the snapshot can be stale or gamed. As the industry matures, the question of how to close the gap between what PoR shows and what it hides will only grow more urgent.
