Exchange Total Reserves (Proof of Reserves)

What is it?

Exchange Total Reserves measures the aggregate value of the crypto assets that can be observed on-chain for a set of centralized exchanges (CEX). Think of it as a "bank account balance" for these platforms - except instead of relying on self-reported numbers, most of the data comes directly from the blockchain, where every balance is verifiable. After the FTX collapse in November 2022, Proof of Reserves became the reference for exchange transparency. Several exchanges began publishing their wallet addresses so that anyone could independently verify the assets they hold. Trinity covers venues through three routes: (1) a book of known addresses restricted to venues that publish their own addresses - OKX, KuCoin, Gate.io, Bitget, Crypto.com, Deribit, Phemex, Kraken and MEXC - whose balances are read directly on the Bitcoin and Tron blockchains; (2) the official wallet list that Bitfinex publishes, scanned the same way; (3) the public proof-of-reserves data published by HTX. A venue whose measured reserves fall below $10 million on a given day is left out of that day's total. Only two chains are scanned: Bitcoin (BTC) and Tron (USDT). Assets held on Ethereum are not counted, and some of the largest venues are not covered yet: Binance and Bybit are being recalculated on Trinity's own Bitcoin node from the addresses they publish themselves, and are shown as "being recalculated" until then, never as zero. The total is therefore a floor for the covered set, not a measure of the whole industry. Asset prices come from Coinalyze spot markets, completed by CoinGecko; stablecoins are valued at par. This chart shows three overlaid metrics: (1) Total Reserves - the sum of the counted assets in USD, (2) Clean Reserves - total minus exchange native tokens, which can inflate balance sheets, and (3) Stablecoin Reserves - the "dry powder" sitting on exchanges ready to buy. Because the address scans only count BTC and USDT, native tokens enter the total only through proof-of-reserves data that declares them (the HTX token, for instance); the gap between Total and Clean reserves measures that share.

How to read

This chart has two visual panels. **Panel A (primary) - Horizontal bar ranking by exchange**: each covered venue (OKX, Bitfinex, Kraken, HTX, Gate.io, etc.) gets a horizontal bar color-coded gold-to-teal, length proportional to its current USD reserves. KPI cards at the top show aggregated snapshot values: Total Reserves, Clean Reserves (excluding exchange native tokens) and Stablecoin Reserves. The ranking view is designed to be immediately readable with even 1 day of data - it shows the current state of the covered set. **Panel B (secondary, below) - time-series at reduced height (280px)**: shows the evolution of total/clean/stablecoin reserves over time. This panel becomes informative as more history accumulates. Focus on: (1) **Ranking order** - who holds the most within the covered set. A healthy industry distributes across many platforms. (2) **KPI cards** - the Total vs Clean gap shows how much of the counted reserves rests on exchange native tokens (large gap = warning sign). (3) **Time-series trend** (secondary panel) - rising/falling aggregate reserves. Sharp drops >5% in 24h on the time-series warrant attention, after checking that the number of covered venues has not changed that day. The BTC price overlay (optional) helps correlate reserve changes with market movements. After the FTX collapse, for example, reserves dropped dramatically across all exchanges as users withdrew to self-custody, then slowly recovered as trust rebuilt. **Data scope note**: the current method (Bitcoin and Tron address scans, HTX proof of reserves, Bitfinex wallet list) starts on September 23, 2026, and this chart's visible timeframe starts from that date. Earlier data was produced by a retired method and is not shown, because the two series are not comparable. Historical PoR events from the 2022-2024 era (FTX collapse, first wallet publications, ETF launch) are documented in the Global Liquidity events table and display as markers on macro charts with deeper coverage (M2, US Net Liquidity, Global Net Liquidity). On this specific chart, vertical markers will display events falling within the visible window. Toggle marker visibility via the 'Events' toolbar button.

Key zones

**November 2022 - FTX Collapse**: The most dramatic event in PoR history. Exchange reserves dropped by an estimated $20-30B in weeks as users panicked and withdrew en masse. This event single-handedly created the Proof of Reserves movement and fundamentally changed how the industry approaches transparency. **January 2023 - Post-FTX Low**: Reserves hit their lowest point as the full contagion from FTX spread to Genesis, BlockFi, and other entities. The surviving exchanges had to prove their solvency, leading to a new era of on-chain transparency. **Q3 2023 - Recovery Phase**: As surviving exchanges rebuilt trust and the market stabilized, reserves began climbing back. BlackRock's ETF application in June 2023 boosted confidence, driving new deposits. **January 2024 - ETF Launch**: The approval of spot Bitcoin ETFs created a new dynamic - while some BTC moved from exchanges to ETF custodians, fresh capital also flowed in, and stablecoin reserves surged as buying power increased. These episodes predate the visible window of this chart: they frame how to read its moves, and they appear as markers on macro charts with deeper history.

What to observe

**Pattern 1 - Reserve Drawdowns**: A sustained decline in total reserves over 2+ weeks often precedes or accompanies bearish price action. Users withdraw to self-custody when they lose confidence, reducing exchange liquidity and selling pressure potential. **Pattern 2 - Stablecoin Accumulation**: When stablecoin reserves rise while BTC reserves remain flat or decline, it indicates capital moving to the sidelines - waiting to buy. This "dry powder" buildup has historically preceded significant rallies. **Pattern 3 - Exchange Concentration**: If one exchange's share of total reserves grows disproportionately, it creates systemic risk. Pre-FTX, concentration was dangerously high. Monitor the top exchange chart for healthy distribution - keeping in mind that it measures concentration within the covered venues only. **Pattern 4 - Clean vs Total Divergence**: If the gap between total and clean reserves widens, the counted reserves rest increasingly on exchange native tokens - a warning sign similar to what happened with FTX and FTT. **Pattern 5 - Coverage changes**: A step in the series on a single day can come from a venue entering or leaving the covered set (for example crossing the $10 million threshold) rather than from real flows. Check the venue count before reading a move.

Historical context

The concept of Proof of Reserves emerged from one of crypto's darkest chapters. Before November 2022, exchange transparency was essentially trust-based - users deposited funds and hoped the exchange was solvent. The collapse of FTX revealed that Sam Bankman-Fried had been misappropriating billions in customer funds, creating a massive hole between what users thought they owned and what actually existed. The aftermath was swift: major exchanges published their wallet addresses within days, and others followed. Third-party auditors like Mazars and Armanino were engaged, though some later withdrew due to the complexity of crypto audits. The industry settled on a pragmatic approach: publish known wallet addresses and let anyone track them continuously. Trinity's method follows that logic, with a deliberately narrow scope: it reads balances on-chain for venues that publish their own addresses, uses the official wallet list published by Bitfinex and the public proof-of-reserves data of HTX, and counts only Bitcoin and USDT on Tron. This conservative approach means actual reserves are higher than reported - assets on Ethereum and other chains, and assets in undisclosed cold wallets, are not counted. This data represents a fundamental shift in financial transparency. Traditional banks don't publish real-time balance sheets. Crypto exchanges that publish their addresses effectively do, making this one of the most transparent sectors in all of finance.

Expert notes

**Methodological caveat**: Trinity counts only Bitcoin and USDT on Tron held at known addresses of venues that publish their own addresses, plus the official Bitfinex wallet list and the HTX proof-of-reserves data. Assets on Ethereum are not counted, some large venues are not covered, and exchanges also hold assets in undisclosed cold wallets. The data represents a *floor*, not a ceiling. **Liabilities gap**: PoR shows assets but not liabilities (what exchanges owe to users). Without liabilities, a 1:1 reserve ratio cannot be confirmed. This is the key limitation - an exchange could show $10 billion in reserves while owing $15 billion. **Native token risk**: Exchange tokens (FTT being the textbook case) are the most dangerous reserves because they're reflexive - if the exchange has problems, the token crashes, which makes the reserves problem worse. FTX's reserves were >50% FTT, which became worthless. Always compare clean vs total reserves. **Cross-exchange flows**: When reserves drop on one exchange and rise on another, it's usually migration, not net withdrawal. Look at aggregate trends, not individual exchange movements in isolation - and remember that flows toward venues outside the covered set appear as a drop.

Common mistakes to avoid

**Mistake 1 - "Higher reserves = safer exchange"**: Not necessarily. Reserves show assets, not the ratio of assets to liabilities. An exchange with $50 billion in reserves but $60 billion in liabilities is insolvent despite large reserves. Always consider the context. **Mistake 2 - "Declining reserves mean an exchange is in trouble"**: Sometimes users simply migrate to competitors, to DeFi, or to self-custody for personal security reasons. Look at aggregate industry reserves, not just one exchange. **Mistake 3 - "Total reserves = user deposits"**: Total reserves include exchange treasury, operational wallets, insurance funds, and sometimes staking rewards. Clean reserves are closer to user deposits, but even they include non-deposit assets. **Mistake 4 - "PoR proves solvency"**: Proof of Reserves proves asset holdings at a point in time. It does NOT prove solvency (assets >= liabilities), does NOT prove those assets aren't borrowed, and does NOT prevent an exchange from moving assets immediately after a snapshot. **Mistake 5 - "This total is the whole exchange industry"**: It covers a defined set of venues, on Bitcoin and Tron only. It is a floor for that set, not an industry-wide figure.

Programmatic access

REST API

curl -sS \
  'https://api.trinityinsights.io/api/v1/macro-intelligence/metrics/exchange-total-reserves/data?timeframe=90d' \
  -H 'X-API-Key: $TRINITY_API_KEY'

MCP server

{
  "tool": "get_chart_value",
  "metric_id": "exchange-total-reserves",
  "timeframe": "1y"
}

Required tier: pro. See the pricing grid for the tier list and the MCP documentation for multi-client configuration.

Related metrics

• exchange-reserves-top
• exchange-stablecoin-reserves
• us-net-liquidity

Institutional disclaimer

Trinity Insights is an educational and analytical tool. The metric above does not constitute investment advice. Trinity Insights is not a Crypto-Asset Service Provider (CASP) registered under MiCA Regulation (EU) 2023/1114. See the full disclaimer.