Bitcoin Exchange Reserves Tracker: BTC Inflows vs. Outflows

 

Exchanges are holding less bitcoin than at almost any point in the past three years, and the price has not rewarded it. CryptoQuant's exchange reserve reading puts total BTC held across exchanges at roughly 2.68 million coins, the lowest on that chart since at least September 2023. Against a circulating supply just above 20 million, that leaves only about 13% of all bitcoin sitting where it can be sold into an order book on short notice.

 

The tape disagreed. According to Yahoo Finance's October 9 price report, BTC opened at $81,690.49, down 1.9% on the day and 3.7% on the week, down 33.8% year over year, having dipped below $81,000 during a fourth consecutive losing session while total crypto market capitalization slid under $2.8 trillion. Multi-year lows in exchange supply alongside a four-day decline is precisely the configuration that exposes how this metric gets misread. It describes where sellable supply sits. It does not describe direction.

 

 

 

Reserves sit at a multi-year low. Roughly 2.68 million BTC remain on exchanges, the least since at least September 2023, which leaves the readily sellable float historically tight.

 

Withdrawals are accelerating. Stocktwits reported that CryptoQuant analyst Darkfost tracked Binance balances falling from 704,800 BTC on September 20 to 663,100 BTC by October 6, more than 40,000 coins or about $3.3 billion in roughly 15 days, including a single week of about 23,100 BTC, the fastest pace since June 2023.

 

Tighter supply has not lifted price. Over the same stretch BTC retreated from the September 21 high of $87,402.34 and touched $80,324.53 on October 8. Thin supply changes elasticity, not whether bids show up.

 

Inflows are not proof of selling. Internal address rotations, derivatives margin transfers, custodian migrations and market-maker rebalancing all appear as inflows on-chain, and Glassnode states in its exchange data transparency notice that large flows should be treated as preliminary.

 

Outflows are not proof of conviction. Withdrawn coins may go to cold storage, but they also go to ETF custodians, OTC counterparties and lending platforms where they can become forced supply later.

 

Cross-checking matters. IG's October 9 market analysis records about $484.9 million of net outflows from US spot bitcoin ETFs on October 7, the largest single day since June, pointing the opposite way from the withdrawal data.

 

 

 

Three numbers frame the supply picture. Total exchange reserves near 2.68 million BTC sit at multi-year lows, Binance alone accounts for roughly a quarter of all exchange-held coins at 663,100 BTC, and that single venue has shed more than 40,000 coins in half a month. On the standard interpretation, this is textbook supply compression.

 

Price has been telling a different story. IG's analysis notes BTC peaked at $87,402.34 on September 21, the highest since January, then fell about 5% and traded as low as $80,324.53 on October 8. The former support band between $82,814.03 and $82,035.16 flipped to resistance, while Coinbase Institutional pegs the ETF cohort's break-even near $81,300. The market is fighting over the average entry price of its newest buyer base rather than trending on a scarcity narrative.

 

 

Supply data cannot be read alone, and the flow side supplies the direct counterexample. Per IG, US spot bitcoin ETFs took in about $2.39 billion in the week to September 25, about $241.1 million in the week to October 2, then turned to roughly $89.9 million of outflows on October 5 and about $484.9 million on October 7, the heaviest single day since June, with IBIT down about $207.7 million and FBTC about $105.1 million. August and September delivered roughly $3.52 billion and $2.65 billion of net inflows, and year-to-date flows only turned positive in September after sitting about $5.8 billion negative in July. The marginal buyer is oscillating, not accumulating steadily. Tracking that series is the job of our bitcoin ETF flow tracker.

 

Macro is not helping either. The same analysis puts the US 10-year Treasury yield near 5.32% on October 8, close to levels last seen in 2002, with Brent crude around $105, after the Federal Reserve raised rates by 25 basis points on September 16. With risk-free yields that high and inflation unresolved, a structural supply story rarely moves price on its own.

 

 

 

CryptoQuant's metric documentation defines exchange reserve as the total quantity of coins held in addresses controlled by an exchange, available for trading, withdrawals and user balance management. Inflow counts coins arriving at those addresses, outflow counts coins leaving, and netflow is the difference. Reserve is cumulative while netflow is a point-in-time measure, so the former is simply the latter accumulated across the period.

 

CryptoQuant's directional reading is straightforward: a rising reserve implies more coins positioned to be sold and therefore greater potential selling pressure, while a falling reserve implies the opposite. Worth noting is the hedged language in that documentation, where readings are framed as what "typically" or "may" happen and the numerical examples are explicitly illustrative rather than real data. This is a probabilistic statement about supply positioning, not a causal mechanism.

 

 

Daily prints are noisy enough that one custodial migration can double a day's netflow. The more reliable practice runs two windows. The 7-day change captures tactical repositioning, such as a burst of event-driven withdrawals or deposits. The 30-day change captures structural drift, such as whether long-horizon capital is steadily moving coins off venue.

 

Scale matters as much as direction. CoinMarketCap Academy, citing CryptoQuant, records a peak daily inflow near 60,000 BTC on February 6, 2026, as price slid toward $60,000, against a seven-day moving average near 23,000 BTC, about 60% below that peak. Those two figures make a usable yardstick for judging whether any given day is a tail event or ordinary churn.

 

 

 

Glassnode's transparency notice is unusually candid about this. Exchange balances come from an address-label database built on exchange-verified addresses including proof-of-reserves disclosures, vetted public tags, and clustering plus machine-learning identification. Because the team works to suppress false positives, published balances are explicitly lower bounds on the true figure. More importantly, exchanges change addresses and move funds internally, and some of those shifts require manual correction, so large inflows and outflows should be treated as preliminary.

 

In practice that means the hundred thousand coins that appeared to flood onto exchanges over a weekend may well have been one venue rotating a cold wallet rather than anyone preparing to sell. Historical balances also get revised when new addresses are identified, which is why two data providers routinely disagree about the same day's reserve level.

 

 

Even with perfect chain data, intent is not legible from a transfer. A deposit may fund a sale, or it may top up futures margin, subscribe to a yield product, or route a cross-chain transfer. A withdrawal may move coins into cold storage, or into a custodian, an OTC counterparty, or a lending platform as collateral, and that last case becomes passive supply precisely when price falls.

 

ETFs complicate the plumbing further. Yahoo Finance reports that as of September 18, US spot bitcoin ETFs managed about $102.532 billion, equal to 6.29% of bitcoin's market capitalization, or roughly one coin in every sixteen. Those coins sit with custodians rather than on exchange books, so creations register as exchange outflows and redemptions as inflows, neither of which says anything reliable about conviction. Address changes at Coinbase and other custodians muddy the series the same way.

 

 

 

February 2026 is the clean positive sample. CryptoQuant recorded the exchange whale ratio reaching 0.64, meaning whales accounted for 64% of deposit volume, the highest since October 2015, while average inflow per transaction climbed to 1.58 BTC, the highest since June 2022. Price slid toward $60,000 over the same window. The signal came from concentration and transaction size rather than raw totals, and on that basis it did lead price.

 

 

The counterexample is on the screen right now. Reserves are at multi-year lows and Binance withdrawals are running at the fastest pace since June 2023, yet BTC fell four sessions straight in October. An earlier instance came in February, when Cointelegraph reported that wallets holding 1,000 to 10,000 BTC added roughly 230,000 coins over 90 days while Glassnode measured gross whale withdrawals averaging 3.5% of exchange-held BTC over 30 days, the strongest since November 2024. The accumulation was real, and price still fell for months, bottoming only in June.

 

The conclusion is consistent either way. Reserve data is good for judging supply elasticity and the scope for amplified moves. It is poor for timing. Tight supply means a given bid pushes price further, and it equally means a given offer digs a deeper hole, because thin books cut both ways.

 

 

Reserves are a slow variable and need fast ones beside them. On the demand side, watch spot ETF net flows, currently the most concentrated channel of marginal buying. On the leverage side, watch funding rates and open interest to size the cascade risk that thin books create. On the liquidity side, watch stablecoin balances on exchanges, where CryptoQuant has tracked daily net USDT inflows falling from a one-year peak near $616 million on November 5, 2025, a dry-powder signal that often says more about buying capacity than BTC reserves do. For cycle context, read it alongside our bitcoin market cycle indicators and the fuller map of what actually drives bitcoin's price.

 

A workable sequence runs like this. Use the 30-day reserve change to establish the supply direction, the 7-day netflow to confirm the move is still underway, ETF flows to check whether marginal demand agrees, and funding rates to see whether leverage has already priced the thesis. Four aligned readings give the highest-quality signal. Contradictory readings, which is the situation today, argue for lowering the weight you give any of them.

 

Reserves tell you where supply is standing. Price is still settled in the order book. Open BTC spot pricing and test the read against the tape

 

 

 

The most common error is treating low reserves as a bullish conclusion. Tight supply raises sensitivity to flows; it does not create them. The second is reacting to a single day, when Glassnode explicitly flags that large movements may need manual verification and internal rotations get misreported as sell pressure every year. The third is ignoring methodology gaps, since label coverage differs by provider and CryptoQuant and Glassnode can diverge by tens of thousands of coins on the same date. Comparing absolute levels across providers is close to meaningless, while tracking the trend within one provider is not.

 

 

If the supply story converts, withdrawals continue, ETF flows return to steady net inflows, and thin books amplify the bid back toward the September 21 high of $87,402.34 and then $90,000.

 

If the market grinds, reserves keep drifting lower while ETF flows stay choppy in the current pattern of large alternating days, and price keeps contesting the roughly $81,300 ETF cost basis without resolving.

 

If supply returns, further weakness triggers collateral liquidations or capitulation from long-term holders, reserves turn higher, and inflows finally do mean selling. The February peak near 60,000 BTC in a day is the yardstick for how extreme that can get, with the September low of $74,919.36 as the reference below.

 

 

Near term, watch whether ETF flows recover from the $484.9 million outflow of October 7, the cleanest read on marginal demand. On chain, watch whether Binance withdrawals hold the pace set since late September, since single-venue shifts tend to lead the aggregate. On macro, watch whether the 10-year yield retreats from 5.32% and what the Fed does on October 27 and 28. On price, the $81,300 ETF cost basis, the $82,035.16 former support now acting as resistance, and the September low at $74,919.36 define the current structure.

 

 

For James Mitchell, the real use of exchange reserve data is calibrating elasticity rather than forecasting direction. With 2.68 million BTC on venue, a $100 million order moves price more than it would in a period of abundant exchange supply. That makes reserves an input to volatility assumptions, not to directional ones, and it argues for treating the metric as a dial on position size and stop width rather than as an entry signal.

 

Two misreadings stand out. The first equates falling reserves with locked-up coins. Withdrawals land in cold storage, but they also land with custodians, OTC counterparties and lending desks, and the last two convert into passive supply exactly when price falls, which is how supposedly departed coins reappear on the book. The second treats the current divergence between record-low reserves and a sliding price as evidence the indicator has broken. It has not. It is saying that supply is tight while demand is absent. The largest single-day ETF outflow since June landing in the same week as the fastest Binance withdrawals since June 2023 means long-term holders and ETF buyers are currently doing opposite things.

 

What matters next is which of those two forces turns first. If ETF flows return to the monthly pace of August and September, in the $2.5 billion to $3.5 billion range, while reserves keep falling, the scarcity thesis finally has the demand it needs. If withdrawals stall while macro pressure persists, low reserves become a liability instead, because thin books offer no protection on the way down either. The 10-year yield near 5.32% deserves equal attention, since at that level bitcoin's competition with other long-duration risk assets explains marginal pricing better than any on-chain supply series.

 

The cross-asset lesson is that bitcoin has become an asset with a transparent supply side and an opaque demand side. Commodity markets have lived with this for decades through inventory data, and commodity traders long ago internalised that low inventories make price more sensitive to demand shocks rather than guaranteeing higher prices. Exchange reserves play the inventory role in crypto, with the differences that they update hourly and invite far more over-reading. The mature application is to pair them with demand-side flow data rather than asking a single supply series to carry the whole judgment.

 

 

 

They are the total amount of bitcoin held in addresses controlled by exchanges, which per CryptoQuant's definition can be used for trading, withdrawals and managing user balances. The series is tracked because only coins sitting on an exchange can be sold into an order book quickly, which makes it a proxy for readily available supply and potential selling pressure. Reserves currently stand near 2.68 million BTC, the lowest since at least September 2023.

 

 

The net direction is off. Aggregate reserves are at multi-year lows, and single-venue data is sharper: CryptoQuant analyst Darkfost tracked Binance balances falling from 704,800 BTC on September 20 to 663,100 BTC by October 6, more than 40,000 coins in about 15 days, with one week alone near 23,100 BTC, the fastest since June 2023. Price nevertheless fell for four consecutive sessions, so the withdrawal trend has not yet translated into upward momentum.

 

 

No. Falling reserves reduce the supply available for immediate sale, but they do not generate demand. The present market is the counterexample, with reserves at multi-year lows while BTC slipped under $81,000 on October 9 and sat 33.8% lower year over year. What low reserves really imply is a thinner order book and greater sensitivity to flows, an amplifier that works in both directions.

 

 

Not necessarily. Internal address rotations, derivatives margin transfers, ETF redemptions and market-maker rebalancing all register as inflows. Glassnode's own guidance is that large inflows and outflows should be treated as preliminary, since some internal movements require manual correction to identify. A better test of genuine pressure is deposit concentration and average transaction size, which is why February 2026 mattered, with the exchange whale ratio at 0.64 and average inflows at 1.58 BTC preceding further weakness.

 

 

Both, for different purposes. The 7-day change shows tactical repositioning and tells you whether a run of withdrawals or deposits is still live. The 30-day change shows the structural drift of long-horizon capital. Daily prints carry the most noise, since one custodial migration can distort them, so unless you see a tail event on the scale of the roughly 60,000 BTC inflow of February 6, 2026, a single day is a poor basis for a decision.

 

 

CryptoQuant and Glassnode are the main sources, both publishing aggregate and per-exchange reserves, inflows, outflows and netflow. Mind the methodology gap: label coverage differs, absolute levels for the same day can diverge by tens of thousands of coins, and Glassnode states outright that its published balances are lower bounds on true balances. Comparing levels across providers adds little, while following the trend within one provider is sound. Current pricing is on the BTC price page.

 

 

They represent supply and demand respectively. Falling reserves say sellable supply is shrinking; ETF net inflows say marginal buying is arriving. The signal is strongest when they agree. At the moment they do not, with reserves falling while ETFs posted $484.9 million of net outflows on October 7, the largest single day since June, a divergence that typically favours range trading over trend. Pair them with funding rates and open interest and bitcoin dominance to see how leverage and rotation are positioned.

 

 

Treat it as a dial on position size and stop width rather than as a trade signal. When reserves are low the book is thinner, so the same position absorbs more slippage and sharper swings, which argues for wider stops or smaller size. The decision itself should still rest on flows, macro conditions and personal risk tolerance. For the mechanics of buying and allocating, see the guide to buying bitcoin, the BTC purchase walkthrough and the current BTC campaign page.

 

 

The information above is provided for general market information and analysis only and does not constitute investment advice, financial advice, legal advice, tax advice or a recommendation to trade. Prices of crypto assets, equities and other related financial assets can fluctuate sharply, and past performance, technical indicators and on-chain data do not guarantee future results. The exchange reserve figures, flow data, price levels and research conclusions cited here reflect publicly available information at the time of publication, methodologies differ between data providers, and the underlying values may be revised or updated at any time, so the latest disclosures from the relevant data platforms and institutions should be treated as authoritative. Readers should conduct their own research and make decisions based on their own financial circumstances, investment objectives and risk tolerance, consulting a qualified professional where appropriate. The MEXC Crypto Pulse team accepts no liability for any direct or indirect loss arising from the use of this information.

 

 

James Mitchell specializes in technical analysis, market trends, and trading strategies for both Bitcoin and altcoins. Based in London, he has over 10 years of experience in financial markets. Before joining MEXC Learn, James worked as a senior analyst at a leading European investment firm, where he developed expertise in risk management and quantitative trading. His transition to cryptocurrency markets began in 2017, and he has since become recognized for his data-driven approach. He holds a Master's degree in Financial Economics from the London School of Economics. His analytical approach combines traditional technical analysis with on-chain metrics to provide readers with actionable insights.

 

Areas of Expertise: Technical Analysis, Market Trends and Cycles, Trading Strategies, Bitcoin and Altcoin Analysis, Risk Management.

 

 

 

Want the fastest access to MEXC's latest updates? Join our official Telegram group now!

MEXC Platform: Download App | Zero Fee Trading | Buy USDT Now

Join MEXC Community: X (Twitter) | Telegram | Discord

Get Insights: MEXC Learn | Terms of Service | MEXC Blog | MEXC News

Account Verification: Understand KYC | How to Complete KYC

External Content Platforms: Substack | Medium | Paragraph | LinkedIn | X(News)