ETF withdrawals, corporate Bitcoin sales and exchange trading data show multiple sources of pressure, while allegations that Binance sold assets to cover hack losses remain unverified.
MARKET INSIDER — Claims that Binance triggered Bitcoin’s latest decline by selling assets to cover losses from its October security breach have circulated as cryptocurrency markets retreat. Available evidence does not establish that allegation. Exchange trading data show stronger seller-initiated activity on Coinbase than on Binance during a recent six-day period, while U.S. Bitcoin ETFs recorded sustained withdrawals and miner MARA disclosed substantial sales. These figures cover different periods and measure different activities, but together demonstrate why attributing the entire sell-off to one exchange is premature. They also leave a crucial question unanswered: who actually placed the sell orders?
Key Highlights
- Reported spot trading data show Coinbase’s negative cumulative volume delta was nearly three times Binance’s during November 9–14.
- U.S. spot Bitcoin ETFs lost $1.91 billion across seven consecutive trading sessions through November 13.
- MARA disclosed selling 13,597 BTC after September, while Binance co-founder Changpeng Zhao denied that the exchange had sold cryptocurrency.
Exchange activity does not identify the seller
Data attributed to Anchorage Digital and reported by Gate News showed cumulative volume delta, or CVD, of approximately negative 8,900 BTC on Coinbase between November 9 and 14, measured in UTC.
The comparable readings were negative 3,100 BTC on Binance and negative 700 BTC on Bybit. OKX recorded positive CVD of roughly 1,400 BTC. Coinbase therefore had the largest seller-initiated imbalance among the four exchanges in that comparison. (gate.com)
CVD tracks the cumulative difference between buyer-initiated and seller-initiated trading volume. A negative reading indicates that traders selling into available bids were more active than traders buying at available offers.
Every completed trade still has both a buyer and a seller. The measure describes which side initiated the transaction; it does not identify the beneficial owner behind an order.
Consequently, negative CVD on Binance does not demonstrate that Binance sold its own assets. The same limitation applies to Coinbase. Establishing proprietary selling would require evidence connecting transactions to the exchange’s own holdings, rather than customer trading or custody activity.
The comparison also covers selected trading pairs and venues. It is not a complete accounting of global Bitcoin transactions, and its reliability depends on consistent data coverage and classification.
ETF redemptions show sustained withdrawals
U.S. spot Bitcoin ETFs recorded $1.07 billion in net outflows during the week ended November 13, their second consecutive week with withdrawals exceeding $1 billion, according to The Block’s analysis of SoSoValue data.
The funds experienced seven consecutive sessions of outflows totaling approximately $1.91 billion. Across the two full weeks, withdrawals reached about $2.35 billion. (theblock.co)
Those figures describe overlapping periods and should not be added together.
They demonstrate weaker demand through regulated Bitcoin investment products, although ETF investors include both institutions and individuals. They also do not establish where any related underlying Bitcoin transactions were executed.
For that reason, ETF outflows and exchange CVD cannot simply be treated as separate, additive amounts of selling. Some activity may overlap, while their units and measurement methods differ.
MARA confirms a substantial corporate sale
Unlike anonymous exchange activity, MARA’s disclosures identify a specific corporate seller.
In its third-quarter shareholder letter filed November 9, the Bitcoin miner said it sold 13,597 BTC after the September quarter-end at an average price of approximately $81,039. The transactions generated about $1.1 billion.
MARA used $750 million to repay its Bitcoin-backed term loans in full, resulting in the release of 17,477 BTC previously pledged as collateral. The released coins represent collateral returned to the company, not an additional sale. (ir.mara.com)
The disclosure documents a significant source of supply over several weeks. However, it does not establish how much of that selling occurred during the latest decline or quantify its effect on prices.
Its timing also differs from the November 9–14 CVD window, making a direct comparison inappropriate.
CZ denies Binance sold cryptocurrency
Binance co-founder and former chief executive Changpeng Zhao rejected the allegation that the exchange had sold cryptocurrency following the breach.
He also said he had used recently received stablecoins to buy crypto over the preceding days. His comments were reported by The Block on November 15. (theblock.co)
Zhao’s statement is a denial from a closely connected party, rather than an independent audit. His personal purchases, if confirmed, would also be separate from Binance’s corporate transactions.
The available evidence therefore supports a narrower conclusion than either categorical blame or categorical exoneration: the allegation that Binance sold assets to cover hack losses remains unverified, while other sources of selling pressure are documented.
For investors, the useful distinction is between trading observed on an exchange and transactions proven to belong to that exchange. Without that distinction, market data can give an unsupported rumor the appearance of evidence.