Crypto Liquidations Reach $545 Million as BTC Slips Below $84,000

Key Facts

Crypto liquidations reached about $545 million over 24 hours as BTC traded below $84,000 on September 23, 2026, according to TokenPost. The outlet put liquidations of leveraged long positions at $447 million, making them the majority of the reported total. ETH touched $2,635 during the same episode, the report said. Separate reporting confirmed a brief BTC move below $84,000, although it used a shorter window for its liquidation figure. Together, those observations establish a substantial wave of forced position closures; they do not establish that those closures initiated the price decline.

The reported liquidation figures cover a rolling 24-hour window, rather than a trading session with one closing time across crypto exchanges. TokenPost listed $98.75 million in short liquidations alongside $447 million in longs and an approximately $545 million overall total. These are figures captured for the report and can change as older events leave the window and new ones enter it. The notional value of a liquidated position also need not equal the cash loss suffered by its owner; it describes exposure that was closed. For that reason, the $545 million figure alone cannot measure investors’ net losses or the volume of spot selling in BTC and ETH.

A liquidation begins when a price move against a margined position erodes its collateral below the required maintenance margin. Binance says the liquidation of its futures contracts is based on a contract’s mark price, rather than a single last traded price. Once a position reaches the threshold, the exchange’s system may reduce exposure through market orders to restore the required margin. Such orders can add selling pressure during a rapid decline and push other positions toward their limits, explaining a possible amplification mechanism without identifying the original cause. The spot price of BTC and a derivative’s liquidation threshold are different measures, so touching $84,000 does not mean every long was closed at that price.

The breakdown shows that long positions accounted for most forced closures, although short positions were liquidated during the broader decline as well. TokenPost counted 126,630 affected traders, a reported count that does not disclose what any individual lost. It identified the largest single liquidation as a $10.04 million ETH/USDT position on Binance, illustrating that the aggregate combines events of very different sizes. The figures do not provide a reliable division of the full $545 million between BTC and ETH. Readers should therefore distinguish the documented long-side concentration from an unsupported claim about the precise allocation by asset or exchange.

In TokenPost’s account of the price path, BTC fell to $83,500 before rebounding to $84,314, leaving it down 2.46% when the report was prepared. ETH dropped to $2,635 and later traded at $2,682, a decline of 2.79% on the report’s measure. The recovery from both lows shows how spot prices can rebound while the running total of earlier liquidations remains high. Any comparison between prices and liquidations depends on when each was captured: a price is a point observation, while liquidations accumulate over a period. The $83,500 and $2,635 lows should therefore not be presented as current prices or compared without qualification with figures compiled over a different window.

TokenPost associated part of the pressure with higher U.S. Treasury yields, but its report did not isolate the effect of yields from the unwinding of leveraged positions. In financial terms, a higher bond yield can raise the relative cost of holding a non-yielding asset for some investors. Coincident moves in yields, BTC and ETH, however, cannot show which factor led the liquidation wave or how much each contributed. That limitation matters more when liquidation data combine several exchanges whose contracts, execution prices and margin settings differ. The disciplined reading separates measured events, namely prices and forced closures, from a macroeconomic explanation that would require further evidence.

For an unleveraged holder of BTC or ETH, another trader’s liquidation does not close the holder’s position, although the resulting market move can affect an available trading price. For a leveraged long, the immediate variables are the distance to the contract’s liquidation threshold and the collateral left to support it. Binance’s protocol allows a position to be reduced before a full closure, another reason that reported liquidation value should not be treated as the trader’s loss. Heavy long liquidations do not prove that the decline has ended: leverage can fall while spot selling continues. Nor does a large aggregate prove that underlying demand for BTC and ETH has vanished or that $84,000 must remain broken.

The next useful comparison is whether prices stabilize above the reported lows as liquidation totals stop expanding in comparable measurement windows. If BTC holds above $84,000 while long liquidations slow, that would fit an easing of forced-selling pressure without guaranteeing a reversal. If it falls below that level again alongside additional liquidations, the case that leverage is amplifying the move would strengthen, subject to matching the data timestamps. For ETH, the reported $2,635 low is a historical reference for this episode, not a forecast of fixed support. Any trading assessment needs fresh prices and fresh liquidation figures with a stated time window, rather than extending the September 23, 2026 snapshot indefinitely.

The CFTC has scheduled its next Commitments of Traders release for September 25, 2026, at 3:30 p.m. Eastern time. That report generally describes futures and options positions as of the preceding Tuesday, making it a delayed weekly view of positioning. It does not directly measure the cross-exchange crypto liquidations TokenPost described for a 24-hour window around September 23, 2026. It may put some regulated-futures positioning in broader context, but cannot by itself establish why BTC or ETH fell in this episode. The nearer test is to compare prices and liquidations over the same period and assess whether selling persists after forced closures subside.