Open Interest Rose 7.59% While $844M of Shorts Were Liquidated: Reading the September 21 Squeeze

Bitcoin traded above $85,000 on September 21 and touched roughly $87,300, its highest print since January 2026. CoinGlass data cited by The Block put total crypto liquidations at $1.06 billion over the following 24 hours, with $844 million of that on the short side. The number that matters more is the one that did not fall: aggregate crypto open interest rose 7.59% to $156 billion during the squeeze, according to CoinDesk data cited by Invezz.

Open interest is supposed to shrink when positions get force closed. It grew instead. That divergence describes the session far better than the candle does.

Why Rising Open Interest During a Short Squeeze Is the Real Signal

A liquidation is a position closing. When a venue force closes a short, it buys the contract back and open interest on that contract falls. A cascade of $844 million in short liquidations should leave a visible hole in the aggregate number.

Instead the aggregate climbed. ChainCatcher, citing Coinglass, reported bitcoin network wide contract positions up 7.97% in 24 hours as of September 22. Twenty four hour volume rose 39% to $224 billion over the same window, per the CoinDesk figures relayed by Invezz.

The arithmetic resolves only one way. New positions opened faster than old ones were destroyed. The market did not step back from risk during the move. It added risk into the move, at higher prices, with the short side that supplied the fuel now removed from the book.

That is a different structure from a squeeze that ends with a flushed book. A flushed book is clean and can absorb the next shock. A book that replaces every liquidated short with a fresh long near $86,000 carries the same fragility, just pointing the other way.

The Funding Rate Flip Shows Who Pays to Hold the Trade Now

Funding sat at 0.0075% per eight hours after the squeeze, which annualizes to roughly 8.18%, per CryptoTicker. That is not an extreme reading. Crypto has run funding at three and four times that level during genuine mania phases.

What changed is direction and who pays. Before September 21 the short side was collecting. Now longs carry an eight percent annual cost of holding, and that cost is charged against positions opened after a 5% move rather than before it.

A senior research analyst at Nansen described the session to crypto.news as a combination of returning ETF demand and a large short squeeze, noting that price turned bullish faster than positioning did. That gap between price and positioning is exactly where funding does its work. Either spot demand catches up and the funding bill gets paid out of unrealized profit, or it does not and the traders who paid to enter become the next block of supply.

Forced Buying and Voluntary Buying Look Identical on a Price Chart

This is the practical problem with reading any squeeze. Both produce green candles. Neither is distinguishable on a daily or even hourly bar.

Cumulative volume delta separates them. CVD tracks aggressive market buys against aggressive market sells, so it measures intent rather than outcome. Forced buying from liquidations arrives as a vertical CVD burst clustered precisely at the price levels where short margin ran out. Voluntary buying arrives as a slower continuous CVD climb that persists after the cluster ends.

The test on this move is simple and it plays out over the next several sessions. If CVD keeps grinding higher above $85,000 while open interest holds flat, the forced buying was handed off to real buyers. If CVD flattens or rolls over while open interest stays parked at $156 billion, the only bid that ever existed was the one the exchanges sent involuntarily.

Charles Schwab head of crypto research attributed the 5% morning move directly to short perpetual contracts being liquidated in comments to CoinDesk. That attribution is the starting point of the analysis, not its conclusion. What replaced those shorts is the open question.

What the 50-Week Reclaim Does and Does Not Prove

The structural argument for this being more than a squeeze is that bitcoin closed last week above its 50 week moving average for the first time in 45 weeks.

Galaxy Digital research head, in work summarized by Invezz, found that across the five completed bitcoin bear markets that lost the 50 week average, the first successful weekly reclaim after the cycle low was not followed by another lower low in four of them. The 2021 to 2022 cycle was the exception.

Four out of five is a pattern with a sample size of five. Worth knowing, not worth sizing a position around on its own. What it does suggest is that the squeeze may have accelerated a structural shift already underway rather than manufacturing the entire move from nothing.

The cleaner confirmation is behavioral rather than statistical. Holding above the former $82,000 resistance zone after forced buying stops is evidence. The moving average reclaim is context around that evidence.

The Levels and Flows That Settle the Question

Three things resolve this over the next week. Whether open interest stays near $156 billion or bleeds lower as the replacement longs get shaken out. Whether funding holds a mild positive or climbs toward the levels where carry starts forcing exits on its own. And whether spot CVD confirms what the derivatives book is currently pricing.

The Buildix screener tracks CVD, open interest, funding and order book imbalance across 530 plus pairs in a single view, and the pair page at buildix.trade/pair/BTC shows where liquidation clusters sit relative to current price. Watching open interest and CVD on the same screen is how you separate a squeeze that finished from a squeeze that simply got refinanced.

Bitcoin near $85,400 on Tuesday morning is a higher price than it was on Friday. Whether it is a better price depends entirely on who owns it now, and the book says a meaningful share of it is owned by traders who bought after the move, on margin, paying to hold.