Bitcoin Squeezed the Bears. Now the ETFs Have to Carry It.

Bitcoin just finished its best week in years, climbing from roughly $62,000 to nearly $80,000, and it did it in a way that looks backward on the futures ledger. Open interest — the number of derivative contracts still open — fell to a five-month low even as price jumped. A rally built on a collapsing book is a rally paid for by forced buying, not fresh conviction, and that distinction is the whole interesting question: who bought, and who has to keep buying now that the squeeze is over?

Start with who got hit, because the trade was not evenly stacked. Going into last week, shorts held just over half of open interest on the major exchanges, and funding rates — the periodic fee one side of a perpetual swap pays the other — were negative, meaning bearish traders were paying rent to keep their positions on. That is the signature of a crowded short: everyone piled on one side, confident, paying to wait for the drop. When price instead blew through resistance, the crowded side had no choice but to buy back. In a violent stretch from August 19 into the 20th, over $3 billion of leveraged positions were liquidated, roughly 92% of it short sellers; Bitcoin shorts alone lost about $1.67 billion. The exchanges and data houses labeled it the largest short squeeze since late 2021.

So what tripped the pile? Not the Fed. The Fed sat flat, holding its policy rate at 3.5%–3.75%. The trigger came from the Treasury, and it is worth pausing on because this is exactly the kind of operation people cheer without reading the entries. On August 19, Treasury Secretary Scott Bessent said the department would at least double the maximum size of its long-dated bond buyback program, from $2 billion to $4 billion per operation, starting September 9. It came after the 30-year Treasury yield had spiked to 5.31% on August 17, its highest in nineteen years. On the announcement, the 20-year yield dropped 11 basis points and the 30-year fell 9, while the 2-year did not move at all — the market read it as a supply event, not a change in rate expectations.

Now the entries question, because this is not quantitative easing and saying it is would be a lie. A Treasury buyback creates no new reserves and no new money. The Treasury repurchases its own older, off-the-run bonds and finances the purchase by issuing new short-term bills — a swap, old long paper for new short paper, total debt unchanged. What changes is still real at the margin: it lifts illiquid long bonds off dealer balance sheets that are constrained by capital rules, it presses long-end yields down, and it broadcasts that a Treasury worried about its own borrowing costs has a hand on the tiller. Bitcoin, sitting at the far end of the fiat-liquidity pipe, reacted before anyone stopped to audit the scale.

And scale is where the honesty check lives. Four billion dollars per operation is a trickle in a market that pays out roughly $125 billion in coupons in a normal week, and the whole program represents under 1% of more than $30 trillion of Treasury debt. There is even a counter-entry: funding the buyback by issuing extra bills pushes supply onto the short end, which can press repo and short-term funding markets and offset part of the easing. So this is a genuine liquidity channel, but a modest one — the early spike owed as much to the signal as to the dollars actually spent.

Here is what the plumbing says happened next. The 22% weekly gain ran the squeeze's course in roughly a week: aggregated open interest in coin terms fell about 11% while price rose 22%, and analytics firms read that as short covering and position closure, not new leverage stacking in. Margin-futures open interest touched a record low of about 52,000 BTC, a sliver of the whole market, and annualized funding on perpetuals stayed in single digits after flipping positive. Put plainly: the fragile, crowded short is gone, and nobody leveraged up to replace them. The tinder has been burned.

That leaves one buyer standing, and it is the one that matters for durability: the exchange-traded funds. U.S. spot bitcoin funds pulled in roughly $1.92 billion in the week of August 17–21, their strongest week since October 2025, with inflows accelerating in the two sessions right after the Treasury announcement. This is the load-bearing difference between a real rally and a derivative mirage. Short covering is one-time buying — it stops the moment the shorts are finished being forced, and the falling open interest is the forensic proof the stack is spent. An ETF creation is recurring: money arrives because institutions decide bitcoin belongs in the portfolio, and the fund operator then has to go buy physical coin. A rally that rests on ETF demand can outlast a squeeze; a rally that rests only on a squeeze cannot.

Keep the base rate honest. At about $79,000, bitcoin still trades roughly a third below its record near $126,000 from October 2025, and it is still down on the year. This is a bear-market relief rally with a real catalyst behind it, not a new all-time regime. The concentration is telling too: with bitcoin dominance near 60% and the altcoin-season gauge reading 27, the money is hiding in the quality name — that looks like institutional and ETF flows, not the broad retail euphoria ramp that marks blow-off tops.

So the thing to watch is not price; price is the outcome. Watch open interest and funding — if they rebuild with longs paying, fresh leveraged fuel is arriving and the move has wind. If they stay flat while ETF inflows continue, the rally is being financed by cash, the healthier kind. And watch November 4, when the Treasury announces the next round of buyback sizes at the quarterly refunding — the one place where this trickle could plausibly become something bigger. The squeeze already spent itself last week; what happens next is a cash question, not a panic question. Read the futures book and the ETF feed, not the tape.