Bitcoin's 23% Week: Price Round-Tripped, Usage Didn't

Bitcoin's 23% Week: Price Round-Tripped, Usage Didn't

At mid-May, bitcoin traded near $77,000 — about $77,400, to be specific. By late June it had touched prices under $60,000. Through the first half of August it was still hunkering around $64,000, volatility at six-month lows, as if the market had quietly given up. Then, in seven sessions through August 24, it climbed 22.7% to close the week at $77,489 — almost exactly the price it had left behind three months earlier. In one week, the entire spring-summer slide was erased.

Start with the distinction the headline blurs. "Wiped out three months of losses" is not the same as "recovered." Even after the burst, bitcoin was down roughly 9% for 2026 and about 30% from a year earlier. The week did not take the asset anywhere new. It took it back to May. That is a round-trip inside a year-long downtrend, not a breakout, and it changes how the move should be read. Notice the froth isn't spreading either: the S&P 500 has been climbing quietly in a calm uptrend with a neutral momentum reading — this is a bitcoin-specific event, not a broad speculative re-rating.

So what actually moved the price in those seven days? Three things, and only one of them is durable evidence.

The first leg was mechanical. On August 20, short sellers lost nearly $2.7 billion in 24 hours as bitcoin surged — reported as the largest one-day wave of forced short liquidations. A squeeze works like this: as price rises and shorts run losses, margin calls force them to buy back and close, and that forced buying is itself buying. It is convexity, not conviction. It explains speed; it explains nothing about demand that would still be there a month from now.

The second leg is the real one. Spot bitcoin ETFs pulled in $1.92 billion net in the week ending August 24 — roughly 71% of the $2.71 billion that flowed into all crypto ETFs that week — pushing August's cumulative intake to a 2026 high. Regulated funds absorbing billions a week is institutional demand on the record, the cleanest evidence the move has. This is the fact that makes the rally something more than a short-squeeze headline.

The third leg is a story. The narrative, as Bloomberg reported, leans on fears that U.S. fiscal strains are pushing money outside the government system — with Ray Dalio pointing at the debt, a Treasury plan to roughly double long-dated bond buybacks, a softer dollar, gold running alongside, and the administration pressing Congress on crypto market-structure legislation. That is a mood, not a network metric. It explains why people wanted the trade. It says nothing about whether the asset's own economics improved.

And here is the part every chart-touting post leaves out: bitcoin's own usage data has not caught up. The most recent detailed on-chain reading I can verify, from Glassnode in mid-July, described the recovery as driven by ETF flows on thin liquidity, with weak organic network demand "despite higher prices" — active addresses declining, transaction fees falling, spot volume down more than 20%. When active addresses did spike this month — near 980,000 a day, an eight-month high — the cause was a security scare, users migrating funds after a wallet-firmware vulnerability, not a wave of new economic use.

In stock terms, this is the divergence I always hunt for. With a company, I call it valuation disconnected from business fundamentals: the price moves one way while the operating data moves the other. Bitcoin is the version without an income statement, so the business thermometer is network usage — and it is reading cold. One side of this market writes checks (the ETF buyer). The other side, the users, is still on pause.

Timing has also turned against the new buyer, even if the longer story is right. The 14-day RSI — a momentum gauge where anything above 70 is already stretched — sits near 82, and price is roughly 14% above its 200-day average. More telling, the market has already shown what that looks like. On August 22, bitcoin ran toward $79,500, then flash-crashed to about $76,300 within hours as more than $1.3 billion of leveraged positions were liquidated across crypto and the four-hour RSI hit its most extreme reading in over seven years. The same weekend, XRP gave back 37% in one violent snap. The squeeze ran on leverage, and that leverage is now pointed the wrong way. The precedent is unhelpful too: the last time the daily RSI ran this hot at nearly these prices, in early May, the same overbought warning flashed — and bitcoin then fell more than 20% to its late-June lows, the very losses this week "wiped out."

Here is the honest part. The low-risk entry was not this week. It was in late June below $60,000 and through July in the low $60,000s, when the price was pricing doom and nothing structural had broken. That window has closed. Buying after a 23% week at an RSI extreme is entering on an event, not a setup — and events have a way of delivering their worst prices to the people who panic hardest about missing out.

None of this makes the move fake. The flows are real, the fiscal-fear audience is durable, and a structural case for bitcoin doesn't need this week's permission. What it needs is proof. Watch the next several weeks for three things: ETF inflows staying positive, network usage turning up — active addresses and fees, excluding security-driven one-offs — and price holding around the old $70,000 to $72,000 zone near its 200-day average on the first real pullback. If those hold, the round-trip earns the word recovery. If flows stall and $70,000 fails, this week gets a simpler name: a squeeze on thin liquidity.

A one-week round-trip is a price, not a verdict. The evidence that turns it into a recovery is still unearned — and the disciplined position is to wait for it rather than pay up for the emotion.