August 2026: Bitcoin Rallied 24% and the Bear Call Didn't Move

Three weekly runs, 13 to 27 August 2026.

Lead

On 13 August bitcoin traded at $63,776, sitting on its 200 week average with the Fear and Greed index at 29. Thirteen days later it closed at $78,955. Nothing in the month’s research moved the cycle word off bear.

The break came on 19 August. A put to call ratio in the 99th percentile met one headline about Treasury buybacks, and between $1.1 billion and $3.0 billion of short positions were liquidated inside a few hours. Then the flows followed, eight straight days of them. Futures open interest fell to a five month low of 587,584 BTC while price climbed, which is shorts closing rather than longs opening.

Price moved 24% and the evidence behind the bear call didn’t move at all. Every run of the month kept the base case low below spot, and the last one said so plainly: price “trades above all three ranges after the August squeeze.” If the models are right, this rally is a bounce with a long fall still underneath it. If they’re wrong, a framework built on four halving cycles just lost its steadiest clock.

Institutions

Twelve firms ended August with a book I could source, and nine of them carry no bitcoin at all. The 1.29M coins counted above belong to a different set, the five firms whose disclosure is denominated in bitcoin rather than in fund shares, and two of those five publish no book. Across the four that appear in both the month’s first and last research runs, the total didn’t move.

Strategy is most of it: 840,447 bitcoin at an aggregate average cost of $75,385, which left the position under water for most of the month. It bought nothing for a second straight week, sold stock instead, and parked $1.59 billion in a new cash pool. The donuts show what that does to a balance sheet, and what it never did to Jane Street’s: the largest book on the page keeps a rounding error in crypto, an authorised participant’s ETF inventory rather than a view.

The banks and the endowments split. JPMorgan added across assets, the broadest multi asset move any bank made. Two sovereign funds sat still, and Harvard’s endowment went flat for the first time after cutting 43% the quarter before. Selling stopped. Buying didn’t start.

Four firms sit below the ten rows above and matter more than their size. Renaissance Technologies added 312.9% to a stake that wouldn’t cover a day of ETF inflows. Macquarie cut 61.8%, though it clears ETF creations for a living, so its number moves with client flow. Intesa Sanpaolo cut its IBIT holding 93.7% and flipped its options posture bearish. Banco Santander filed its first ever bitcoin ETF position, worth 0.027% of its book. That’s European bank adoption seen from inside Europe. One large lender leaving through the side door, another arriving with a rounding error, in the same fortnight. Two of the continent’s biggest lenders are telling you it’s still a test position here.

Outside bitcoin only ether drew a bullish signal, in two of the month’s three runs, on spot ETFs that matched bitcoin’s eight day inflow streak. Solana stayed mixed in every run. Ripple and the equity proxies drew bearish, and the equities lagged the rally they are meant to track.

Cycle clock

Two clocks in this series barely move. Peak to bottom has run 364 and 366 days in the two falls the brief averages, and 406 days in the thin early cycle it excludes from every average. The lead from a bottom into the next halving has run 542, 513 and 528 days. The provisional low of $58,525 on 30 June landed 267 days after the October 2025 peak, early against every one of them.

Measured against days from each cycle’s own peak, this fall has reached 53.1% at its deepest. The two the brief averages went to 83.8% and 76.7%, and the excluded thin cycle went deeper than either, to 84.5%. Either bitcoin’s holder base changed when the spot ETFs opened in 2024, or the fall isn’t finished.

Daily close, drawdown from each cycle's peak. The band and the months below it are this cycle's. The 2013 line traded in a thin market. The 2013 and 2017 lines run off the top of the chart to 17.3x and 3.4x.

Falls from the 2017 and 2021 peaks, 2 cycles

Metric	Mean
Halving to peak, days	540
Peak to bottom, days	365
Max drawdown	-80.2%

Counter argument

The strongest case against all of it is the 30% the research assigns to it. The low of 30 June already holds. The evidence is decent: this fall is 23 to 27 points shallower than any predecessor at the same point in a cycle, which is what a permanent bid looks like. Eight of VanEck’s twelve capitulation signals had fired by 18 August, and in 2015, 2018 and 2022 the low arrived within weeks of readings like that. Mining difficulty is down 15.1% since January. The largest treasury holder stopped selling. Bernstein published a $150,000 target for mid 2027.

Put together, that’s a market where the worst is priced and the marginal holder is an ETF that doesn’t panic. If that’s right, the base case band of $46,000 to $55,000, due between 4 October and 30 November, never prints and the window closes with nothing in it.

The weakness is timing rather than logic. The June low sits 97 days short of the tightest peak to bottom interval on record. To believe it’s in, you have to believe the ETF changed a cycle’s amplitude without touching its calendar. That’s possible. It’s also the one claim in the whole argument that nobody can test until December.

Scorecard

The wrong call this month was mine. On 13 August this brief wrote that “seven independent on chain gauges all sit to the right of their prior bottom bands,” and that a 2022 style MVRV on that day’s realized price “implies about $39,500, a further 38% from here.” Bitcoin rose 24% instead, and on 21 September it closed at $86,597. The gauges still haven’t reached any floor, so the premise survived. The direction it implied did not.

The bull case from the same run reads better. It argued that “the cascading liquidations that made prior bottoms are largely spent.” That was right, and right for the wrong reason. The flush hadn’t cleared longs out of the way of a rally. It had cleared them out of the way of shorts, who then paid for the rally themselves.

Coins to follow

Seven of the ten appear nowhere else in this brief. Read the two Federal Register clocks first: the SEC’s Regulation Crypto Assets comment period closes on 20 October and the Treasury’s stablecoin rule closes on 19 October. The one that tells you most about sentiment is Balancer, where the DAO is voting on winding itself down.

Prediction

Bitcoin closes above $55,000 every day through 30 November 2026, so the base case bottom band of $46,000 to $55,000 expires with the window untouched.

Scored in the September issue.

Sources

• Coinbase, 2026‑08‑13, https://api.coinbase.com/v2/prices/BTC‑USD/spot
• Coin Metrics, 2026‑08‑27, https://community‑api.coinmetrics.io/v4/timeseries/asset‑metrics?assets=btc&metrics=PriceUSD&frequency=1d
• CoinGecko, 2026‑08‑26, https://api.coingecko.com/api/v3/coins/bitcoin/history?date=26‑08‑2026
• Alternative.me, 2026‑08‑26, https://api.alternative.me/fng/
• Newhedge, 2026‑08‑13, https://newhedge.io/bitcoin/200‑week‑moving‑average‑heatmap
• The Block, 2026‑08‑19, https://www.theblock.co/news/markets/2026‑08‑19‑bitcoin‑69000‑ether‑jumps‑treasury‑buybacks‑sec‑crypto‑proposal‑fuel‑market‑rally‑412248
• CoinDesk, 2026‑08‑25, https://www.coindesk.com/markets/2026/08/25/a‑bitcoin‑short‑squeeze‑for‑the‑ages‑as‑futures‑open‑interest‑collapses
• CoinDesk, 2026‑08‑27, https://www.coindesk.com/business/2026/08/27/live‑updates‑bitcoin‑etf‑inflows‑hit‑eight‑straight‑days‑as‑august‑tops‑usd3‑billion
• VanEck, 2026‑08‑20, https://www.vaneck.com/us/en/blogs/digital‑assets/matthew‑sigel‑vaneck‑mid‑august‑2026‑bitcoin‑chaincheck/
• news.bitcoin.com, 2026‑08‑19, https://news.bitcoin.com/crypto‑news/vaneck‑bitcoin‑capitulation‑signals‑august‑2026/
• checkonchain, 2026‑08‑13, https://charts.checkonchain.com/
• CryptoSlate, 2026‑08‑14, https://cryptoslate.com/bitcoins‑q2‑selloff‑split‑wall‑street‑as‑banks‑bought‑hedge‑funds‑cut‑and‑sovereigns‑held/
• Coinpedia, 2026‑08‑14, https://coinpedia.org/news/jpmorgan‑reveals‑bitcoin‑ethereum‑and‑xrp‑etf‑holdings‑in‑q2‑filing/
• Crypto Briefing, 2026‑08‑14, https://cryptobriefing.com/harvard‑blackrock‑bitcoin‑etf‑stake‑q2/
• Spoted Crypto, 2026‑08‑04, https://www.spotedcrypto.com/intesa‑sanpaolo‑ibit‑stake‑cut‑2026/
• TFTC, 2026‑07‑29, https://www.tftc.io/banco‑santander‑ibit‑bitcoin‑etf‑13f‑q2‑2026
• SEC, 2026‑08‑24, https://www.sec.gov/Archives/edgar/data/1050446/000119312526361845/mstr‑20260824.htm
• gncrypto.news, 2026‑08‑23, https://www.gncrypto.news/news/bitcoin‑mining‑difficulty‑near‑2026‑low‑150‑ehs‑idles/
• Cryptonomist, 2026‑08‑26, https://en.cryptonomist.ch/2026/08/26/bitcoin‑price‑forecast‑2029/
• Federal Register, 2026‑08‑21, https://www.federalregister.gov/documents/2026/08/21/2026‑17183/regulation‑crypto‑assets
• Federal Register, 2026‑08‑18, https://www.federalregister.gov/documents/2026/08/18/2026‑16796/genius‑act‑regulations‑on‑payment‑stablecoin‑issuance‑offer‑and‑sale
• CoinGecko, 2026‑09‑21, https://www.coingecko.com/en/coins/bitcoin/historical_data

The author holds no crypto assets, as of 22 September 2026.

Informational research, not licensed investment advice. Crypto assets are highly volatile and historical cycles do not reliably predict future prices.