BTC Squeeze Breakout — 1h
Status: Regime entry: worked in one test window, not the other · Regime #3 · Listed 2026‑10‑14 Instruments: BTCUSDT perpetual only · Timeframe: 1h · Direction: long only Frequency: ~28 trades/yr · Fees: 0.1%/side plus real funding (sweep #1's spot‑taker rate; the perpetual's taker fee is 0.05%)
Why it is in the regime section, not the library
It failed the six‑check bar. It is listed under the regime bar, which asks four things of one of the two test windows: a net profit after fees and funding, still a profit without the best trade, a result above 95% of 400 random‑entry runs with the same exit, and at least 20 trades. It clears all four from August 2018 to August 2022 (58 trades). From August 2022 to August 2026 it lost 37.3%, and more without its best trade. Everything below the setup is disclosure, not a pass.
Where it fails the full bar:
- Two windows: −37.3% from August 2022 to August 2026 (PF 0.53, 168 trades), −43.1% without its best trade.
- Cross‑asset: ETH made +0.7% on 17 trades in 2018–22, then +8.5% on 65 in 2022–26, short of the random‑entry bar (91st percentile). BNB made +7.9% on 13 trades, then lost 4.8% on 128.
- Drawdown: 6.3 summed points in its good window, inside the bar of 20; 40.9 in the other.
It came out of sweep #1, pre‑registered on 2026‑08‑05, which killed it on its second window: "Same spec, adjacent windows, opposite result." That sweep ran before the random‑entry null joined the bar. The re‑audit of sweeps #1–#10, pre‑registered on 2026‑10‑14, ran the null on every window test from those sweeps that made money, made money without its best trade and had at least 20 trades: thirty in all. With a 95th‑percentile bar, about one window test in twenty clears by luck alone. Of the thirty, one or two would by chance; eight did, this one at the 99.8th percentile.
The setup (compilable spec)
Market: crypto BTCUSDT perpetual, 1h trading timeframe. Longs only. Squeeze (1h): the prior 48 bars' highest high is within 3% of their lowest low,
highest(48) ≤ 1.03 × lowest(48). Entry (1h):close crosses above highest(48). The range is the 48 bars before the breakout bar. Exit: trailing stop 2×ATR(14) below the highest price since entry. Fees: 0.1%/side (the spot‑taker rate sweep #1 charged) · funding charged and credited at every settlement held through.
Evidence: two windows on Binance spot candles
| Window | Trades | PF | Return | Strip‑best | Sum DD | Account DD | Random‑entry percentile |
|---|---|---|---|---|---|---|---|
| Aug 2018 → Aug 2022 | 58 | 3.73 | +51.8% | +39.1% | 6.3 | 4.5% | 99.8th |
| Aug 2022 → Aug 2026 | 168 | 0.53 | −37.3% | −43.1% | 40.9 | 39.5% | not run |
Returns are per‑trade percentages summed, like every return in the research log. The price path is Binance spot candles, as in every August sweep; the funding is the real USDT‑M perpetual series, which begins in September 2019, so holds before then paid none. Holding BTC made +203.6% and +169.0% over the same windows.
- Random entries, 2018–22: 400 runs of the same exit on random bars had a median of −11.9% and a 95th percentile of +19.9%. They took a median of 57 trades to its 58 and were in the market about as long (1.7% of the time, against its 1.6%).
- 2022–26 was not null‑tested: the re‑audit ran the null only on windows that made money with and without their best trade.
- Fees and fills: sweep #1 charged 0.1% a side, twice the perpetual's taker fee. It was never re‑run at 0.05%, and no run filled it at the next bar's open.
The record by year
| Year | 2018* | 2019 | 2020 | 2021 | 2022 | 2023 | 2024 | 2025 | 2026* |
|---|---|---|---|---|---|---|---|---|---|
| Trades | 6 | 19 | 29 | 2 | 14 | 45 | 32 | 56 | 23 |
| Return | +10.2% | +11.9% | +25.2% | 0.0% | +7.5% | −14.1% | −3.2% | −14.7% | −8.4% |
*2018 from August, 2026 through July. Each trade counts in the year it was entered.
- 2018 to 2020 made nearly all of the good window's money: 54 trades, +47.3%. In 2021 the setup came twice.
- Since 2023 it has lost money every year: 156 trades, −40.4% between them. The setup also came far more often: 45 times in 2023 and 56 in 2025, against 19 in 2019.
- Last high: October 2022, at +55.3% summed since August 2018. It ended July 2026 at +14.5%, 40.9 points below.
- Last twelve months (August 2025 to July 2026): 52 trades, −18.4%.
What the result depends on
Nothing around it was run. Sweep #1's pre‑registration ran a family's neighbors only when its primary passed, and this one failed its second window, so no other range width, lookback or trail was tested (the log's H2 section). At 4h the setup never occurred on any of the three coins in either window: an eight‑day range within 3% is that rare. It was priced at 0.1% a side only, and filled only at the signal bar's close. The rest of the evidence is the same rules on ETH and BNB, in the cross‑asset line above.
Before you adopt it
- It has lost money every year since 2023, 2026 so far included: 156 trades, −40.4% between them. Adopting it is a bet that BTC's tight ranges break out again the way they did from 2018 to 2020.
- One window on one coin is the whole case. No neighbor, other fee level or other fill was run, and on ETH and BNB it never cleared the bar.
- Paper‑trade it first. Adopting it watches BTCUSDT on the 1h chart and alerts every setup, about 28 a year; in 2025 there were 56.
Every number here can be re‑run: npx tsx scripts/research/replicate-sweep1-lowtf.ts for both
windows and every trade (the record by year sums them), and
npx tsx scripts/research/null-reaudit-sweeps-1-10.ts --only "#1 H2" for the null.