Harmonic Gartley (ratios, 15% tolerance) · single target
Holding to the final target captures the whole tail of the move instead of cutting it short with partial profit-taking.
Measured in crypto — Binance spot · 540 pairs, delisted ones included · 0.2% per round trip
- ✗invariance
- ✗costs
- ✗placebo
- ✗benchmark
- ✗out of sample
- ✓multiple testing
At its worst the account was worth 91% less than its own best previous moment, and it spent 1,860 days below that peak.
What was measured
How many of those trades actually count
What it paid, before any deductions
Why it did not pass — it died here · where the result came from
What would have happened to the money
What this result does NOT say
- One market, one universe
- Measured on 441 spot cryptocurrency pairs, delisted ones included. It says nothing about futures, equities or indices, nor about how the same technique behaves in another market.
- One window of time, not every window
- The measured period runs from 2017-08-17 to 2026-07-28. A market moves through regimes, and a technique can work in one and fail in another — the card measures the regimes that fit inside this window, not the ones still to come.
- One cost structure
- The cost charged is 0.10% per leg, in and out. Anyone paying more than that gets a worse result, and anyone paying less gets a better one — the verdict holds for this fee.
- One exit rule
- The trade was closed by: stop or target, whichever comes first. The same entry measured with a different exit is a different strategy, and can earn a different verdict — it happens in this archive.
- The number of trades is not the sample size
- There are 3,599 trades, but only 326 independent market episodes: a single move fires the technique across dozens of assets at once, and counting those as separate observations inflates any result. It is the smaller number that governs the arithmetic. With 326 episodes, what the data supports is a range from -6.18% to -1.84% per trade — the published average is the centre of it, not the exact measurement.
- Daily bars
- Measured at the daily close. Nothing here measures what happens inside the day, and an intraday technique is not auditable with this data.
Numbers and reproducibility
The six controls
| control | status | t | threshold | episodes |
|---|---|---|---|---|
| invariance | failed | — | — | — |
| costs | failed | -3.82 | 1.97 | 326 |
| placebo | failed | -6.37 | 1.97 | 326 |
| benchmark | failed | — | — | — |
| out of sample | failed | — | — | — |
| multiple testing | passed | — | — | — |
- invariance88% of the gross profit comes from 179 trades (5% of the total) — lottery
- costsgross -4.011% · cost 0.200% · net -4.211% (t=-3.82)
- placeboactual -4.211% · placebo +3.038% · excess -7.249% ± 1.137% (t=-6.37 against a threshold of 1.97, 326 real groups, 177,911 sham dates, draw error ±0.049%)
- benchmarktechnique -4.21% · buy and hold (same horizon) +0.36% · excess -4.57%
- out of sampleasset half A: -4.169% (t=-3.98, 271 episodes) · asset half B: -4.250% (t=-2.77, 284 episodes) · liquid half (>= US$ 2,240,681/day): -3.927% (t=-3.02, 288 episodes) · illiquid half: -4.544% (t=-3.22, 255 episodes) · period 1/4 (2017-11-14 a 2022-06-16): -3.948% (t=-1.59, 128 episodes) · period 2/4 (2022-06-17 a 2024-02-13): -2.745% (t=-2.30, 84 episodes) · period 3/4 (2024-02-14 a 2025-01-12): -4.714% (t=-2.02, 43 episodes) · period 4/4 (2025-01-13 a 2026-07-27): -5.432% (t=-4.43, 74 episodes)
- multiple testing4 variation(s) tested · t=-3.82 across 326 episodes (equivalent to t=-3.81) · p≈0.0001 · false positives expected by chance ≈ 0.00
Equity — outside the six controls, and here is why
The t of the trade series is invariant to bet size: 0.5%, 1% and 3% agree to the sixth decimal. Nothing here moves the verdict — it moves what the account would have lived through.
- risking 1.0% per trade
- ×0.22
- worst drawdown from the peak
- −91%
- days below the previous peak
- 1,860
- signals refused for lack of capital
- 0%
- paths where the account halved (out of 12)
- 12
Reproducibility
- period
- 2017-08-17 to 2026-07-28
- assets that traded
- 441
- variations tested before this one
- 4
- gross per trade
- −4.01%
- net per trade
- −4.21%
- exit rule
- stop or target, whichever comes first
- median duration bars
- 1
- max duration bars
- 565
- intrabar ambiguity
- 0.2
- fee per leg
- 0.001
- episode days
- 7
- requires ratios
- 1
- fibonacci tolerance
- 0.15
- confirmation bars
- 5
- point D window
- 60
- partial profit-taking
- 0
Binance spot klines (delisted pairs included) · collected from 2026-07-27 23:31 to 2026-07-28 14:29 · 540 assets · 750,934 bars · 2017-08-17 to 2026-07-28
Seed not recorded: audits before 2026-07-29 used 20260728 by convention in the code, and the value is not in the record.
Hypothesis, filed before the result
Harmonic Gartley. The description does not close on its own arithmetic: the two routes to point D (the CD extension, and 78.6% of XA) disagree by 10.4% and 17.0% of the XA leg, and the BC=38.2% branch is IMPOSSIBLE — it would require CD=1.712, outside the canonical range. Point D was taken at 78.6% of XA, which the description itself calls the most important number. Four forms were tested to separate what the technique actually claims: with the ratios at 3% and at 15% tolerance, with no ratio at all (the zigzag alone), and with a single target in place of partial profit-taking.
filed on 2026-07-28, before the number existed
The original, as it was filed
Gartley harmônico. A descrição não fecha na aritmética: os dois caminhos até o ponto D (extensão de CD, e 78,6% de XA) discordam em 10,4% e 17,0% da perna XA, e o ramo BC=38,2% é IMPOSSÍVEL — exigiria CD=1,712, fora do intervalo canônico. Adotado o ponto D pelos 78,6% de XA, que a própria descrição chama de o número mais importante. Testadas quatro formas para separar o que a técnica alega: com as razões a 3% e a 15% de folga, sem razão nenhuma (só o ziguezague), e com alvo único no lugar da realização parcial.
Pre-registration exists to keep prediction apart from rationalisation: written after the number, every hypothesis is right.