Trading Auditor
← back to every claim
FAILED

Bollinger Bands (20/2σ)

Price spends 95% of the time inside the bands, so touching them means being statistically cheap or dear. The band adjusts itself to volatility — widening in an agitated market and narrowing in a calm one — which avoids the false signal a fixed threshold gives when volatility changes.

Measured in cryptoBinance spot · 540 pairs, delisted ones included · 0.2% per round trip

Net per trade
−1.18%
after fees
The fee is charged on both legs: every trade pays to open and pays to close.
Died at
placebo
random dates pay the same or more
Worst drawdown
−99%
3,235 days underwater
How far the account fell below its own best previous moment.
Timeframe
1 day
The time grid this was measured on. The same technique on a coarser or finer grid is a different measurement, and can earn a different verdict.
  • invariance
  • costs
  • placebo
  • benchmark
  • out of sample
  • multiple testing
Equity curve
Risking 1.0% per trade, marked to market every day — not only when a trade closes. The line at 1.0 is the capital it started with.

At its worst the account was worth 99% less than its own best previous moment, and it spent 3,235 days below that peak.

The technique against chance
Same number of trades, same holding time, same assets — only the dates were drawn at random.
0%barthe technique−1.18%random dates−0.36%
The average trade and the typical one
The mean sits BELOW the median: most trades win a little and the tail loses a lot. It is the lottery inverted, and no less dangerous — what decides the outcome is the rare loss.
0%mean−0.98%median+0.62%

What was measured

We coded the rule exactly as it is described and let it trade on its own from 2017-08-17 to 2026-07-28. It found trades in 540 coins, 45,522 in total. Each one enters and exits at the price that existed on that day — the program never sees what comes next, which is the most common mistake people make testing strategies at home. Wherever the description was ambiguous we took the reading least favourable to the technique; whatever was left open is stated in the hypothesis, at the foot of this page.

How many of those trades actually count

Only 29 of the 45,522. Cryptocurrencies rise and fall almost in unison, so the same rule fires on dozens of coins on the same day — and that is one bet repeated, not dozens of different bets. Counting them all separately is the trick that makes a bad test look impressive.

What it paid, before any deductions

-0.98% per trade. Look at the sign: the average is already negative before any fee is taken out. The controls below still hold, but here they are not knocking down an edge — they are confirming there never was one.

Why it did not pass — it died here · against randomly drawn dates

We ran the whole thing again entering on randomly drawn dates and changing nothing else: same number of trades, same holding time, same assets. Chance did as well or better. That means the signal was not picking the moment: any moment would have done the same.

What would have happened to the money

Risking 1.0% of capital per trade — the most widely taught rule — across the whole period: The capital would have ended at 0.03× what it started with. At its worst the account was worth 99% less than its own best previous moment, and it spent 3,235 days below that peak. In 12 of the 12 paths tested, the account fell to less than half of what it started with.

The account would not fit every signal

59% of the trades had to be turned down: by the time they appeared, the money was already tied up in other positions. That matters because the average return per trade the technique claims is computed over trades that nobody could have taken all of.

What this result does NOT say

For an edge to be assertable here it would have to reach 4.54% per trade — that is the size that survives this archive's multiple-testing correction, and it rises as the archive grows. So FAILED means “we found nothing above that size”, and never “it cannot possibly work”. The instrument is far more sensitive than that: on synthetic data, with a clean effect, it separates from 0.69% upwards. The distance between the two numbers is the price of a real market and the price of publishing many claims. The difference matters, and it is the rule of this house: the card confronts the claim, never the person who made it.
What this card does not measure
Every verdict holds for the conditions it was measured under. These are this card's — and outside them the result does not apply.
One market, one universe
Measured on 540 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: the auditor's fixed horizon (10 bars). The same entry measured with a different exit is a different strategy, and can earn a different verdict — it happens in this archive.
The same technique, measured by another exit
This technique was also audited under a different exit rule — the technique itself (held until the opposite signal) — and there the verdict is <strong>FAILED</strong> (the most favourable result of 3 measurements on that exit). They are different questions about the same technique, and both are published: the claim that passes on a fixed horizon and the claim that passes on the exit the technique itself teaches <strong>are not the same claim</strong>.
The number of trades is not the sample size
There are 45,522 trades, but only 29 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 29 episodes, what the data supports is a range from -3.36% to +1.40% 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

controlstatustthresholdepisodes
invariancepassed
costsinconclusive-1.022.0529
placebofailed-0.702.0529
benchmarkfailed
out of samplefailed
multiple testinginconclusive
  • invariance45522 signals across 750934 barsconcentration: +31.23% of the profit sits in the top 5% of trades — with the dates shuffled, +35.96% (fails above +50.00%)
  • costsgross -0.981% · cost 0.200% · net -1.181% (t=-1.02) · the range runs from -3.562% to +1.200%
  • placeboactual -1.181% · placebo -0.362% · excess -0.820% ± 1.166% (t=-0.70 against a threshold of 2.05, 29 real groups, 2,276,100 sham dates, draw error ±0.016%)
  • benchmarktechnique -1.18% · buy and hold (same horizon) +0.03% · excess -1.21%
  • out of sampleasset half A: -1.238% (t=-1.23, 29 episodes) · asset half B: -1.126% (t=-0.85, 29 episodes) · liquid half (>= US$ 2,066,613/day): -1.645% (t=-1.43, 29 episodes) · illiquid half: -0.674% (t=-0.65, 28 episodes) · period 1/4 (2017-09-14 a 2022-04-10): -3.376% (t=-1.66, 15 episodes) · period 2/4 (2022-04-11 a 2023-11-08): -1.324% (7 episodes — too small, does not count) · period 3/4 (2023-11-09 a 2025-03-08): -1.718% (5 episodes — too small, does not count) · period 4/4 (2025-03-09 a 2026-07-18): +1.642% (5 episodes — too small, does not count)
  • multiple testing1 variation(s) tested · t=-1.02 across 29 episodes (equivalent to t=-0.97) · p≈0.3308 · false positives expected by chance ≈ 0.33

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
×0.03
worst drawdown from the peak
99%
days below the previous peak
3,235
signals refused for lack of capital
59%
paths where the account halved (out of 12)
12

Reproducibility

period
2017-08-17 to 2026-07-28
assets that traded
540
variations tested before this one
1
gross per trade
−0.98%
net per trade
−1.18%
exit rule
the auditor's fixed horizon (10 bars)
horizon bars
10
fee per leg
0.001
episode days
112
seed
20260728
band period
20
deviations
2.0

Binance spot klines (delisted pairs included) · collected from 2026-07-27 23:31 to 2026-08-03 10:49 · 540 assets · 750,934 bars · 2017-08-17 to 2026-07-28

Hypothesis, filed before the result

Bollinger Bands are the RSI with a moving threshold, and the specific promise is that adapting to volatility avoids the false signal of a fixed threshold. Specific, falsifiable prediction: the moving band does NOT change the verdict relative to the RSI — because the RSI's problem was never the threshold, it was that there is no reversal left to capture after costs. I expect, as with the RSI, a gross edge close to zero and a loss to buy and hold.

the filing date is not in this audit's record

The original, as it was filed

Bandas de Bollinger são o RSI com limiar móvel, e a promessa específica é que a adaptação à volatilidade evita o sinal falso do limiar fixo. Previsão específica e falsificável: a banda móvel NÃO muda o veredito em relação ao RSI — porque o problema do RSI não era o limiar, era não haver reversão a capturar depois do custo. Espero, como no RSI, vantagem bruta próxima de zero e derrota para comprar e segurar.

Pre-registration exists to keep prediction apart from rationalisation: written after the number, every hypothesis is right.

The same technique in the other market

The verdict held in the other market too, on independent data.

Earlier audits of the same technique

Each variation an author teaches enters as its own test, so that whatever might work in the strategy gets covered. The verdict held in all of them.

  1. this measurement →FAILED
  2. 2026-08-03FAILEDopen ↗
  3. 2026-07-28FAILEDopen ↗
  4. 2026-07-28FAILEDopen ↗
  5. 2026-07-28FAILEDopen ↗
  6. 2026-07-28FAILEDopen ↗
  7. 2026-07-28FAILEDopen ↗
  8. 2026-07-28FAILEDopen ↗

record 3ce70a6e639b · 2026-08-03 22:42

This code comes from this card's content: if anything here changed after publishing, the code would stop matching — that's how a change gets caught. We audit the technique, never the person — no record names an author, a channel or a brand.

The full record behind this verdict.