Independent strategy evidence library

Most trading strategies don’t work. We publish the evidence.

Every strategy faces the same test: multi-year market data, exact parameters, and unseen out-of-sample periods. We show where it worked, where it fell apart, and whether it beat simply holding. Most fail. A rare few survive. We have no signals to sell and no exchange to push, so we can say plainly which is which.

Inspect all 48 tests Read the testing standard The leverage guide See how a backtest lies The full scoreboard (48 tested) Why most strategies die

01 / DEFINE Rules before results Entries, exits, fees, timeframes and parameters are recorded exactly.

02 / EXECUTE Filled at the next bar A signal read at the daily close is filled at the NEXT close — a full bar of lag, never look-ahead.

03 / HOLD OUT Judged on recent unseen data The most recent ~45% of each coin’s history is held back as an out-of-sample check (not rolling walk-forward).

04 / COMPARE Doing nothing is the benchmark Returns mean little unless they beat buy-and-hold after costs and risk.

The evidence wall

Forty-eight beliefs entered. Eight held up.

This is not a leaderboard. It is a public record of what happened under the same honest standard. “Survived” means the edge persisted out-of-sample—not that future profit is guaranteed.

Showing all 48 records Verdicts include fees and out-of-sample performance

The rare survivors

Not miracles. Just edges that refused to disappear.

Surviving strategies tend to be simple, patient and uncomfortable. They trade less, endure long flat periods, and still need disciplined risk management.

Survivor 01

9/21 EMA crossover

A plain momentum rule with no decorative filters. It remained useful across different assets and market regimes.

Beat holding on 7 of 10 coins · Held out-of-sample

Survivor 02

Ichimoku cloud

Complex on the chart, but structurally coherent: trend, momentum and support were tested as one fixed system.

Beat holding on 7 of 10 coins · Held out-of-sample

Survivor 03

EMA-50 daily flip

A slow regime switch that accepts late entries in exchange for avoiding parts of prolonged market declines.

Out-of-sample validated · held up recently

A survivor is not a promise. It is simply a strategy that earned the right to be taken seriously. Tomorrow’s data may still kill it—and if it does, the record will change.

Featured autopsy · No. 002

The Golden Cross

Six trades. Seven years. Behind doing nothing.

The Golden Cross sounds sensible: buy when the 50-day moving average crosses above the 200-day average. On Bitcoin, it made +506%. That looks impressive—until the honest benchmark enters the room.

Test period 7 years

Golden Cross return +506%

Buy-and-hold return +883%

Total trades 6

Cross-asset result Beat hold 5/10

Verdict NO EDGE / COIN FLIP

Growth of $100

Bitcoin · Seven-year test
Illustrative equity path

The strategy made money. It still failed the useful question: did the signal add value? Here, the answer was no.

The one principle

We would rather publish an ugly result than sell you a beautiful lie.

No hidden losing periods · No optimized hindsight · No verdict for sale

No code. No jargon. Just your idea.

Got a strategy in your head? You don’t need to build it.

You don’t have to be technical or know Python. Just describe the idea in plain English to Vera, our assistant — she works out exactly what you mean, turns it into a real strategy, and runs it through the same honest test as everything here. Most ideas don’t survive. A few do. You’ll know in minutes, not in losses.