An edge that only exists in the last three years
Ninety-eight percent of a six-stream book's lifetime profit sat after 2022, which is why every result is now split by era before it ships.
- Verdict
- Settled by the data
- Decided by
- Era split
- The number
- 12-year out-of-sample: avgR +0.010, p = 0.39
The hypothesis, as pre-registered
A strategy validated on recent data with a clean walk-forward is ready to deploy.
What the test found
Taken back to a deep history, a book of six separately validated streams produced an average R of +0.010 with a p-value of 0.39 over twelve years — statistically indistinguishable from nothing — while ninety-eight percent of its lifetime profit came from the most recent stretch.
That pattern has two readings and they are not equally comfortable. Either the mechanism genuinely did not exist before and is real now, or the recent period is where the fitting happened. Distinguishing them requires knowing when the mechanism's ingredients were born: some data simply did not exist earlier, and an era split across a feed's birth date is not a fair test.
The discipline that comes out of this: split every result by era before believing it, and know the birth date of every input. An edge that cannot be found in any era but the one it was discovered in is a hypothesis, not a strategy.
What would make this work
The edge would have to be detectable in at least one era outside the one it was discovered in — or its ingredients would have to demonstrably not exist before that era.
Every insight ends with this line: the exact condition that would turn the finding into a tradeable edge. That is what makes each entry a next experiment rather than a full stop.
Screened September 5, 2026
Run the arithmetic on your own numbers with the deflated Sharpe ratio calculator.
Testing something similar?
The screener puts your version of this idea through the same gate stack, on your market, and answers in seconds. Your first verdict is free.
Related insights
Picking the strategy with the best return-to-drawdown ratio
Return persists weakly out of sample and the drawdown ratio does not, so rank on return and correlation and let the portfolio shape the drawdown.
SettledA strategy that only works on the 30-minute chart
Hold the entry instant and the stop distance fixed and an 18x span of bar sizes pays the same, which makes the strategy far more robust than it first appeared.