Research
What we measured,
including what failed.
Selepro publishes trading systems with their complete record. This section is the working half of that promise: the tests behind the numbers, the ideas that were tried and thrown away, and the corrections to things we got wrong and had already published.
Why the failures are here. A backtest anyone can produce. What is hard to fake is a trail of hypotheses that were written down, measured, and mostly rejected. In the work behind these notes, five were tested and four died. That ratio is the point, not an embarrassment.
Notes
We published a failed test. Then we found out why it failed.
The strategy was tested on Ethereum, lost money, and the failure was published in the script description. Re-measured under control, the failure turned out to be ours: we had run Bitcoin's stop widths on an instrument that moves about a third more per bar. Same code, same year, one variable changed — and the result flipped.
What is coming
Regime Report — a monthly reading of six markets through Selepro Flow: how much of the time each one actually trends, how long its longest trending run has ever been, and what changed since the previous report. Descriptive only; it does not forecast anything.
Further notes in the “tested and discarded” series: stop widths as a multiple of ATR, regime-switched stops, efficiency-ratio thresholds, and a three-tranche exit. All four were measured, all four were rejected, and each has a number attached.