Five years of logged breakout trades: most of them fail, and that's fine. The base rates.

I run a rules-based breakout system and log every trade. Base rates from
the 131 breakout entries in its locked five-year test window (2021 to
2026), because almost nobody posts these:

- 56.5% of entries lost money (74 of 131).
- The single most common outcome was a loss between 5 and 10 percent.
That's not a coincidence, it's where the volatility stop lives.
- Median trade: minus 4.0%.
- And the book still made money, decisively: mean +10.5% per trade,
average winner +34.2% against an average loser of minus 7.7%. Roughly
one trade in eleven ran +40% or more. No trade lost more than 20%.

What those numbers taught me about the craft:

1. The stop IS the strategy. A breakout entry is a cheap option on a
trend. The stop caps the premium. Mine is ATR-scaled and sits below
structure, not at a round number, which is why the losses pile up in
one narrow bucket instead of bleeding everywhere.

2. No profit targets. Every cap on the right tail attacks the only part
of the distribution that pays. Winners get a slow moving-average
trail, confirmed on the close so one bad intraday print doesn't shake
the position out.

3. Win rate is the wrong thing to optimize. A 43.5% hit rate with this
shape beats a 60% hit rate with a clipped tail. Chasing win rate
quietly caps your winners.

4. Regime does the heavy lifting. The same entries during weak breadth
just bleed. The system only trades breakouts when breadth is broad and
rising, and it stands down to cash entirely on roughly a third of all
days. Doing nothing is a position.

Honesty footnote: these are hypothetical backtest numbers, not live
results. I paper-trade the system in public and publish every signal and
every miss daily, because I think the industry's cherry-picked screenshots
are a plague. Nothing here is advice; it's just what a breakout
distribution actually looks like when it's logged honestly.

Happy to go deeper on the stop math or the regime gate in the comments.