Our Backtest Results: 63.9% Win Rate Across 1,394 Trades
Every claim on UnxEdge is backed by data you can inspect. The scanner logic was backtested across 200+ symbols on the 30 minute timeframe over twelve months before any of it went live. This page summarizes what we tested, what worked, and what did not. The live numbers update daily on the results page.
Methodology
The test rules are mechanical, so the results are repeatable:
- Timeframe. 30 minute bars.
- Universe. 200+ liquid US stocks and ETFs.
- Entry. At the breakout level, after the level is exceeded on volume.
- Exit. At the first profit target or the stop, whichever comes first.
A trade counts as a win if it reaches at least the first target before the stop. No discretionary overrides, no curve fitting after the fact.
Why we skip the first bar
One rule made an outsized difference: do not enter on the opening bar. The opening 30 minutes is the noisiest part of the day, full of gaps and reversals that have nothing to do with the pattern.
In testing, opening-bar entries showed negative expectancy. Waiting for the first bar to close before acting turned a losing rule into a profitable one. The scanner and the bots both wait.
Overall results
Across the full sample:
- 63.9% win rate across 1,394 trades.
- 1.78x profit factor (dollars won divided by dollars lost).
- +0.281R average per trade.
A profit factor above 1.3 indicates a real edge, and the sample is large enough that luck is not a reasonable explanation. See R-multiples and profit factor for why the average R matters more than the win rate.
Top performing symbols (the 100% Club)
A handful of symbols never lost a backtested trade with five or more trades. We call them the 100% Club: GOOGL, PANW, V, GDX, INTC, ZM, LULU, COIN, and SMCI.
When one of these appears in the scanner approaching a breakout, Arxe sends a gold alert. Each symbol has its own trade-by-trade scorecard at /ticker/SYMBOL, so you can see the full history before you act.
Sweet spot results
The strongest pattern in the data is the sweet spot: pressure below 60 with a three to five bar consolidation. Those setups produced a 64% win rate, well above the 63.9% baseline.
This is the setup the Top Picks filter is built around. The full breakdown is in the sweet spot guide.
What does not work
Being honest about the losers is how you trust the winners. Two patterns underperformed badly:
- Two-bar consolidations. 37% win rate. Too short to be a real base, so the breakout has nothing behind it.
- RVOL 3.0+ at entry. 39.6% win rate. By the time volume has tripled, the move is usually already extended and you are buying the top of the first push.
Both are filtered down in the grading and in the bot rules. Knowing what to avoid is half the edge.
See the live results
Backtests describe the past. The results page shows the present: live scanner resolutions, the bots scaling out of real positions, and the running scoreboard updated daily. Nothing is cherry-picked. The misses are published next to the wins.
Frequently asked questions
What is the UnxEdge backtest win rate?
The scanner backtest shows a 63.9% win rate, a 1.78x profit factor, and a +0.281R average across 1,394 trades on 200+ symbols over twelve months.
Why does UnxEdge skip the first bar?
The opening 30 minutes is the noisiest part of the day. In testing, opening-bar entries had negative expectancy, so the scanner and the bots wait for the first bar to close before acting.
Which symbols perform best in the backtest?
The 100% Club never lost a backtested trade with five or more trades: GOOGL, PANW, V, GDX, INTC, ZM, LULU, COIN, and SMCI.
What setups should I avoid?
Two-bar consolidations won only 37% of the time, and entries with RVOL above 3.0 won only 39.6%. Both are too late or too thin, and both are filtered down in the grading.
See these patterns live
The scanner watches 200+ symbols and flags every setup as it forms, with a grade, a pressure score, and a full trade plan.
Open the live scanner