BigWay Bot pilot run on our own internal strategy. Tested on 40 days of Binance Futures data, 25 liquid pairs. Strategy failed success criteria — paper deploy SKIPPED. This page exists as transparent proof of process.
📊 Backtest results
Trades
766
Win rate
10.7%
Profit Factor
0.20
Max DD
573%
PnL net
-$57,523
Sharpe (est)
-14.4
Criterion
Target
Result
Pass?
Profit Factor
> 1.5
0.20
❌
Sharpe Ratio
> 1.0
-14.4
❌
Max Drawdown
< 20%
573%
❌
Min Trades
> 100
766
✅
🛑 What this means
We do NOT proceed to paper deploy. A strategy with -575% return in backtest will lose money in live. Saving compute + saving from running a losing system.
No regime filter. Strategy entered все reclaims regardless of higher-timeframe trend.
Bull traps caught the strategy 85-90% of the time.
🔄 Tune attempt
We tried tightened parameters (volume 2.5x, TP 4×ATR, exit 24 bars). Result: still failed (PF 0.21, win rate 15.1%, DD 154%). Strategy concept is structurally vulnerable on this data, не a parameter problem.
📋 Lessons (grabли пройдены)
Grabля #1: Data period limitation. Spec asked 24mo; we had 40d. Real client needs longer history before backtest.
Grabля #2: Strategy looks good on paper, dies on data. LLM-designed spec was logical (extended downturn + reclaim = textbook setup). Real data: 10% win rate.
Grabля #3: Don't tune until it works (overfitting trap). After 2 failed variants → stop. Honest verdict more valuable than fake "we made it work."
Grabля #4: LLM bias toward bull market thinking. Real strategies need regime awareness baked in.
💡 The value of this exact gate
This pilot demonstrates exactly what BigWay Bot service buys you: a backtest gate before you commit live capital. The process here saved a hypothetical client from deploying a strategy that would lose 575% of equity.
If this were a paid pilot ($2-5k tier), the client would receive:
Honest backtest report (this page)
One tuning attempt (also documented)
Recommendation to pivot strategy OR refund
🛤️ BigWay Bot pipeline status
Phase
Status
0. Intake / strategy brief
✅ Done
1. Strategy spec generation (Pro)
✅ Done
2. Code (backtest engine)
✅ Done (reusable)
3. Backtest execution
✅ Done
4. Decision Gate
❌ NO-GO ← stopped here
5. Paper deploy
⏭️ SKIPPED (gate failed)
6. Live dashboard with auto-PnL
⏭️ SKIPPED
7. Hand-off package
📝 Replaced with verdict report (this page)
Pipeline ran end-to-end in ~90 minutes. Cost ~$0.50.
🤝 Want this for your strategy?
The BigWay Bot service gives you exactly this honest pipeline — strategy spec, code, backtest, decision gate. If your strategy passes backtest → paper deploy. If fails → honest verdict (like this) and recommendation.