20260115_060001_bitfinex_tDOGE_price_relative_to_mu.md

# Discovery Run: 20260115_060001

**Exchange:** bitfinex
**Symbol:** tDOGE
**Focus:** price relative to multiple EMAs
**Started:** Thu Jan 15 06:00:01 AM CET 2026

---

## Summary

I've created a new trading strategy: **DOGE EMA Cascade Momentum**

**Strategy File:** `/var/lib/scriptrunner/staging/doge_ema_cascade_momentum.py`

**Concept:**
A trend-following strategy that uses a strict EMA cascade filter (EMA20 > EMA50 > EMA100 > EMA200) to only trade during strong uptrends. It enters on pullbacks to EMA20 and exits when the trend structure breaks.

**Training Results (2024-01-01 to 2025-06-30):**
- 2024: +243.9% | 28 trades | 32% win rate
- 2025 (partial): +4.8% | 6 trades | 33% win rate
- **Total: +248.8% | 2/2 years profitable**
- Max drawdown: ~37%

**Key Design Decisions:**
1. **Strict regime filter** - Full EMA cascade required (20>50>100>200) keeps strategy flat during 54% of the period that was bearish
2. **Pullback entry** - Only enters after price dips to EMA20, avoids chasing
3. **5% stop loss** - Caps individual trade losses
4. **Standard parameters** - Uses round EMA periods (20, 50, 100, 200), no magic numbers
5. **Role: momentum** - Strategy is expected to lose bounded amounts in bear markets, which is acceptable per validation rules

**Robustness Testing:**
Tested multiple parameter variations (stop loss 4-6%, pullback window 3-10 bars, threshold 1-3%) - all configurations remained profitable in both training years.

The strategy is now in staging and ready for automated validation on the unseen 2025-H2 data.

---
**Completed:** Thu Jan 15 06:06:16 AM CET 2026