L-Shape Long Strategy Note

Strategy Overview

Name: Enhanced L-Shape Long Strategy
Asset: ETHUSDT Perpetual Futures
Exchange: Binance Futures
Version: 1.0
Date: 2026-04-25

Pattern Definition

L-Shape 패턴은 급락 후 횡보 구간을 거쳐 MA 돌파 시 롱 진입하는 패턴이다.

     ┌──────┐
     │ HIGH │ ← 이전 고점
     └──┬───┘
        │
        ▼ ATR 기반 하락 (drop_atr_mult × ATR)
        │
   ┌────┴────────────────────┐
   │  횡보 구간 (consolidation) │ ← ATR 기반 범위
   │  range < consol_atr_mult × ATR │
   └─────────────────┬───────┘
                     │
                     ▲ MA 돌파 → LONG 진입

Entry Conditions (Enhanced Rules)

1. 선행 하락: 최근 20봉 내 drop_atr_mult × ATR(14) 이상 하락
2. 횡보 확인: 최근 N봉 범위가 consol_atr_mult × ATR(14) 이내
3. MA 돌파: 이전 봉 종가 < MA, 현재 봉 종가 > MA + 양봉

Optimal ATR Multipliers

Parameter	Value	Description
drop_atr_mult	2.5	하락 감지 임계값
consol_atr_mult	1.5	횡보 범위 임계값

Exit Conditions

• Take Profit: 진입가 +15%
• Stop Loss: 진입가 -2%
• Half Close: 진입가 +5% 도달 시 50% 익절 + SL을 본절로 이동

Strategy Profiles

5M Balanced (Recommended)

Parameter	Value
Timeframe	5m
Position Size	20%
Leverage	10x
Take Profit	15%
Stop Loss	2%
Half Close	5%

Backtest Results (2021-2025, with fees):

• Return: 510.9%
• CAGR: 33.4%
• Max DD: 32.0%
• Calmar: 1.05
• Win Rate: 45.5%
• Trades: 334
• Avg Holding: 12.7h

5M Aggressive

Parameter	Value
Position Size	30%
Leverage	10x

Backtest Results:

• Return: 1,012.0%
• CAGR: 46.8%
• Max DD: 46.3%
• Calmar: 1.01

15M Aggressive

Parameter	Value
Timeframe	15m
Position Size	30%
Leverage	10x

Backtest Results:

• Return: 1,862.4%
• CAGR: 60.7%
• Max DD: 69.9%
• Calmar: 0.87

Fee Structure

Fee Type	Rate
Maker	0.02%
Taker	0.05%
Funding (avg)	0.01% / 8h

5M 전략이 최적인 이유: 평균 보유 시간 12.7시간으로 펀딩 비용 최소화

Risk Management

1. Position Sizing: Kelly Criterion 기반
   • Full Kelly: 15.9%
   • Half Kelly (권장): 8.0%
   • 실제 적용: 20% (보수적 버퍼)
2. Stop Loss: 2% (타이트)
3. Half Close: 5% 도달 시 50% 익절 → 본절 SL

Implementation Status

Backtest ✅

• Basic backtest engine
• Parametric study (MA, TP, SL, HC)
• Enhanced rules (ATR-based detection)
• Realistic fees simulation
• Multi-timeframe comparison

Live Trading ✅

• Go engine with Binance API
• Enhanced L-shape detector (ATR-based)
• 3 strategy profiles
• Svelte frontend (Dashboard/History/Settings)
• F/E → B/E static serving integration
• Testnet validation

Validation Checklist

• Backtest CAGR > 30%
• Max DD < 50% (5M Balanced: 32%)
• Calmar > 1.0 (5M Balanced: 1.05)
• Win Rate > 40%
• Sufficient trades (>300)
• Testnet paper trading (1 week)
• Production deployment

Files

• Backtest engine: src/crypto_backtest/long_strategy.py
• Parametric study: scripts/run_realistic_backtest.py
• Go engine: live-trading/engine/
• Svelte frontend: live-trading/frontend/
• Results page: docs/realistic-backtest.md

References

• Enhanced Strategy
• Realistic Backtest Results