trader-bot

Crypto trading bot. Real Binance data, no API key needed for paper trading.

Strategy

• EMA crossover (9/21) with trend filter (EMA50)
• RSI(14) with divergence detection
• MACD histogram for direction strength
• ADX(14) market regime detection (trending vs ranging)
• Bollinger Bands squeeze for breakout anticipation
• Multi-timeframe filter (4h confirms 1h signals)
• ATR-based stops that adapt to volatility
• Progressive trailing stop (3 phases: breakeven → 1x ATR → 0.75x ATR)
• Circuit breaker pauses after 3 consecutive stop-losses
• Correlation limit caps altcoin exposure when BTC is held
• Cooldown prevents re-entering after a stop-loss

Fees (0.1% per side) are deducted from every trade.

Setup

pip install -r requirements.txt

Paper trading (no API key needed)

python main.py --limit 1000

With web dashboard

python main.py --limit 1000 --dashboard

Open http://localhost:8080

Dashboard includes:

• TradingView candlestick charts with volume
• Equity curve
• Trade history with filters
• Market regime indicator
• Circuit breaker status
• Progressive trailing stop phases

Backtesting

Test the strategy on historical data before risking capital:

python main.py --backtest --months 6 --limit 1000

Output includes: win rate, profit factor, Sharpe ratio, max drawdown, per-symbol breakdown.

Live trading

1. Create API key on Binance (Spot Trading only, never enable Withdrawals)
2. Copy .env.example to .env and fill your keys:

BINANCE_API_KEY=your_key_here
BINANCE_SECRET=your_secret_here

3. Run:

python main.py --limit 1000 --live --dashboard

Config

Edit config.yaml:

symbols:
  - BTC/USDT
  - ETH/USDT
  - SOL/USDT

timeframe: "1h"
htf: "4h"                     # higher timeframe confirmation
scan_interval: 60
max_positions: 5
risk_per_trade_usd: 15        # fixed $ risk per trade (ATR-based sizing)
max_correlated: 2             # max altcoins when BTC is held
circuit_breaker: 3            # consecutive SLs before 4h pause

# Regime detection
adx_trend_threshold: 25       # ADX > 25 = trending
adx_range_threshold: 20       # ADX < 20 = ranging

Stops are dynamic (ATR-based):

• SL = entry - 2x ATR
• TP = entry + 4x ATR (2:1 R/R)
• Trail Phase 1: at 1R profit → stop to breakeven
• Trail Phase 2: at 2R profit → trail at 1x ATR
• Trail Phase 3: at 4R profit → trail tight at 0.75x ATR

Structure

trader-bot/
  main.py           # entry point (paper/live/backtest)
  config.yaml       # settings
  .env              # API keys (live mode only)
  server.py         # dashboard API + WebSocket
  dashboard/
    index.html      # web UI (TradingView charts)
  bot/
    data.py         # Binance connection + orders + retry
    strategy.py     # signal engine (EMA+RSI+MACD+ADX+BB)
    risk.py         # ATR sizing + correlation + circuit breaker
    portfolio.py    # positions, progressive trailing stops
    tracker.py      # trade history
    engine.py       # main loop + multi-TF + reconciliation
    backtest.py     # backtesting engine

Disclaimer

No guaranteed profits. Always backtest and paper trade first. Don't risk money you can't lose.