TrendFollowingBot

A pragmatic trend-following strategy that pairs an EMA trend filter with two complementary entry patterns:

1. a crossback into the trend (price crosses the EMA and the very next bar opens on the trend side), and
2. a volatility-scaled breakout away from the EMA (the current open is beyond the EMA by k × ATR).

Stops, trailing logic, and position sizing come from your configured ExitStrategy (e.g., ATR-trailing), so this bot focuses purely on when to enter. Signals are read from the previous closed bar, and orders are placed at the current bar open to avoid look-ahead.

Overview

• What it does: Uses an EMA to define trend and ATR to demand “enough distance” before entering.
• Two entry modes:
  • Crossback confirms that price actually crossed the EMA and opened on the trend side.
  • ATR breakout requires the open to be at least k × ATR beyond the EMA (default k = 1.5).
• Risk first: Exits and sizing are handled by your exit module (e.g., ATR-trailing, fixed-ratio).
• Timing: Compute on bar t-1; place at the open of bar t (no look-ahead).
• Use cases: Clean template for experimenting with EMA spans, ATR multipliers, and different exit policies.

How it works (at a glance)

Let:

• ema_prev = EMA value on the previous (closed) bar
• prev_close = previous bar’s close
• entry_open = current bar open (execution price)
• atr_prev = previous bar’s ATR
• k = breakout_atr_mult (default 1.5)

1) Crossback into trend

• Long: prev_close ≤ ema_prev and entry_open > ema_prev → buy
• Short: prev_close ≥ ema_prev and entry_open < ema_prev → sell

2) EMA ± (k × ATR) breakout

• Long: entry_open > ema_prev + k × atr_prev → buy
• Short: entry_open < ema_prev − k × atr_prev → sell

If neither rule fires, the bot stays flat for that bar.

Key parameters

Open the notebook