Strategies

Each strategy is a self-contained strategy() script. It replicates the signal logic of one indicator and adds entry and exit management. There is no generator: a change to an indicator does not reach its strategy by itself, and pulling a strategy back in line is a deliberate step.

Scripts

Strategy	Based on	Stop
chandelier_flip_radar_strategy.pine	Chandelier Flip Radar	trailing, ATR envelope
commodity_pulse_matrix_v4_strategy.pine	Commodity Pulse Matrix v4	directional fixed SL/TP
market_average_relationship_engine_strategy.pine	Market–Average Relationship Engine	trailing, MA ∓ ATR multiple
oscillator_divergence_zones_strategy.pine	Oscillator Divergence Zones	pivot plus ATR buffer
reversal_engine_score_strategy.pine	Reversal Engine Score	structural, R-multiple target
smooth_trend_radar_strategy.pine	Smooth Trend Radar	fixed, TP1/TP2/TP3 levels
wavetrend_v4_strategy.pine	WaveTrend v4	trailing

What every wrapper adds

• Trade direction: both, long only or short only.
• Entries on the confirmed bar, on by default, so historical and realtime entries match.
• Cooldown in bars after an exit, and an optional break-even stop that moves the existing stop instead of adding a second exit.
• Filters: session, date range, maximum drawdown, maximum intraday loss, losing-streak limit.

Standard candles only

Every strategy draws a red warning label on Heikin Ashi, Renko, Kagi, Line Break, Point & Figure and Range charts. Orders there fill at synthetic bar prices, and the result says nothing about the market — it is reliably flattering and never reproducible.

Shared defaults

Setting	Value	Note
Position size	10 % of equity	change in the strategy’s properties
Commission	0.02 %	raise it for crypto maker/taker fees
Slippage	1 tick	raise it for illiquid instruments
Entries	confirmed bar	prevents repainting

Overnight and roll financing on CFDs is not priced in. On multi-day holds that is a known cost the Strategy Tester does not show.

Known limitations

• Entry timing: signals fire on the bar close and the strategy enters at the next bar’s open, which is the realistic case. Filling at the exact close is useful for comparison only.
• Pivot-anchored stops can sit several bars in the past. On illiquid instruments, check that the resulting distance still describes a tradable risk.
• Higher-timeframe pivots use the non-repainting pattern of request.security() with lookahead against an offset series. It needs enough history to warm up.

Testing a strategy

1. Run it on at least two instruments and two timeframes before drawing a conclusion.
2. Record the baseline with default settings. Fewer than 30 trades is not a result.
3. Change one input group at a time, measure again, keep or revert.
4. Validate on data outside the window you tuned on. A parameter set that only works where it was fitted has been fitted, not found.
5. Test long only and short only separately; an average can hide one healthy side and one broken one.

Backtest figures are kept in each strategy’s assessment file in the repository, together with instrument, timeframe and sample size. This documentation does not quote them.