Optimise deeper.
Validate harder.
Version 6.12 exposes more of Titan's inner engine to cTrader optimisation while adding a custom fitness function and stronger parameter validation.
Turn the black box into a testable system.
Periods
Local/trend EMA, MACD, RSI, Bollinger, ATR, ADX and Fibonacci lookback.
Signal weights
Fine-tune how strongly each vote contributes to the final score.
Thresholds
Score, score lead, RSI zones, ADX minimum, Fibonacci retracement bounds.
Initial exits
Stop floor, target floor, ATR stop multiplier and ATR target multiplier.
Management
Break-even trigger/offset, partial trigger/percentage, Runner start/target and trail geometry.
Safety
Recovery scaling, recovery score boost, near-floor risk usage and black-swan multipliers.
Optimise in layers, not chaos.
Entry selectivity
Start with Score Needed To Trade and Minimum Score Lead. Prove Titan can trade and identify stable trade-count regions before changing the whole engine.
Initial exits
Optimise stop floor, target floor, stop ATR multiplier, target ATR multiplier and realistic spread limits.
Core periods
Use small sensible ranges for local/trend EMAs and MACD periods. Huge ranges multiply the search space quickly.
Secondary filters and weights
Test RSI, Bollinger, ADX, Fibonacci and individual signal weights. Keep score thresholds consistent with the enabled maximum score.
Management
Test break-even, cost offset, partial trigger, Runner start/target and Runner/Protect trail geometry after the entry structure is stable.
Safety
Recovery and Black Swan settings should be validated after the core system is stable so protection does not hide signal problems.
Out-of-sample validation
Choose broad stable parameter plateaus, then test unseen dates and the actual broker feed. Never treat the single highest-profit historical pass as automatically superior.
Optimise for more than raw net profit.
Titan's custom fitness can reward profitable, better-sampled passes while penalising equity drawdown and rejecting passes that fail minimum-trade or drawdown requirements.
View included optimisation resources →fitness = log-scaled net profit + profit factor weight + win-rate contribution + average-trade contribution + sample-size contribution - equity drawdown penalty Reject when: trades < configured minimum or drawdown > configured maximum or non-positive profit (optional)