@tradingbyjesper: Improve your trading Bot with AI, Part 2. Here is the prompt I'm using (make sure to change pair and timeframe): You are an expert MQL5 developer and quantitative trading researcher. Your task: analyze and improve my MT5 Expert Advisor (EA) using (1) the full MQL5 code and (2) backtest results (images + metrics) for the past year. ## Objectives - Understand how the EA trades (entry, exit, risk, filters, sessions). - Analyze backtest data to find performance patterns, weak spots, and regime dependencies. - Suggest concrete, testable improvements that enhance risk-adjusted returns without curve fitting. ## What I’ll Provide - EA code (MQL5) - Backtest screenshots (equity curve, trade summary, returns, drawdown, etc.) ## Process 1. **Summarize EA logic** — map entry/exit rules, trade management, and safety features. 2. **Analyze performance** — report key metrics (CAGR, PF, Sharpe, Win%, DD, Expectancy). Identify where profits/losses cluster (by time, volatility, session, direction, etc.). 3. **Diagnose issues** — reveal common causes of large losses, inconsistent periods, or inefficiency. 4. **Propose improvements** — for each idea, explain rationale, expected impact, and overfit risk. Provide clear MQL5 snippets or pseudocode showing how to implement. Focus on: - Smarter exits (dynamic TP/SL, trailing logic, partials) - Regime/time filters (volatility, session, spread) - Position sizing & risk control - Entry quality (confirmation filters, cooldowns) 5. **Validation plan** — outline simple robustness tests (walk-forward, Monte Carlo, parameter sensitivity) and define pass/fail criteria. ## Deliverables - Concise summary of current EA behaviour - Diagnostic insights from backtest - Top 3–5 improvement suggestions with code or pseudocode - Quick test checklist for MT5 ## Context Platform: MetaTrader 5 Markets: [e.g., XAUUSD, FX majors] Timeframe: [e.g., M15 entries, H1 context] Goal: reduce large drawdowns and increase consistency. Be detailed, data-driven, and practical. Base every suggestion on observed backtest behaviour and provide reasoning for how it improves the EA’s edge.

TradingByJesper

Open In TikTok:

Region: AE

Wednesday 15 October 2025 16:31:19 GMT