--- title: "Strategy Discovery and New Edge Research - Examples - MCP and AI - MetaTrader 5 Hilfe" description: "Breakout edge: Mean-reversion conditions: Trend persistence: Multi-timeframe edge: New edge prototype: Path-signature pairs... - Strategy Discovery and New Edge Research - Examples - MCP and AI" image: "https://www.metatrader5.com/i/logo_metatrader5.png" url: "https://www.metatrader5.com/de/terminal/help/practical_examples/strategy_discovery" --- [MetaTrader 5 Hilfe](https://www.metatrader5.com/de/terminal/help) → [MCP and AI](https://www.metatrader5.com/de/terminal/help/mcp_and_ai) → [Examples](https://www.metatrader5.com/de/terminal/help/mcp_and_ai/practical_examples) → Strategy Discovery and New Edge Research # Strategy Discovery and New Edge Research Using AI with MetaTrader > Adapt the symbols, periods, position sizes, risk limits, and any criteria to your approach or preferences. Carefully study every prompt before using it and review any proposed action before executing it. > > You are solely responsible for AI-generated content, AI-assisted actions, trading decisions, trading results, and any resulting losses. Breakout edge: Analyze USDJPY H1 data from the last six months for breakout setups. Compare range length, volatility compression, breakout size, tick volume, trend alignment, and session, then identify the simplest conditions with the strongest forward results. Mean-reversion conditions: Analyze USDJPY H1 data from the last six months for extended moves. Compare volatility, trend strength, distance from recent value, and session to identify when mean reversion has positive expectancy and when continuation is more likely. Trend persistence: Analyze EURUSD H1 and H4 data from the last six months. Measure how long strong moves persist after different confirmations, comparing trend strength, pullback depth, volatility, and timeframe alignment. Multi-timeframe edge: Using EURUSD D1, H4, and H1 data from the last six months, test recurring trend, momentum, volatility, and structure combinations. Identify states that occur often enough and produce favorable forward returns. New edge prototype: Use GBPUSD H1 data from the last six months to define one recurring pattern with precise entry and exit rules. Create and compile a prototype EA, run a six-month backtest, and compare the result with the researched edge. Path-signature pairs edge: For NAS100 and US500 CFDs, use the latest five trading days of M1 data and 20-bar windows to calculate path signatures of order two. Use the first 70% of windows to define interaction and direction filters and compare their results with ordinary spread z-scores on the remaining 30%. Directional-change overshoot law: Convert the latest 5,000 EURUSD ticks into directional-change and overshoot events at thresholds of 0.05%, 0.10%, and 0.20%. Compare median overshoot size and duration across the three thresholds and define one rule from their common normalized relationship. Matrix-profile motif edge: For GBPUSD and its latest 500 M15 bars, build a matrix profile over normalized 20-bar price and tick-volume subsequences. Inspect the five strongest motifs and three largest discords, then compare their next four-bar returns and verify whether each discord is valid data. Transfer-entropy lead network: For EURUSD, GBPUSD, USDJPY, XAUUSD, and US500, use the last 10 trading days of M15 returns to estimate transfer entropy at one- and two-bar lags. Test each link with 50 shuffled surrogates, retain the three strongest significant links, and compare them with ordinary lagged correlation on the last two trading days. Ordinal-entropy state transitions: For EURUSD and its latest 300 H1 returns, encode three-bar ordinal patterns and calculate permutation entropy in 20-bar windows. Divide entropy into three fixed states, estimate their transition table from the first 240 bars, and test transition-triggered next-four-bar returns on the latest 60. Multifractal asymmetry edge: Using the latest 500 XAUUSD H1 returns, estimate simplified multifractal detrended-fluctuation statistics for positive and negative returns at scales of 8, 16, 32, and 64 bars and moments of -2, -1, 1, and 2. Compare the first and second 250-bar halves and check whether asymmetry predicts the latest half better than realized volatility. Capital-gains-overhang proxy: For AAPL, MSFT, NVDA, AMZN, and META and their latest 250 D1 price-volume bars, approximate each investor reference price with a decaying volume-weighted cost basis. Compare capital-gain overhang with 60-day momentum and test both measures against the latest 20-day return after estimated spread. Overnight-intraday decomposition: For AAPL, NVDA, and TSLA and their latest 120 trading days, decompose returns into close-to-open and open-to-close components. Compare next-session returns after top- and bottom-quartile moves and report whether either component shows reversal or continuation that is absent from close-to-close returns. Topological regime recurrence: Using EURUSD, XAUUSD, US500, and NAS100 and their latest 120 D1 returns, form six non-overlapping 20-day point clouds. At five fixed distance thresholds, summarize connected components and loops, group the first four clouds by similar topology, and check whether either of the latest two clouds has matching correlation and return behavior. Eigenportfolio residual edge: For EURUSD, GBPUSD, USDJPY, XAUUSD, US500, and NAS100 and their latest 250 D1 returns, fit three principal components on the first 200 bars and calculate residuals without refitting. On the latest 50 bars, compare next-five-day returns after residuals exceed plus or minus two standard deviations and include estimated spread. ## In this section - [Quick Analysis for Market Entry](https://www.metatrader5.com/de/terminal/help/practical_examples/market_entry_analysis) - [Risk Management Operations](https://www.metatrader5.com/de/terminal/help/practical_examples/risk_management_operations) - [Account and Behavioral Analysis](https://www.metatrader5.com/de/terminal/help/practical_examples/account_behavioral_analysis) - [Quick Workspace Preparation](https://www.metatrader5.com/de/terminal/help/practical_examples/workspace_preparation) - [Backtesting and Optimization](https://www.metatrader5.com/de/terminal/help/practical_examples/backtesting_optimization) - [Reverse Engineering of Strategy](https://www.metatrader5.com/de/terminal/help/practical_examples/strategy_reverse_engineering) - Strategy Discovery and New Edge Research - [Scenario Planning and What-If Analysis](https://www.metatrader5.com/de/terminal/help/practical_examples/scenario_planning) ```json {"@context": "https://schema.org", "@type": "BreadcrumbList", "itemListElement": [ {"@type": "ListItem", "position": 1, "item": {"@id": "https://www.metatrader5.com/de/terminal/help", "name": "MetaTrader 5 Hilfe"}}, {"@type": "ListItem", "position": 2, "item": {"@id": "https://www.metatrader5.com/de/terminal/help/mcp_and_ai", "name": "MCP and AI"}}, {"@type": "ListItem", "position": 3, "item": {"@id": "https://www.metatrader5.com/de/terminal/help/mcp_and_ai/practical_examples", "name": "Examples"}}, {"@type": "ListItem", "position": 4, "item": {"@id": "https://www.metatrader5.com/de/terminal/help/practical_examples/strategy_discovery", "name": "Strategy Discovery and New Edge Research"}} ]} { "@context": "https://schema.org", "@type": "FAQPage", "mainEntity": [{ "@type": "Question", "name": "MetaTrader 5: Strategy Discovery and New Edge Research", "acceptedAnswer": { "@type": "Answer", "text": "Breakout edge: Mean-reversion conditions: Trend persistence: Multi-timeframe edge: New edge prototype: Path-signature pairs edge..." } }] } ```