Add methodology documentation for EA-119 strategy

Document the methodology for EA-119 Three Candle Momentum strategy, including strategy identification, signal definition, baseline test configuration, metrics, assessment, research protocol, reporting requirements, and current status.
2026-10-19 13:12:00 +07:00
parent bbcf2411ab
commit bd890d978c
1 changed files with 84 additions and 0 deletions
@@ -0,0 +1,84 @@
+## EA-119 — Three Candle Momentum
+
+### Strategy Identification
+
+- **EA:** `EA-119_Three_Candle_Momentum`
+- **Baseline ID:** `EA119-M1-BASELINE-001`
+- **Strategy Type:** Short-term price momentum
+- **Baseline Symbol:** XAUUSD.PRO
+- **Baseline Timeframe:** M1
+
+### Signal Definition
+
+The EA evaluates the closing prices of three completed candles when a new bar forms.
+
+- BUY when `Close[1] > Close[2] > Close[3]`.
+- SELL when `Close[1] < Close[2] < Close[3]`.
+
+The implementation uses the chart's current timeframe. Reproduce the baseline using M1.
+
+### Baseline Test Configuration
+
+- Period: 2026-01-02 to 2026-04-01
+- Initial deposit: $1,000
+- Leverage: 1:500
+- Lot size: 0.01
+- History quality: 100% real ticks
+- Stop Loss: 300 points
+- Take Profit: 600 points
+- Break-even: enabled, trigger 150 points, offset 0
+- Trailing stop: enabled, start 200 points, distance 200 points
+- Maximum spread: 30 points
+
+### Baseline Metrics
+
+| Metric | Result |
+|---|---:|
+| Net Profit | -$992.47 |
+| Profit Factor | 0.71 |
+| Expected Payoff | -$0.37 |
+| Sharpe Ratio | -5.00 |
+| Maximum Equity Drawdown | 99.27% |
+| Total Trades | 2,650 |
+
+### Assessment
+
+The baseline is classified as `FAIL`. It exhibits negative profitability metrics and extremely high equity drawdown. It is not validated for live trading.
+
+The baseline must remain unchanged as the reference for subsequent experiments. A failed baseline must not be presented as a successful strategy merely because individual parameter combinations perform better.
+
+### Research and Validation Protocol
+
+1. Verify signal implementation and order execution.
+2. Evaluate signal length and momentum strength separately.
+3. Test entry confirmation rules in controlled experiments.
+4. Analyze BUY and SELL performance separately.
+5. Evaluate exit management only with clearly documented settings.
+6. Assess session and market-regime filters.
+7. Validate selected candidates out of sample.
+8. Conduct robustness and walk-forward testing.
+9. Require forward-test evidence before any live-trading decision.
+
+### Reporting Requirements
+
+Each experiment must record:
+
+- Experiment ID and EA version
+- Exact input parameters
+- Symbol, timeframe, and test period
+- Deposit, leverage, and tick-model settings
+- Net profit, Profit Factor, expected payoff, and drawdown
+- Trade count and directional results
+- Differences from the baseline
+- Decision and rationale
+
+### Current Status
+
+- Baseline: `COMPLETE`
+- Baseline Result: `FAIL`
+- Controlled Research: `REQUIRED`
+- Broad Optimization: `BLOCKED`
+- Out-of-Sample: `NOT STARTED`
+- Walk-Forward: `NOT STARTED`
+- Forward Test: `NOT STARTED`
+- Live Trading: `NOT APPROVED`