Glandslam goldfleet

**GRANDSLAM GOLDFLEET — FOUR ENGINES. ONE GOLD FLEET.** A multi-engine Expert Advisor for GOLD (XAUUSD) on MetaTrader 5. Instead of depending on one trading model, GOLDFLEET combines four independent engines — **DARUMA × FUJI × KAGURA × KABUTO** — into one coordinated system with account-level protection (daily loss stop 3%, equity drawdown kill 12%). The objective is not the most spectacular backtest; it is to combine several different sources of trading activity into one systematic XAUUSD portfolio. ■The four engines: **DARUMA** = DEMON2 swing engine (five sub-strategies on RSI / Bollinger / CCI / Stochastic / MFI, one position each, SL/TP plus step trailing). **FUJI** = DemonScalp M5 engine (pullbacks and breakouts inside an efficiency-ratio trend, ATR-based stop and trail). **KAGURA** = Asia-box breakout: the 03:00–09:00 server-time range is traded once per day in the London breakout direction, only when volatility qualifies, with box-based SL/TP and a 22:00 close. **KABUTO** = previous-day high/low stop orders with a D1 SMA20 trend gate and a previous-day H1 volatility filter, SL 0.35 / TP 1.0 x ATR(D1), orders expire after 24 h. Each engine carries the exact GOLD settings validated in its stand-alone product, and each was verified trade-for-trade against that product inside the fleet. **No martingale, no grid, no averaging in any engine — every position has a stop-loss.** ■Reference result (GOLD, real ticks, Mar 2020 – Aug 2026, $10,000, 0.03 lot per engine): **2,241 trades, PF 1.45, +$10,521, equity DD 8.2%, Recovery Factor 8.1, in-sample PF 1.67 / out-of-sample PF 1.40** (OOS = May 2023 onward). Honest notes: 2022 was slightly negative (−$7); the PF is 1.45, not 5; the drawdown is 8.2%, not zero. ■Three engines vs four: at the same 0.03, net profit +$9,471 → +$10,521 (about +11%), RF 7.3 → 8.1, trades 2.7x — while PF fell from 1.65 to 1.45. We disclose this trade-off deliberately: the four-engine version was chosen for broader activity and portfolio characteristics, not because every metric improved. ■Lot frontier (four engines): 0.01 = 1,822 trades, PF 1.44, +$3,500, DD 3.7%, RF 7.5 / **0.03 = reference** / 0.05 = 2,226 trades, PF 1.44, +$17,074, DD 11.7%, RF 7.3. 0.05 made the most money but with more drawdown and a lower RF, so 0.03 per engine is the reference for a $10,000 test account. This is not individual investment advice; appropriate sizing depends on capital, leverage, contract specification, execution and risk tolerance. Included SET files: CONSERVATIVE (0.01), BALANCED (0.03, default), HIGHER_RISK (0.05). None of them is "safe" or "guaranteed". ■Evidence stages: the evidence that exists today is **backtest and out-of-sample**. Walk-forward, Monte Carlo, forward and live stages have not been run yet, and only stages with real evidence are ever shown as verified. Backtests are never presented as live results. ■Requirements and cautions: a hedging account is recommended (engines can hold opposite directions at the same time; up to about eight concurrent positions). Session-based engines assume a GMT+2/+3 server with US daylight-saving switching. Do not run the stand-alone DARUMA, FUJI, KAGURA or KABUTO on GOLD at the same time (they share magic numbers). Results vary with spread, commission, swap, tick data and execution — test on your own broker. ■How to use: attach to your broker's GOLD/XAUUSD chart (M5 recommended), load a SET file, enable algo trading. Built with Anthropic's Claude Code as a development partner under a research-first, disclose-everything policy. Past performance does not guarantee future results.