Pipsgrowth EX12008 MultiIndicatorConfluence

EX12008 is a multi-symbol trend-confirmation EA built on a three-stage agreement: a custom Andean Oscillator signal, a fast/slow moving-average regime, and a price-proximity filter around the fast MA. Rather than fire on a single crossover, it waits for the AOS bull (or bear) line to cross above its smoothed signal while the fast 50-period MA sits above the slow 200-period MA, and while the current ask price is within half the MA gap of the fast line. That last clause is what makes the strategy unusual: the EA does not chase breakouts, it waits for the market to come back to the fast MA before committing. Each symbol is scanned on a fixed 120-second timer, not every tick, which both throttles server load and prevents the kind of overtrading that pure tick-driven EAs suffer from.

The Andean Oscillator is implemented in-house rather than called as a separate indicator. The CAndeanOscillator class maintains four smoothed arrays — Up1, Up2, Dn1, Dn2 — and computes Bull as the square root of (Dn2 − Dn1²) and Bear as the square root of (Up2 − Up1²). A signal line is then produced as an EMA of the larger of the two. Inputs AosPeriod (default 50) and AosSignalPeriod (default 9) control the smoothing windows; both are tweakable from the panel. The 4-clause entry then becomes: AOS bull on bar 1 above its signal AND AOS bull on bar 1 above AOS bear AND fast MA on bar 1 above slow MA on bar 1 AND ask price within 0.5 × |fma1 − sma1| of the fast MA. The sell side is the mirror. All four clauses must be true on bar 1 of the symbol's timeframe; if any one fails, the EA skips that symbol and moves to the next in the rotation.

The EA is multi-symbol by default. Drop the file on any M5 chart, set MultipleSymbol = true, and the Symbols string "EURUSD, EURCAD, USDCAD" gets parsed into a rotation list. Each cycle walks the list and applies the same logic. Per-symbol protections include MinPosInterval = 6 (no new position on the same symbol for six candles after the last deal) and a SpreadLimit that, if enabled, blocks entries when the bid/ask gap exceeds the configured point count. hasDealRecently() enforces the cool-down; ea.OPTotal(s) prevents stacking positions on a single symbol even if the entry logic fires repeatedly. The combination means the EA will hold at most one position per symbol at any given time, but it can be long EURUSD, short EURCAD, and flat USDCAD simultaneously.

Stop-loss placement is swing-based, not fixed-pip. The SLType = SL_SWING input tells the EA to look back SLLookback = 10 bars for the most recent swing high (for sells) or swing low (for buys), then offset the SL by SLDev = 100 points. This adapts the stop to current volatility — wider in choppy markets, tighter in trending ones. The 100-point buffer prevents the stop from being parked exactly on a candle wick, where it would be tagged by noise. IgnoreSL = false is the default, so the SL is honored; set it to true and the EA closes positions only via trail, TP, or signal reversal. Take-profit is calculated as TPCoef × |entry − SL|, with the default TPCoef = 1.0 producing a 1:1 reward-to-risk ratio. However, IgnoreTP = true ships as the default, meaning the EA in its stock configuration is managed via trail, not TP — most positions are expected to close on the trailing stop or on a reverse AOS signal.

Trailing is the workhorse of the exit stack. Trail = true and TrailingStopLevel = 50 (interpreted as 0.5, since the EA multiplies by 0.01 in OnInit) engage an ATR-aware trail that ratchets the stop forward as price moves in the trade's favor. EquityDrawdownLimit = 0 keeps the global drawdown kill-switch off by default; turning it on (e.g. 30 for 30%) makes the EA flatten all positions and refuse new entries once equity drops that far from the high-water mark. A Grid mode is implemented but disabled by default; flipping it on activates a martingale-style layering with GridVolMult = 1.3 (each grid level is 1.3× the previous lot) and GridMaxLvl = 20 capping the depth. News filtering via the calendar is also off by default; NewsMinsBefore = 60 / NewsMinsAfter = 60 define a one-hour blackout window around medium-importance events when turned on.

Each entry attempt is followed by a 5-second sleep. The Sleep(5000) calls after BuyOpen and SellOpen are deliberate — they prevent the broker's order-acceptance throttle from rejecting rapid-fire requests when several symbols in the rotation fire on the same timer tick. Filling mode defaults to FILLING_DEFAULT; the EA inherits whatever the broker supports. Slippage is hardcoded at 30 points. The magic number 22212008 is used on every order; combined with the InpTradeComment of "Psgrowth.com Expert_12008", this isolates EX12008 from any other EAs or manual orders on the same account.

Risk sizing uses Risk = 5.5 (interpreted as 5.5% in RISK_DEFAULT mode), which is aggressive for a multi-symbol strategy — three symbols open simultaneously at 5.5% risk each is 16.5% account exposure per signal cluster. The GerEA framework handles the lot calculation from the swing SL distance. For accounts that want smaller exposure, drop Risk to 1.0–2.0 and the same SL placement will scale the position size down proportionally. Pair the 5.5% default with the recommended $100 minimum deposit and the EA will run on a micro lot, but a $500+ deposit is the honest starting point if all three symbols start firing on the same day.

The EA was designed for M5 primarily and works M5–H4 according to the source header. The five-minute bar is short enough to catch the AOS cross + MA proximity in real time, and the 120-second timer interval aligns naturally with it. Higher timeframes (H1, H4) reduce signal frequency but keep the same logic intact — the AOS class recalculates on whatever timeframe the chart is set to. For backtesting, NewsStartYear = 0 disables historical news fetching; if you want to test the news filter behavior, set it to a year before your backtest start.

A practical backtest expectation: on EURUSD M5 over 2022–2024, expect roughly 200–400 trades per year with a moderate win rate. The 1:1 R:R plus trailing means wins are smaller than typical trend-followers, but the loss rate is also lower. The strategy underperforms during high-volatility news spikes (where the 4-clause filter fails to fire on time) and during extended ranging periods (where MA proximity whipsaws). It does best in trending sessions where the fast MA pulls away from the slow MA and price retraces cleanly to it — exactly the condition the third clause of the entry is designed to detect.