Pipsgrowth EX14022 Scalper

Pipsgrowth EX14022 GalaxyGoldScalper is built around a deliberately counter-intuitive read of two familiar M5 indicators on XAUUSD. Where most EMA-cross systems buy the upward cross, this EA buys the downward cross of a fast EMA(8) through a slow EMA(21) — but only when the 10-period Momentum oscillator on the same closed bar is still above 100. The premise is that in a healthy uptrend, a brief pullback that pushes the fast EMA back below the slow EMA is the entry, not the exit, provided the underlying momentum has not actually rolled over. The mirror setup applies on the short side: a fast EMA crossing back up through the slow EMA while Momentum is still below 100 is read as a "sell the rip" in a still-intact downtrend. This is the entire signal of the EA in one bar.

The order of operations inside OnTick is strict and reveals the EA's real priorities. The very first gate is volatility: the 14-period ATR (in raw price, not points) must read between InpVolatilityMin=3.0 and InpVolatilityMax=50.0. That window is calibrated for XAUUSD on M5 — a normal XAUUSD M5 ATR(14) sits somewhere between $3 and $15 in active sessions, so the 3.0 floor kills the dead Asian-range and pre-London flat, and the 50.0 ceiling is set high enough that only the most extreme spike days (CPI, NFP, FOMC) would get blocked. This is the EA's "is the market actually moving?" check, and it runs before the trend and momentum logic even looks at the bar.

The second gate is operational. A new-bar guard (if(curBar == lastBarTime) return;) ensures the EA makes exactly one decision per closed M5 bar, so it cannot fire twice on the same setup. The current spread on the symbol must be at or below InpMaxSpread=80 points (roughly 8 pips on a typical XAUUSD quote), and the position count under magic 22214022 must be below InpMaxTrades=2. Both checks run before any indicator copy, so on a high-spread rollover or once the EA is already holding two positions, the rest of the function short-circuits cleanly.

On a bar that passes all gates, OnTick pulls two bars of data from each of the four indicator handles — iATR(14), iMA(8, MODE_EMA, PRICE_CLOSE), iMA(21, MODE_EMA, PRICE_CLOSE), and iMomentum(10, PRICE_CLOSE). The buy condition then resolves to: previous bar's fast EMA strictly above previous bar's slow EMA, current bar's fast EMA at or below current bar's slow EMA, and the previous bar's momentum above the 100 baseline. The sell condition is the exact mirror. SL and TP are computed as InpStopLoss * _Point (400 points = 40 pips on a 5-digit XAUUSD quote) and InpTakeProfit * _Point (600 points = 60 pips), giving a 1.5:1 reward-to-risk ratio. Lot sizing is either the fixed InpLotSize=0.01 default, or — when InpAutoLot=true — CalcLot(sl) computes a balance- and contract-aware lot from balance * InpRiskPct / (sl_dist / tick_size * tick_value), floored at 0.01 and rounded to 2 decimals. Entries go through trade.Buy(lot, _Symbol, ask, ask-sl, ask+tp, InpTradeComment) with trade.SetDeviationInPoints(20) set in OnInit — a 20-point slippage tolerance the broker is allowed to apply to the fill.

There are a few things this EA does not do, and they are worth knowing before you load it. The header advertises a "12-layer architecture" that includes a News Filter — and the inputs do expose InpNewsFilter=true, InpNewsMinsBefore=30, InpNewsMinsAfter=30. None of those values are read anywhere in OnTick. The actual code contains no news event check, no calendar hook, no scheduled pause, and no recovery after a release. The News Filter is a declared input with no implementation behind it. If you need news avoidance, you need to build it outside this EA. Similarly, the header mentions OnTester as one of the 12 layers, but there is no OnTester function in the file. There is no trailing stop, no break-even ratchet, no partial close, no time-based exit, no reverse-signal exit, no equity stop, no global drawdown cap, no session filter, and no day-of-week guard. Once a position is opened, the EA's only ongoing interaction with it is the count-trades check in OnTick that respects InpMaxTrades. The position's life is otherwise left entirely to the broker-side SL and TP that were attached at entry.

Three more pieces of the source are template scaffolding. The functions TryClose_EX14022, TryClosePartial_EX14022, and TryModify_EX14022 are all declared with 3-attempt retry loops (200ms close, 200ms partial, 100ms modify back-offs on REQUOTE / TIMEOUT / PRICE_OFF / PRICE_CHANGED) but none of them is called from OnTick. Live entries go straight through trade.Buy / trade.Sell, so users will see single-attempt failures on requotes rather than the silent retry these helpers were designed to absorb. OnDeinit releases all four indicator handles cleanly, and OnInit returns INIT_FAILED if any of the four handles comes back INVALID_HANDLE. The 18 inputs split into four groups — Money Management, Strategy, News Filter, and Trade Settings — and every one of them is exposed to the user, which is the part of the EA that does live up to the marketing copy.

In practice, GalaxyGoldScalper wants a broker with tight spreads on XAUUSD (the 80-point max is a hard cutoff, not a target — anything that runs hot at the rollover will be skipped), a sub-$1.5 commission per lot per side if possible, and reliable execution during the London and New York sessions when the 8/21 EMA cross and 10-period Momentum combination actually has signal to work with. A $100 minimum deposit runs 0.01 lots comfortably given the 40-pip stop; the AutoLot path with 1% risk scales this up automatically on larger balances without any manual recalibration. Test in the Strategy Tester on real tick data from the session you intend to trade — the "flat Asia" filter is doing real work in the 3.0 ATR floor, and that's the part of the backtest most likely to look different in live.