Pipsgrowth EX14011 Scalper

Pipsgrowth EX14011 is a closed-bar, single-position Gold scalper that fires only when the Williams Percent Range(14) and the Relative Strength Index(14) sit on the same side of their extreme zones on the same M5 bar. The whole entry decision is the AND-gate between two oscillators — there is no regime filter, no trend overlay, no confirmation candle, no news blackout, and no multi-timeframe scan. When both indicators agree that price is stretched, the EA opens one market order in that direction with a hardcoded 300-point stop and a hardcoded 500-point target, then waits for the trade to close before the next signal is even considered.

The bar timing sits in OnTick. The handler refreshes the symbol, checks mysymbol.Spread() > InpMaxSpread (default 250 points, roughly 25 pips on a 5-digit Gold quote) as a hard no-trade gate, and then compares the current bar's iTime against a static lastBar timestamp. If the timestamp matches, OnTick returns immediately — meaning every evaluation, every indicator read, every lot calculation only happens once per fresh M5 bar. Inside CheckSignal the EA first checks PositionsTotal() > 0 and returns if anything is open, so this is strictly a one-position-at-a-time design: the EA will never pyramid, never hedge, never stack a grid, and never enter a second trade while the first is still running.

The signal logic in CheckSignal reads the previous bar's WPR(14) and RSI(14) values via CopyBuffer(hWPR, 0, 1, 1, bufWPR) and CopyBuffer(hRSI, 0, 1, 1, bufRSI). A buy fires only when wpr < -80 AND rsi < 30 — both oscillators saying oversold at the same time. A sell fires only when wpr > -20 AND rsi > 70 — both saying overbought. There is no in-between trigger: if WPR says oversold and RSI says neutral, nothing happens. This AND-conjunction is the entire entry engine. It is also why the EA is so sparse — two handles, two threshold pairs, one buy side, one sell side.

Lot sizing runs through CalculateLotSize(InpStopLoss). When InpRiskPercent > 0 (default 1.0), the function computes riskMoney = account.Balance() * InpRiskPercent / 100.0, then solves lot = riskMoney / riskPerLot where riskPerLot = stopLossPoints * (tickValue / (tickSize / pointVal)). That is a textbook stop-distance risk formula: the broker's tick value and tick size are used to translate the 300-point stop into the per-lot dollar exposure for XAUUSD, and the lot is rounded DOWN to the symbol's LotsStep, then clamped to LotsMin/LotsMax. When InpRiskPercent = 0 the function returns the fixed fallback InpFixedLot = 0.01. There is no compounding, no balance-based scaling beyond the 1% rule, and no per-trade risk override.

Order execution is delegated to the standard CTrade wrapper. OpenBuy sets the SL to Ask - 300 * Point and the TP to Ask + 500 * Point, then calls trade.Buy(lot, _Symbol, mysymbol.Ask(), sl, tp, InpTradeComment). OpenSell mirrors with Bid + 300 * Point and Bid - 500 * Point and trade.Sell(...). trade.SetExpertMagicNumber(22214011) ties every fill to that magic so the EA can identify its own orders in a multi-EA account. trade.SetTypeFilling(ORDER_FILLING_FOK) is hardcoded at OnInit, which is correct for most MT5 brokers that allow Fill-Or-Kill on Gold.

What the EA does NOT do is just as important as what it does. There is no trailing stop, no breakeven move, no partial close, no time-based exit, no equity-protection kill switch, no daily loss limit, no drawdown cap, no margin guard, no news filter, no session filter, no magic-per-symbol guard beyond the magic number, and no OnTester pass. The header advertises a 12-layer stack (REGIME, SIGNAL, ENTRY, CONFIRM, NO-TRADE, CAPITAL CAP, RISK, SIZING, MANAGE, EXIT, SCALING, OnTester) but the code wires only the SIGNAL, ENTRY, NO-TRADE, RISK, and SIZING layers — everything in the MANAGE / EXIT / SCALING / OnTester / REGIME / CONFIRM / CAPITAL CAP section is marketing copy. The file also defines three retry helpers — TryClose_EX14011, TryClosePartial_EX14011, TryModify_EX14011 — that each loop 3 times on REQUOTE / TIMEOUT / PRICE_OFF / PRICE_CHANGED with a 200ms sleep, but none of them are wired into a caller inside the EA. They are dead code, left from a heavier version of the template that this simpler Gold-scalper variant was forked from.

The retry helpers' existence is the strongest tell about the EA's design intent: the author was probably going to add a manage-position pass that calls TryModify_EX14011 to ratchet the stop, and a partial-close pass that calls TryClosePartial_EX14011 to scale out 50% at 1R and let the rest run to TP. None of that is active. The actual position lifecycle is: open with SL/TP set, walk away, let the broker hit one of the two prices, then PositionsTotal() drops to 0 and the EA is allowed to look for the next signal on the next fresh bar.

In backtest this means the equity curve will be a sequence of fixed 5:3 reward-to-risk trades (500 TP / 300 SL) interspersed with skipped bars where one oscillator failed to confirm. The win rate required to break even on a 5:3 bracket is 37.5%; anything above that is profit. Because WPR and RSI are correlated but not identical, the AND-gate filters out a meaningful slice of marginal signals — the trade count will be lower than a single-oscillator system, but the per-trade quality should be higher. The flip side is that during trends that take Gold deep into overbought or oversold for many bars in a row, this EA will keep taking counter-trend entries at the extremes; the lack of a regime filter is by design but is the single biggest risk to know about before going live.

Practically: the EA is best understood as a small, auditable, indicator-confluence prototype. The entire strategy fits in CheckSignal (12 lines of actual logic), the entire money management fits in CalculateLotSize (12 lines), and the entire execution fits in OpenBuy / OpenSell (5 lines each). If you want a Gold scalper you can read end-to-end in two minutes and modify with confidence, this is exactly that. If you want a finished, multi-layered production system, this is the starting point, not the destination — the retry helpers and the 12-layer header are the roadmap for what was intended next.