Pipsgrowth EX15015 SMC-OrderBlock

EX15015 (file: Pipsgrowth_com_EX15015_XAUUSD_Liquidity_Trap_EA_v3_3.mq5) is the v3.3 refresh of the EX15-family liquidity-trap scalper, and the new piece of logic in this build is the addition of a DEMA (Double Exponential Moving Average) crossover confirmation gate sitting on top of the engulfing, pin bar and RSI gates that already shipped in v3.1. So this is a 7-gate AND-confirmation strategy: 1 gate is always-on (price surge), 4 gates default-on (engulfing, pin bar, RSI, DEMA crossover) and 2 gates default-off (volume surge, higher-timeframe MA trend). All seven must read true before a buy or sell is fired. The brief in the file header — "Liquidity-trap scalper that detects swing-based liquidity zones, price surges and volume spikes, then enters on pin-bar / engulfing / RSI / DEMA-crossover confirmation with R:R SL/TP" — captures it exactly, including the twelve-layer claim of REGIME / SIGNAL / ENTRY / CONFIRM / NO-TRADE / CAPITAL CAP / RISK / SIZING / MANAGE / EXIT / SCALING / OnTester.

How a setup actually forms on the chart. IdentifySwingPoints walks the last ~20 bars and FindSwingHigh / FindSwingLow return the first index where high[i] strictly exceeds both high[i-barsEachSide] and high[i+barsEachSide] (mirrored for lows), with barsEachSide = SwingBarCount = 1 by default — so a true 3-bar fractal pivot. When a new swing is found and not already in the buffer, AddLiquidityZone stores the swing price plus the swing time, de-dups against existing zones at 0.3% proximity, and MarkLiquidityZone draws an OBJ_HLINE in clrCrimson (upper) or clrDodgerBlue (lower) with STYLE_DASH, named EA_LiqZone_Upper_ / EA_LiqZone_Lower_. Zones stay active for MaxZoneAge = 24 hours, and CleanupInactiveZones runs every 4 hours (TimeCurrent() % (3600*4) < 60) to free memory and remove the chart objects.

Once a zone exists, the buy path is CheckBuySignals, which iterates the lowerLiquidityZone array looking for the first zone where the current Ask is inside the 1% band [zoneLower, zoneLower * 1.01]. When that condition holds, all 7 confirmation functions get called and logged: CheckPriceSurge (MathMax of open-to-current percent over 10 bars and 5-bar high-low range, threshold PriceSurgePercent * 100 = 0.1% default), IsVolumeSurge (BYPASSED because RequireVolumeSurge = false by default), IsZoneConfirmedByHigherTimeframe (BYPASSED because UseHigherTimeframeConfirmation = false by default — falls through to a permissive true), IsBullishEngulfingPattern (3-bar lookback, current body engulfs prior opposite-color body), IsBullishPinBar (body/totalRange <= PinBarRatio = 0.3 AND lower shadow >= 2x body AND upper shadow <= 0.5x body), IsRSIOversold (rsiValues[0] <= RSIOversoldLevel = 30 from a 3-bar CopyBuffer) and IsDEMABullishCrossover (the v3.3 addition). Sell path is a mirror in CheckSellSignals with upper zones, bearish engulfing / shooting-star pin, RSI >= 70, and DEMA bearish crossover. After a successful fire the zone is deactivated (zoneActive[i] = false), so the same liquidity level can only trigger one trade per lifetime.

DEMA in this EA is not a native MQL5 indicator. CDEMAWrapper (lines 216-253) builds DEMA = 2 * EMA(length) - EMA(EMA(length)) by holding two iMA handles per wrapper: m_handleEMA and m_handleEMAofEMA. demaFast.Init(..., 8) and demaSlow.Init(..., 21) create both wrappers, and CopyBuffer reads the inner EMA first, the EMA-of-EMA second, then returns the DEMA buffer. The IsDEMABullishCrossover condition is intentionally loose: it returns true on a fresh crossover (fast[0] > slow[0] AND fast[1] <= slow[1]) OR whenever fast[0] > slow[0] — i.e., the fast DEMA is currently above the slow DEMA. That is more of a trend-alignment check than a strict crossover detector, and it makes the DEMA gate weaker than the engulfing / pin bar / RSI gates it sits next to. With DEMAFastPeriod = 8 and DEMASlowPeriod = 21, the wrapper reconstructs DEMA(8) and DEMA(21) on the trading timeframe, releasing both internal handles via demaFast.Release() / demaSlow.Release() in OnDeinit.

Sizing is fixed and non-risk-based. CalculatePositionSize(stopLossPoints) explicitly ignores the stopLossPoints argument and returns NormalizeLotSize(FixedLotSize), which is 0.01 by default and gets clamped to the symbol's minLot / maxLot / lotStep range. SL placement for a buy is zoneLower * (1 - 0.005) — half a percent past the zone on the downside — and TP is currentPrice + (currentPrice - stopLoss) * RewardToRiskRatio. With RewardToRiskRatio = 2.0 that is a clean 1:2 R:R regardless of how far price has run from the zone to entry. Sell path mirrors with stopLoss = zoneUpper * (1 + 0.005) and takeProfit = currentPrice - (stopLoss - currentPrice) * RewardToRiskRatio. Orders are sent via Trade.Buy / Trade.Sell with Trade.SetDeviationInPoints(10) for slippage, and a single position counts against MaxPositions = 5 per EA instance.

Trailing stop is now actually wired. ManageOpenPositions walks all open positions, applies the trailing only when EnableTrailingStop = true, and ApplyTrailingStop computes trailLevel = bid - currentProfit * TrailingStopPercent / 100 for buys and ask + currentProfit * TrailingStopPercent / 100 for sells (default 50% of profit, so half the open profit is locked once the trade is in green). Modify goes through TryModify_EX15015 (3 attempts, 100ms backoff on REQUOTE / TIMEOUT). That is a real improvement over the EX15013 cousin where ApplyTrailingStop was defined but not called. There is still no time-based exit, no opposite-signal close, and no partial close during the trade life — positions exit exclusively at the static SL or TP, or get ratcheted by the trailing.

Risk and session gates. The per-tick spread filter rejects the bar at spread > MaxSpreadPoints * point = 30 points (3 pips on a 5-digit XAUUSD quote). MaxConsecutiveLosses = 3 in UpdateTradeStatistics triggers pauseUntilTime = TimeCurrent() + PauseMinutes * 60 = 60 minutes, and the dashboard's Status line goes red for that window. EnableTimeFilter is OFF by default but accepts HH:MM strings (08:00-20:00 broker time when enabled). AvoidHighImpactNews = true is the default and IsHighImpactNewsTime is implemented as a hardcoded single event at 14:00 same day with NewsBufferMinutesBefore / After = 30, so unless you repopulate highImpactNewsTimes the news filter is functionally dormant. The 5-attempt progressive-delay retry in DetermineTrend (100ms, 200ms, 300ms, 400ms) and the 5x MapTimeframeInt redefinition in the source (lines 33, 56, 80, 115, 142 — code quality issue, not a logic bug) are inherited from the EX15 family.

Version-string inconsistency worth flagging. #property version says "2.00", the #property description line advertises "v3.3", the OnInit Print and OnDeinit Print say "v3.30", the UpdateDashboard title writes "XAUUSD Liquidity Trap EA v3.2", and the dashboard stat line uses the same v3.2 string. Treat the v3.3 description as authoritative for behavior because the 4 default-on gates match v3.3 in every other way. Magic 22215015, comment string Psgrowth.com Expert_15015, and slippage 10 points are the only stable identifiers across all five version labels.

What to expect when running it. On XAUUSD M30 with the 4 default-on gates, the EA will fire only when all of price surge >= 0.1%, an engulfing body covers the prior opposite body within 3 bars, a hammer / shooting-star pattern prints within 3 bars (PinBarRatio = 0.3), RSI(14) sits at the extreme (>= 70 for sells, <= 30 for buys), and DEMA(8) is at-or-above DEMA(21) for buys (or at-or-below for sells). That is a strict set of conditions and the trade frequency will be modest. With FixedLotSize = 0.01 and RewardToRiskRatio = 2.0, a winning streak of three R multiples yields 6% on a $1,000 account, a losing streak of three losses yields -3% before the 60-minute pause kicks in. Recommended setup: $100 minimum, ECN / RAW-spread broker with sub-30-point typical spread on XAUUSD, broker time set to GMT+2 or GMT+3, and keep UseDEMAConfirmation enabled — it is the v3.3 piece of the design and the simplest way to see whether the v3.3 vs v3.1 difference actually changes your trade log.