Pipsgrowth EX15016 SMC-OrderBlock

EX15016 is an XAUUSD scalper built around one design decision: instead of holding positions to a fixed take-profit and hoping, the EA actively ratchets the stop-loss upward as a trade moves into profit, using a dynamic profit-lock system that becomes more aggressive the further price travels in your favor. The version-numbered file ships as v3.5.1, but the source contains three different version strings — #property version 2.00, the description line that says v3.5.1, the OnInit print that says v3.5.1, and the dashboard title that says v3.0 — so treat the build as v3.5.1 (the most recent of the labels) and ignore the dashboard string.

The trading idea starts with swing liquidity zones. On every new bar the EA scans the last Max(20, SwingBarCount*2+5) bars for 3-bar fractals — FindSwingHigh walks the high array and returns the first index where high[i] > high[i-1] AND high[i] > high[i+1] for SwingBarCount=1 bar on each side; FindSwingLow does the mirror. Every fresh swing high becomes an upper liquidity zone, every fresh swing low becomes a lower zone, each one drawn on the chart as a clrCrimson or clrDodgerBlue OBJ_HLINE with STYLE_DASH and a tooltip. The AddLiquidityZone helper de-duplicates new candidates against existing zones at a 0.3% price tolerance, so the chart accumulates a sparse map of recent swing pivots rather than a wall of overlapping lines. Zones auto-deactivate after MaxZoneAge=24 hours, and CleanupInactiveZones runs every four hours to compact the arrays.

Once a zone is on the chart, the entry check is a 5-condition AND gate. For buys, the EA looks at every active lower zone and asks whether the current Ask is inside that zone's 1% band — zoneLower <= currentPrice <= zoneLower * 1.01. If price is in the band, five conditions are evaluated: (1) CheckPriceSurge returns true when the MathMax of the open-to-current percentage over the last 10 bars AND the 5-bar high-low range percentage exceeds PriceSurgePercent*100 = 0.1%; (2) IsVolumeSurge returns true when current tick volume is at least VolumeSurgeMultiplier=1.5x the 20-bar average — but the function short-circuits to true because RequireVolumeSurge=false by default, so volume is currently a dormant gate; (3) IsZoneConfirmedByHigherTimeframe calls DetermineTrend(ConfirmationTimeframe) to read the MA(20) vs MA(50) cross on H4 by default — but this also short-circuits to true because UseHigherTimeframeConfirmation=false, so the higher-timeframe filter is dormant; (4) the price-surge condition is the only always-on filter beyond zone containment; (5) IsBullishEngulfingPattern walks back EngulfingLookback=3 bars looking for a current bullish candle whose body completely engulfs a prior bearish body's range — currentOpen <= previousClose AND currentClose >= previousOpen — and is on by default. The mirror sell path checks upper zones with the inverted band [zoneUpper*0.99, zoneUpper] and a bearish engulfing. In practice the EA is running as a price-surge + engulfing + zone-proximity strategy with two dormant confirmation toggles available if you want to tighten it.

The stop-loss and take-profit geometry is fixed by the zone, not by the indicator state. For a buy, the SL is zoneLower * 0.995 — half a percent below the lower band — and the TP is computed as a clean 1:2 reward-to-risk: takeProfit = currentPrice + (currentPrice - stopLoss) * RewardToRiskRatio. With RewardToRiskRatio=2.0 default, a 1% zone SL gives a 2% target. Sells mirror the geometry on the upper zone. Lot sizing is the simplest part of the system: CalculatePositionSize ignores the stop-loss-points argument entirely and returns FixedLotSize=0.01 after NormalizeLotSize clamps it to the broker's min/max and lot-step. This is a fixed-lot EA — not a risk-percent one — so position sizing scales only by changing the input, not by equity.

The dynamic profit-lock system is the reason this EA exists as a separate file from the rest of the EX15 family. On every tick ManageOpenPositions walks every open position with this EA's magic and calls ApplyDynamicProfitLock(ticket). The function first reads the entry price, current SL/TP, and current Bid (for buys) or Ask (for sells), then calculates profitPoints = (currentPrice - entryPrice) / point. It then derives currentLockLevel = floor(profitPoints / LockProfitEvery_X_Points), which is floor(profitPoints / 30) with the default. Whenever the detected level exceeds the stored lockLevelCount[posIndex], the EA computes a new stop-loss that locks in profit: newSL = entryPrice + (currentLockLevel * 30 - buffer) * point for buys (mirror for sells). The buffer is LockMinusBuffer=15 points by default, but if EnableProgressiveBuffer=true, CalculateProgressiveBuffer multiplies the buffer by ProgressiveBufferRatio=0.8 once per lock level — so level 1 sees a 15-point buffer, level 2 sees 12, level 3 sees ~9.6, level 4 sees ~7.7, and so on, with a hard floor of 1 point. The ratchet is one-directional: a buy only updates SL when newSL > currentSL (or current SL is 0); a sell only updates when newSL < currentSL. If the new SL is too close to the current price relative to the broker's SYMBOL_TRADE_STOPS_LEVEL, the modify is skipped. The actual PositionModify call goes through a fresh local CTrade tradeMod with SetDeviationInPoints(10) — it does not go through the main Trade object and does not use the TryModify_EX15016 retry helper. Every successful lock event increments totalLockEvents and adds the locked points to totalProfitLocked; both numbers are reported in the OnDeinit print.

The partial-close path is a separate, opt-in layer wired into the same lock-step. When EnablePartialClose=true and the position has not yet had a partial close, the EA computes partialCloseDone[posIndex] = true and tries to close PartialClosePercent=50% of the position via a fresh CTrade tradePartial with the same magic and 10-point deviation. With the default config the partial close is disabled and never fires; turning it on lets the EA book half the position at the first lock level and then continue managing the remainder with the ratchet. The standard trailing stop is a third, parallel management path: when EnableTrailingStop=true the ApplyTrailingStop function fires from ManageOpenPositions and uses the TryModify_EX15016 retry helper (3 attempts, progressive 100ms/200ms/300ms backoff on REQUOTE/TIMEOUT/PRICE_CHANGED) to move the SL to entryPrice + profitPoints * TrailingStopPercent/100 — by default 50% of the unlocked profit. This is a separate code path from the dynamic lock: if you enable the trailing stop and the dynamic lock at the same time, both will try to push the SL, and whichever fires last with a higher SL wins.

Risk controls are short. A per-tick spread filter rejects any tick where SymbolInfo.Spread() * point > MaxSpreadPoints * point (3 pips default). On every close, UpdateTradeStatistics increments consecutiveLosses on a loss and resets it to zero on a win; once consecutiveLosses >= MaxConsecutiveLosses=3, pauseUntilTime = TimeCurrent() + PauseMinutes * 60 blocks all new entries for an hour. MaxPositions=5 caps concurrent exposure. EnableTimeFilter=false by default — set to true to gate entries between TradingHoursStart='08:00' and TradingHoursEnd='20:00' broker time. The two inputs AvoidHighImpactNews=true and the news-buffer minutes are decorative: there is no IsHighImpactNewsTime function in the source, no news feed integration, and the filter is functionally dormant.

Code-quality notes worth knowing about: the MapTimeframeInt helper is declared six separate times in the source (lines 35, 58, 82, 106, 133, 163), each with the same switch on tf 1-7 — the duplicate declarations compile because of how MQL5 handles identical signatures, but they are dead-weight clutter. The header advertises a 12-layer architecture (REGIME, SIGNAL, ENTRY, CONFIRM, NO-TRADE, CAPITAL CAP, RISK, SIZING, MANAGE, EXIT, SCALING, OnTester) but the source actually only implements roughly seven of those layers — there is no formal regime classifier, no capital-cap engine, and no OnTester fitness function. The UpdateLiquidityZones call is gated to new bars; the position-management call runs every tick. The dashboard is a 7-line label panel (title / status / lot+RR / zone counts / trade stats / profit factor / current signal) with a separate UpdateProfitLockDashboard section that shows locked levels per ticket — ShowProfitDashboard=true by default.

For backtests expect: 1-4 trades per session depending on how often price tags the recent swing zones, a 1:2 reward-to-risk on the static TP leg, and a ratcheting SL on the dynamic leg that turns roughly half of all trades into breakeven-or-better exits. The fixed 0.01 lot keeps the dollar risk identical regardless of stop distance — if you want the EA to scale position size to account, multiply FixedLotSize manually when changing symbols. This EA is for traders who specifically want a level-ratcheting profit-lock on a swing-zone reversal framework and are willing to accept the two dormant confirmation gates as configured.