Pipsgrowth EX16075 Trend

Pipsgrowth EX16075 Trend is a breakout EA that wraps price action around a Gaussian-smoothed midline and waits for a confirmed piercing move beyond the band. Rather than chasing fast M1 ticks or stacking indicators into a vote, it does one thing deliberately: it draws two ATR-width bands above and below a 10-period Gaussian filter on the working timeframe, then enters only when the close crosses outside the band on the same bar, with the band slope already pointing in the trade direction and a higher-timeframe EMA agreeing with the side. The result is a slower, more selective engine that prefers to miss a fast spike than to fade one.

The signal logic is built around a single indicator — an iMA(close, 0, MODE_EMA, 10) handle that the EA calls GaussianFilter() even though the underlying call is the standard EMA. The midline is what the EA actually treats as the noise-cleared price. Two bands are then placed at a fixed multiple of ATR(7): upper = gauss + 0.8 × ATR, lower = gauss − 0.8 × ATR. The DistanceMultiplier=0.8 input governs the band width. A long entry fires when close crosses above the upper band (close[0] > upper && close[1] < upper) AND the Gaussian slope is up (gauss[0] > gauss[1]) AND close[0] is above the H4 EMA(20) that the EA pulls via GetEMA_HTF(). The short side mirrors every condition. The MinBreakDistance=0.3 input is exposed but explicitly disabled in the OnTick code with a comment that the original extra distance check was removed — the band-cross is the only distance gate.

The higher-timeframe agreement is a meaningful design choice. The EA does not trust a band break by itself; it demands that the H4 EMA(20) sit on the correct side of price. That single filter is what turns the system from a generic Bollinger-band-piercing strategy into a directional breakout engine. It also explains why the default Timeframe input is 4 (M30): the bands are drawn on M30, the H4 EMA is sampled on a higher frame, and the EA waits for both to align. The TrendFilterTF input is exposed separately at 6 (H4) and is the timeframe from which the EMA is pulled, not the signal timeframe.

Filtering happens on three independent gates before any entry is considered. First, a volatility band gate: ATR(7) must be above MinATR=0.04 and below MaxATR=0.2 — a tiny ATR means the band is too narrow and breakouts are noise; a giant ATR means the band is so wide that almost any close crosses it. UseAutoFilter=true calls AutoTuneFilters() which scales MinATR to 1.5× the current spread in points, so the EA can adapt its dead-zone to the broker's spread regime. Second, a regime gate: with UseADXFilter=true, the ADX(7) main line must be at or above ADXTrendThreshold=20. A sub-20 ADX means price is ranging, and band breaks in a range revert. Third, a spread gate: a hard block at MaxSpreadPoints=12 — if the live spread is wider than 12 points, the EA prints a warning and refuses to enter on that tick. CheckMaxDrawdown() also runs on every tick and halts new entries once the open equity drawdown exceeds MaxDrawdownPercent=10%.

Position management is layered. The initial stop is ATR(7) × ATR_MultiplierSL=1.0 behind entry. The take-profit is a fixed FixedTP_Points=100 points. After the position is open, four management modules can intervene: (1) ExitOnGaussCross=true closes the trade the moment close crosses back across the Gaussian midline against the trade direction; (2) UseTrailingStop=true re-anchors the stop to ATR(7) × 1.0 behind the current bid/ask, but only forward — it never widens the stop; (3) UseBreakeven=true moves the stop to entry plus a 1-point slippage buffer once price has travelled 0.8×ATR in profit (BreakevenTriggerATR=0.8); (4) EnableDynamicLockProfit=false, when flipped on, ratchets the stop in $2 steps (LockProfitEvery_X_Dollars=2.0) with a $1 buffer (LockMinusBuffer=1.0), turning a $10 profit into roughly a $9 locked stop. Position sizing uses the standard equity-percent formula: lots = equity × 1% / (SL distance × value per point per lot), snapped to the symbol's volume step and clamped to volume min/max. UseFixedFractionalRisk=true with RiskPercent=1.0 is the default; the FixedLots=0.01 fallback fires only when risk calculation cannot produce a positive lot.

Trade management uses three local retry helpers — TryClose_EX16075, TryClosePartial_EX16075, and TryModify_EX16075 — each of which retries up to three times on requote, timeout, price-off, or price-changed, sleeping 100–200 ms between attempts. TryClosePartial_EX16075 is defined but never called from OnTick (it is dead code, present for future partial-close work). OneTradeOnly=true is the default; with MaxOpenTrades=1 the EA holds a single position per symbol. If the user sets MaxOpenTrades > 1, AllowOnlyProfitableAdditions gates new entries behind AllPositionsHaveMinProfit — every open position must show at least MinProfitPerTradeToAdd=$5 of profit before another layer is added. CloseOpposite() runs before every entry, so a flip in signal direction closes the existing trade and opens the new one in the same tick.

What to expect in backtest: the M30 signal bar combined with the H4 EMA filter is slow. On a quiet FX major or XAUUSD M5 chart, this EA will print a handful of trades a day, most of them in the direction of the dominant H4 trend. Band crosses that happen against the H4 EMA are filtered out silently, so the equity curve should be smoother than a pure band-piercing system but with more trades missed. The Gaussian-cross exit is a distinctive feature: it tends to close trades near the midline before the fixed 100-point TP is reached, which means the realised RR is usually below 1:1 — the trade is small and frequent, sized so that even sub-1R winners compound. Users who want longer holds should set ExitOnGaussCross=false and let the trail + fixed TP carry. AvoidNews=true is wired to a placeholder function that always returns false; the news filter is exposed in inputs but does nothing until the user implements a real calendar. For live deployment, a low-spread ECN broker is required because the spread gate at 12 points will block entries during wide-quote events on retail accounts.