Pipsgrowth EX16048 Trend

Pipsgrowth EX16048 sits at the intersection of classical Japanese reversal patterns and a single-oscillator money-flow filter. Each new bar the EA inspects the most recent three candles through CheckPattern() and looks for one of four shapes: Evening Doji, Evening Star, Morning Doji, or Morning Star. The geometry of each is identical to the textbook definition — a tall bull or bear bar (body larger than the 12-bar average body, InpAverBodyPeriod=12), a doji or near-doji in the middle (body less than 10% of the average body for Doji patterns, less than 50% for Star patterns), a price gap that flips direction, and a confirmation bar closing beyond the middle candle on the reversal side. Where the morning/evening star framework ends, the MFI takes over. The indicator handle is built in OnInit() with iMFI(_Symbol, _Period, 37, VOLUME_TICK) and stored in ExtIndicatorHandle. The 37-period MFI uses tick volume as its raw input, so the oscillator blends price movement with participation data rather than pure price deltas. A 12-bar body average is the only candle-side reference value the EA computes through AvgBody().

CheckPattern() returns the pattern and an ExtSignalOpen (SIGNAL_BUY or SIGNAL_SELL) without committing to a trade. CheckConfirmation() is the gate: it pulls MFI(1), the most recent completed bar's MFI value, and only marks ExtConfirmed=true if MFI(1) sits below 40 for a buy (oversold zone) or above 60 for a sell (overbought zone). If the pattern is detected but the MFI is on the wrong side of the threshold, the EA deliberately downgrades ExtSignalOpen to SIGNAL_NOT and skips the trade for that bar. The pattern's geometric setup is necessary but not sufficient — a Morning Doji in a tape where money is still flowing into the rally is treated as a pattern, but not a signal. The strategy is therefore a four-pattern detection engine wrapped in a single-oscillator vote, and the indicator's role is asymmetric confirmation: it doesn't have to agree with the pattern's direction, but it has to agree with the bar's exhaustion level.

The exit side mirrors the entry philosophy. When no entry signal is present, CheckCloseSignal() reads two MFI bars and watches for a cross of the 70 or 30 line from the opposite side. A long position is closed on a downward MFI cross of 70 (MFI(1)<70 AND MFI(2)>70) or an upward MFI cross of 30 (MFI(1)<30 AND MFI(2)>30). A short position closes on the mirror. These thresholds are tighter than the entry thresholds (60/40) by 10 points, which is a deliberate reinforcement — the EA exits on a less extreme oscillator move than it required for entry, so a position can close out before the oscillator fully reverts to neutral. A 10-bar time-based stop (InpDuration=10) closes any position that has neither been stopped out, taken profit, nor flipped by the MFI cross signal. BarsHold() counts completed bars since the position's open_time by reading the MqlRates history and is the only function that touches the time dimension beyond the static datetime next_bar_open guard at the top of OnTick.

Risk control on the open side is unusually defensive. PositionOpen() in EX16048 implements a spread-fallback safety net that the broker's stops would not normally offer. If the current spread, measured as ExtSymbolInfo.Ask() − ExtSymbolInfo.Bid() and expressed in points, is greater than or equal to InpSL (default 200 points) for the buy side, the EA bypasses the configured stop distance and sets the stoploss to price − spread. The same logic applies symmetrically to the take-profit side: if the spread is wider than InpTP, the TP is collapsed to a one-spread distance from the entry. The reasoning is to prevent the position from being stopped out by the spread the instant it opens on a wide market. A buy order uses ExtSymbolInfo.Ask() and a sell order uses ExtSymbolInfo.Bid(); both are NormalizeDouble'd to the symbol's digits. There is no trailing stop, no break-even logic, and no partial-close path: positions go from open to either SL, TP, MFI cross, or 10-bar time exit. The EA only ever holds one position at a time in any direction (PositionExist() is checked before every open), so there is no scaling-in or grid logic.

On the risk side, EX16048 ships with a fixed 0.1 lot (InpLot=0.1) and no percentage-of-equity or risk-per-trade calculation. Risk per trade on a 200-point stop at 0.1 lot is therefore fixed in dollar terms against whatever symbol it is attached to — on XAUUSD at typical point size this is in the single-digit-dollar range; on FX majors at the same 200 points it is roughly 20 USD per pip times the lot's per-pip value. The magic 22216048 (InpMagicNumber) and the comment string Psgrowth.com Expert_16048 (InpTradeComment) tag every ticket so the EA only ever reacts to its own positions during the close sweep in CloseBySignal() and the time sweep in CloseByTime(). Slippage tolerance is 10 points (InpSlippage), set through ExtTrade.SetDeviationInPoints() at init. There is no equity stop, no daily loss cap, and no maximum drawdown gate at the EA level; account-level protection is the responsibility of the platform.

The indicator stack is deliberately minimal. MFI is the only native handle, and its period 37 is on the slower side of the typical MFI 10-20 default, which dampens noise on M5 bars at the cost of a few more bars of confirmation lag. The 12-bar body average (InpAverBodyPeriod) for the pattern geometry is also on the slower end — most star/doji references in MQL books use 5-10 bars. The combination means the EA is conservative on both axes: it asks for a well-formed reversal pattern against a longer body baseline, and then asks the MFI to agree that the bar in question is in an exhaustion zone. Backtest expectations: this is a low-frequency, low-drawdown system. XAUUSD M5 will produce single-digit signals per week on quiet tape and a couple of clusters when volatility expands. The 1:1 risk/reward (200/200) means a 50% win rate is breakeven after spread, and the time stop will close marginal trades that drift. Run it on FX majors on H1 as a sanity check — the M5/H1 timeframe pairing in the EA's published symbol list points to FX Majors, Gold/Metals as the portable target. The 12-layer claim in the file header is aspirational: the actually-wired layers are SIGNAL (the four pattern checks), CONFIRM (MFI threshold), NO-TRADE (new-bar guard + single-position), RISK+SIZING (fixed lot), ENTRY (market order), EXIT (SL/TP/10-bar time/MFI cross). OnTester, REGIME detection, capital cap, partial-close, scaling, and trailing are declared in the header but not implemented in this build.

What to expect in a live session: the EA prints "Pattern not detected" on most new bars and only logs a signal line when a confirmation bar is also present. The trade comment makes it easy to filter the strategy's own history in the terminal's trade tab. Because PositionExist() blocks new entries while a position in that direction is open, the EA can sit in cash for long stretches. Time-based exits mean an old position cannot hold through a weekend unless the broker's rollover or expiration rules permit it; the 10-bar time stop is the upper bound on duration regardless of broker.