Pipsgrowth EX16021 Trend

Pipsgrowth EX16021 reads three consecutive candles as a single geometric formation rather than trusting any standard indicator stack. The strategy is built around two of the oldest reversal patterns in Japanese candlestick theory — Three Black Crows at tops and Three White Soldiers at bottoms — but the EA defines them with a tighter set of rules than a textbook would. To qualify, each of the last three bars must have a body that exceeds the 12-bar average body size (InpAverBodyPeriod = 12) AND the midpoint of each bar (defined as (High + Low) / 2 inside the MidPoint() helper) must step strictly in the direction of the pattern. A black candle must open above its close, a white candle must close above its open, and the midpoints have to cascade downward for the bearish three crows or upward for the bullish three soldiers. That cascade rule is the heart of the strategy: it filters out noisy three-bar sequences that look bearish in body size but don't actually step lower in their range midpoints, which is a common failure mode of naive candlestick recognition.

The candlestick geometry is only half of the entry decision. The other half comes from the MFI oscillator configured at period 37 with VOLUME_TICK applied volume (InpPeriodMFI = 37, InpVolume = VOLUME_TICK). The 37-period MFI is a deliberately slow choice — most chart packages default to 14, so the EA is reading money flow over more than a full trading day's worth of bars on M5. A buy signal from Three White Soldiers only gets executed when the MFI(1) value is below 40, meaning money flow is depressed enough to make the bullish reversal plausible from an oversold baseline. A sell signal from Three Black Crows only gets executed when MFI(1) is above 60, requiring elevated money flow to support the bearish reversal. This confirmation asymmetry — buys need depressed flow, sells need elevated flow — is the EA's way of forcing the oscillator to agree with the price geometry before any order goes to the broker.

Once a pattern is detected and confirmed, PositionOpen() fires a market order at the current Ask (for buys) or Bid (for sells) with the configured fixed lot (InpLot = 0.10) and the configured 200-point stop loss and 200-point take profit (InpSL = 200, InpTP = 200). The 1:1 risk-reward ratio at fixed pips means the EA expects to lose about as often as it wins on a per-trade basis and relies on a hit rate above 50% (or favorable money flow timing) to come out ahead. There is a useful safety net inside PositionOpen(): if the current spread is wider than the configured stop loss — for example, if the broker is quoting a 250-point spread while InpSL is 200 — the EA overrides the stop to a single spread's distance from the entry price and logs a warning. The same override applies to the take profit. Without this guard, a buy order with a 200-point stop during a 250-point spread would be rejected by the broker, and the trade would simply not happen. With the guard, the trade goes through at a degraded stop distance rather than being skipped entirely.

Exits in EX16021 come from three independent sources, evaluated in order. The first is an MFI cross-signal in CheckCloseSignal(): a long position is closed if MFI(1) drops below 70 while MFI(2) was above 70 (overbought-to-cooling cross) OR if MFI(1) drops below 30 while MFI(2) was above 30 (oversold confirmation). A short position is closed on the mirror: MFI(1) above 30 with MFI(2) below 30, or MFI(1) above 70 with MFI(2) below 70. This MFI cross logic uses the same indicator and the same 70/30 thresholds but applies them to the close decision rather than the entry decision — the same oscillator is doing double duty as both gate and trigger. The second exit source is the hard time stop (InpDuration = 10): the BarsHold() helper copies MqlRates from the position's open time to TimeCurrent() and returns the number of bars in between, and CloseByTime() force-closes any position that has held for 10 or more bars. The third exit source is the static 200/200 SL/TP, which catches anything that the MFI cross and the time stop both missed. There is no trailing stop, no break-even, no partial close, and no scaling in or out — the only position management is the trio of MFI cross, time stop, and fixed pips.

A point worth flagging: the header comment claims a 12-layer architecture (REGIME, SIGNAL, ENTRY, CONFIRM, NO-TRADE, CAPITAL CAP, RISK, SIZING, MANAGE, EXIT, SCALING, OnTester), but only seven of those layers are actually wired into the code. SIGNAL (the candlestick pattern) is wired. ENTRY (market order) is wired. CONFIRM (MFI gate) is wired. NO-TRADE (the new-bar guard via the static next_bar_open datetime in OnTick) is wired. RISK (200-point SL) is wired. SIZING (fixed 0.10 lot) is wired. EXIT (MFI cross + time + SL/TP) is wired. What is NOT wired: REGIME detection (no volatility filter, no session filter, no trend filter — the body-size check inside CheckPattern() is the only proxy), CAPITAL CAP (no daily loss limit, no equity drawdown stop, no max-position guard), MANAGE (no trailing, no break-even, no partial), SCALING (no pyramiding, no martingale, no grid), and OnTester (no custom optimization criterion — the EA does not export a tester fitness function). Three retry helpers are defined at the bottom of the file — TryClose_EX16021, TryClosePartial_EX16021, and TryModify_EX16021 — each attempting the trade action up to three times with 100–200ms sleeps on REQUOTE/TIMEOUT/PRICE_OFF/PRICE_CHANGED. TryClose_EX16021 is wired into CloseBySignal() and CloseByTime(); TryClosePartial and TryModify are dead code that is never called from any active function.

EX16021 carries magic 22216021, which makes it a sibling to EX16020 (magic 22216020). The two EAs share the same strategy, the same fixed 0.10 lot, the same 200/200 SL/TP, the same 10-bar time stop, and the same MFI(37) confirmation thresholds, but they are independent trade streams because of the unique magic. A practical deployment is to run both on the same XAUUSD M5 chart on the same account: because the magic number filters positions, EX16021 will only manage its own trades and will not touch EX16020's open positions or vice versa, so the combined exposure is two concurrent trades when both fire on the same bar. The default timeframe is M5 and the alternative the header recommends is H1; the same source code runs on both because the candlestick definitions are timeframe-agnostic and the MFI(37) period scales naturally. The minimum recommended balance is $100 USD (the InpLot of 0.10 lot on XAUUSD implies a notional of roughly 1,000 USD at XAUUSD = 2,000, so a $100 account runs 10:1 notional leverage, which is high but workable for a 1:1 risk-reward scalper). The risk level is classified MEDIUM because the fixed 0.10 lot and the 1:1 risk-reward create volatility around the equity curve, even though the strategy is mechanically simple.

What to expect in backtest: a relatively small number of trades per week (the three-candle cascade is a strict filter and the MFI gate eliminates more than half of the patterns that form), with a hit rate that needs to stay above 50% for the 1:1 risk-reward to produce a positive expectancy. The 10-bar time stop means the EA never holds through a weekend or through major news, which removes gap risk as a failure mode. The spread-overrides-stop safety means the worst-case loss on a single trade in extreme spread conditions is one full spread, not a blown-through stop. The OnTick function has four phases: Phase 1 calls CheckState() at the new bar to set ExtPatternDetected, ExtConfirmed, ExtSignalOpen, ExtSignalClose, and ExtPatternInfo; Phase 2 opens a position if there is a signal and no position in that direction; Phase 3 closes a position if there is a close signal; Phase 4 force-closes any position that has held for 10 or more bars. The new-bar guard (the static datetime next_bar_open inside OnTick) ensures the EA does not act on the same signal multiple times within a single bar, which is critical because the MFI and pattern readings can flicker on intra-bar ticks.

If you are running this EA, two configuration decisions matter. First, pick a broker with low spread on XAUUSD because the spread-overrides-stop safety will degrade the 200-point stop distance to whatever the spread is whenever the market widens — a 50-point spread broker will preserve most of the 200-point stop, while a 250-point spread broker will collapse the stop to 250 points on most trades. Second, decide whether you want to run the MFI confirmation enabled or disabled by leaving the InpPeriodMFI and InpVolume inputs at their defaults — the defaults (37 and VOLUME_TICK) are the values the strategy was designed around, and changing them will change the entry frequency and the win rate without any guidance from the source on what the new optimum looks like. The EA does not log trade results in a custom file and does not push notifications — only Print() statements to the terminal — so plan to read the MT5 journal to see what happened on any given day.