Pipsgrowth EX18061 TrendFollow

The Pipsgrowth EX18061 is a single-position, ATR-anchored trend follower that anchors its entire risk architecture on a 1:2 reward-to-risk envelope. Every signal flows through the same constraint: if the math cannot justify a TP that is at least twice the stop distance, the trade is rejected before it ever reaches the order ticket. That single filter defines the character of the system — slow, deliberate, and ruthlessly selective.

The signal is built from two inline Hull moving averages, fast at 34 bars and slow at 89 bars, both computed on close prices. There is no iCustom call, no external indicator file: the Hull math is hand-rolled inside HullValue() using weighted moving averages from WMA(). A 34-bar Hull first derives a half-period WMA at 17 bars, subtracts the full 34-bar WMA, doubles the half, then takes a square-root-period WMA of the resulting series. The slow 89-bar Hull runs the same construction with a 45/89/9 split. The signal function then compares the value at bar 1 (the last closed bar) against bar 2 to determine color: +1 if the curve is rising, -1 if falling, 0 if flat. A trade is only valid when the fast and slow Hulls share the same color, and that color is non-zero. To filter out single-bar whipsaws, HullSignal() demands a confirmation count across HULL_CONFIRM_BARS (set to 2): the fast Hull must agree with its color on both of the two most recent closed bars. The returned confidence climbs from 50% to 100% as the agreement count grows.

A seven-state regime filter runs in front of the signal. DetectRegime() reads ADX(14), ATR(14), and Bollinger Bands(20, 2.0) and reduces the market to one of seven states: STRONG_TREND when ADX clears InpAdxMinTrend (20.0), WEAK_TREND between 18.0 and 20.0, RANGE otherwise. Volatility overrides then apply. If the current ATR sits in the 80th percentile of the last 100 closed bars and is rising, the regime is reclassified as EXPAND when ADX is at or above 22, or BREAKOUT otherwise. ATR in the bottom 20th percentile pushes the state to COMPRESS, as does a Bollinger width contraction of more than 15% bar-over-bar. Width expansion of more than 25% lifts the state to EXPAND. ADX below 15 with elevated ATR percentile flags CHOPPY. The tradeable set is exactly four: STRONG_TREND, WEAK_TREND, BREAKOUT, and EXPAND — three states close the door.

After the signal passes, four confirmation gates must all clear. CheckConfirms() evaluates the higher-timeframe EMA first: an H1 EMA(50) (set in HTF_TIMEFRAME and InpHtfEmaPeriod) must sit on the buy side of the closed-bar close for longs, on the sell side for shorts. Second, the Heikin-Ashi confirm is checked when InpUseHeikinAshi is true (default): the constructed HA candle for the most recent closed bar must be bullish for a long and bearish for a short, derived by averaging open, high, low, and close into the HA body. Third, the prior candle body must agree: a long entry requires the previous bar to close above its open, and shorts require the inverse. Fourth, the reward-to-risk ratio of the proposed trade must be at least 1.5. Because stop and target are set at 2.0x and 4.0x ATR(14), the 1.5 threshold is essentially a sanity check that ATR data was successfully copied and the math is non-degenerate.

The capital cap is the other half of the risk architecture. ComputeEffectiveCapital() takes the minimum of InpCapitalCapAmount and current equity when the cap is positive, otherwise falls back to raw equity. That figure flows into the lot calculation: CalcLotByRisk() uses tick value and tick size to translate the 0.5% risk budget into a lot count normalized to the broker's volume step. The daily loss limit (InpDailyLossLimitPct at 3%) and the hardcoded weekly limit (WEEKLY_LOSS_PCT at 6%) both reference the effective capital. A separate 4-hour cooldown fires from g_cooldownUntil after UpdateLossStreak() detects three or more consecutive losses.

In-trade management follows a layered rule set. When price reaches +1R, two adjustments fire: the stop ratchets to break-even plus 2 points (one-shot, controlled by USE_BREAK_EVEN), and half the position is closed at market if the remaining lot is at least twice the broker minimum (USE_PARTIAL_TP). From that point forward, a forward-only ATR trail at InpAtrTrailMult (2.5xATR(14)) ratchets the stop to lock in additional progress — the stop only ever moves in the direction of the trade. The position is force-closed if it survives longer than InpMaxBarsInTrade (200 bars) or if the Hull color flips to the opposite direction with at least one bar of confirmation.

The trade-management side does not run on every tick. OnTick() calls ManageOpenPositions() only when the ATR buffer is fresh, and the entry logic is gated by IsNewBar() so signals can only fire once per closed M5 bar. TryPyramid() is wired but disabled by default (USE_PYRAMID is hardcoded to false); enabling it permits up to three same-direction legs spaced by 2.0xATR of progress, which matches the conservative character of the rest of the design.

Execution safety is built into the path. OrderCalcMargin() runs before every entry, and the EA refuses to open a trade when required margin exceeds free margin. InpDryRun defaults to true, so a fresh install logs trade intentions to the experts tab without sending orders. TryClose_EX18061, TryClosePartial_EX18061, and TryModify_EX18061 retry up to three times on requote, timeout, price-off, or price-changed retcodes with 200ms / 100ms back-off before giving up. InpKillSwitch is a hard master cutoff that prevents new entries without affecting existing positions.

In a backtest, expect a small number of trades per day — the 4-gate confirm plus the regime filter plus the trend requirement naturally compress the entry set. The custom OnTester() formula multiplies net profit by profit factor and divides by one plus the balance drawdown, with a 30-trade minimum to score non-zero. That favors a profile with high reward-per-trade and a low-to-moderate trade count over a high-frequency equity curve. The Heikin-Ashi gate is the most useful off-switch for traders who find the system too selective: disabling InpUseHeikinAshi increases the entry rate while preserving the rest of the risk and regime logic.

The natural fit is a trader with a $100-plus account trading XAUUSD on M5 with an ECN or low-spread broker, a tolerance for waiting out drawdowns, and a willingness to let winners run to the full 4.0xATR target. The 7-21 server-time session matches the London-New York overlap during US hours, and the 3-pip spread cap means a broker that widens aggressively in pre-market or rollover sessions will produce more spread_high log entries than trades.