Pipsgrowth EX02022 Breakout

Pipsgrowth EX02022 Breakout is a pure Donchian-channel breakout EA built for XAUUSD on the 5-minute chart. It runs a deliberately simple signal pipeline — a closed-bar close above the prior N-bar highest high triggers a long, a close below the prior N-bar lowest low triggers a short — and then layers position management on top of that signal. There is no ADX filter, no higher-timeframe trend gate, no body/range strength test, and no regime classifier in this code. The edge is expected to come from how the EA sizes, trails, pyramids, and protects capital, not from a heavily filtered entry.

The channel is constructed inline in CheckSignal() by walking the previous 1..InpPeriod bars and tracking the running maximum high and minimum low. With the default InpPeriod = 20, the EA watches the highest high and lowest low of the most recent 20 closed five-minute bars, which on gold is roughly 100 minutes of price action. A long fires when Close[1] closes above that high, a short fires when Close[1] closes below that low. The decision runs once per new bar (lastBar guard in OnTick()), so signal evaluation is stable and re-attempts won't spam orders within the same candle.

Risk defaults are noticeably more aggressive than the rest of the EX02 family. InpRiskPercent = 1.0 percent of account balance per trade (other EAs in this family default to 0.5%), and InpStopLoss = 400 points while InpTakeProfit = 1200 points — a hardcoded 1:3 reward-to-risk ratio baked into the order ticket, not a multiple of ATR. The lot sizing path goes through CalculateLotSize(), which divides balance * riskPercent / 100 by the per-lot risk implied by the 400-point stop, snaps the result down to the symbol's LotsStep(), and clamps to LotsMin() / LotsMax(). If the user sets InpRiskPercent = 0, the EA falls back to InpFixedLot = 0.01.

The single order retry logic is robust: OpenBuy() and OpenSell() both pre-check free margin via OrderCalcMargin(), then attempt trade.Buy() / trade.Sell() up to three times at 200 ms sleep between attempts, refreshing Ask/Bid between attempts, and breaking out on any non-recoverable retcode. Filling is forced to FOK via trade.SetTypeFilling(ORDER_FILLING_FOK). After the order is in, the same retcode-aware retry pattern runs in TryModify_EX02022() for the trailing-stop updates, also with a 200 ms back-off.

Pyramiding is the headline feature of this EA. With InpEnablePyramid = true (the default) and InpMaxPositions = 3, the EA will stack up to three long or three short positions in the same direction, but only if the existing position is in profit by at least InpMinProfitPoints = 250 points. That gating is enforced by IsGridSafe(), which loops over open positions for this symbol and magic and returns false if any one of them is below the threshold. The effect is that the EA only adds to a trade that is already working in the trader's favour — it does not average down into losers, and it does not pyramid faster than every 250 points (≈ 2.50 USD on a 0.01-lot XAUUSD) of favourable move on the existing position. This is the central risk-control design choice in the EA: aggressive 1:3 RR entries with a single position, optionally scaled up to a three-position stack on confirmed breakouts.

The trailing stop logic in ManageProfitLock() is a classic fixed-distance ratchet. Once price moves InpTrailStart = 300 points in profit on a position, the EA sets the new stop to Bid - InpTrailDistance = Bid - 150 points for longs (and the mirror for shorts). The ratchet only ever moves the stop closer to price — it never widens it. The check runs every tick, so the stop is updated on every price change once the activation threshold has been crossed, with no minimum step gap. With the default 300/150 settings, the trail activates at 3.00 USD of move and locks at 1.50 USD behind price, which leaves a comfortable buffer for the 5-minute volatility of gold.

The no-trade gate is implemented as a single IsSafeToTrade_EX02022() function that returns false on any of the following: market closed (Saturday, Sunday, or the Friday 22:00 cutoff), session filter not active (only checked when InpUseGMTSessions = true, which defaults to off), a high-impact news event inside ±30 minutes (InpFilterNews = true default, queried via CalendarValueHistory() for the symbol's country), equity below 70% of the initial balance, fewer than 10 trades already taken today, an active cooldown after 5 consecutive losses, or realized daily loss equal to or worse than InpDailyLossPct = 5% of the initial balance. The cooldown itself is implemented in OnTradeTransaction(), which is the only place in the EA that updates the cumulative g_realizedToday/g_realizedWeek P&L counters and the g_consecLosses loss streak; hitting 5 in a row sets g_cooldownUntil = TimeCurrent() + 4 hours.

The session filter, when enabled, is set to London 7-16 GMT and New York 12-21 GMT (the default InpNewYorkEndGMT = 21 is the only input visible at the end of the GMT group because the others use literal input declarations inside the body rather than at the top of file — a minor quirk of this source). When InpUseGMTSessions = false, which is the shipped default, the session gate is bypassed entirely and the EA will fire on a clean signal any time the market is open. The GMT offset used by the session and market-open checks is auto-detected by DetectGMTOffset_EX02022() via a brute-force search over offsets -12..+12, picking the one that best fits weekday 1-5 and 7-21 hour, defaulting to +3 if no better candidate scores.

The capital-cap feature is shipped disabled: InpCapAmount = 0.0 means the EA treats the full account balance as effective capital for sizing. Setting it to a positive value (the input expects a USD figure) caps the equity that the EA "sees" at min(balance, InpCapAmount), and InpCapFloor = 1000 enforces a minimum effective capital before trading is allowed. This is intended for users who want to run a small portion of their account through the EA without rewriting the risk percent.

The OnTester() custom fitness function is profit * profitFactor / maxDD, with two guard rails: profitFactor is clamped to the 0.001-1000 range, and the function returns 0 if fewer than 10 trades are present in the test (much looser than the >=30 gate other EAs in this family use). The optimizer will therefore prefer parameter sets with positive net P&L, a profit factor above 1, and low max drawdown, but it will not reject a low-trade-count backtest outright — useful for fast first-pass optimization runs.

In practical terms, this EA is for a trader who is comfortable with a relatively unfiltered breakout entry on gold and wants the position management to do the heavy lifting. The risk profile is medium: a single 1% risk trade with a 400-point stop, optionally scaled to three positions when the breakout extends, and a 150-point trailing stop that locks 300 points in. The recommended broker profile is one that can fill 0.01-lot orders on XAUUSD with reasonable slippage under the 250-point (2.50 USD) max-spread ceiling, an ECN/RAW account for best fill quality, and a server time that aligns with GMT +2/+3 so the auto-detected offset makes the session filter behave intuitively. The EA runs every tick for the position-management loop, so the host should have a low-latency connection to the broker to avoid stale-stop scenarios during fast gold spikes.