Pipsgrowth EX02021 Breakout

EX02021 is a Donchian-channel breakout system aimed squarely at XAUUSD on the 5-minute chart. The core idea is to wait for a 20-bar high/low range to be broken on the closed candle, then let a three-stage confirmation filter decide whether that breakout is tradable. The 20-period Donchian envelope is computed from bars InpDonchianShift to InpDonchianShift + InpDonchianPeriod (default shift=1, period=20) so the EA reads the band off confirmed bars — never the still-forming candle. A long signal requires close[1] > upper and close[2] <= upper; a short signal is the mirror. That double-bar requirement is the most important gate in the EA: a candle can sit above the band for three bars in a row and EX02021 will only fire on the first bar that punched through. After three bars above, the breakout is stale and the EA ignores it until price re-enters and breaks again.

Three filter stages run on every confirmed bar. First the ATR floor: if ATR(14) divided by _Point is below InpMinAtr (0.5 default), the trade is skipped — this is a de-facto no-trade filter against ultra-quiet candles. Second, an RSI(14) directional check: longs require RSI > 50, shorts require RSI < 50, but only when InpUseRsi is true. Third, MACD(12,26,9) on PRICE_CLOSE: longs need the MACD line above the signal line, shorts need it below — both filters can be disabled independently via InpUseMacd. A signal that fails any of these is zeroed out before the order manager ever sees it. The result is a tighter pipeline than a raw Donchian: most candles that touch the band never become trades.

Risk math lives in GetLotSize(). With InpRiskPercent (1.0% default) the EA sizes the position to risk 1% of AccountInfoDouble(ACCOUNT_BALANCE) over the SL distance using m_symbol.TickValue() and m_symbol.TickSize(). Set InpRiskPercent to 0 and the EA falls back to a flat InpFixedLot (0.01). The output is rounded to the symbol's lot step, clamped to LotsMin()/LotsMax(), and then capped at InpMaxLotSize (10.0). That 10-lot ceiling is a guard for prop-firm accounts where a runaway money-management module would otherwise exceed the broker's lot limit.

SL and TP are ATR-anchored, not fixed-pips. Stop loss = ATR(14) × InpSL_ATR_Mult (1.5) divided by _Point and rounded to int points, with a 50-point minimum. Take profit = ATR(14) × InpTP_ATR_Mult (2.0). The ratio is 1.5 : 2.0 = 1 : 1.33, which is tighter than the 1:2 many breakout EAs run. The trade-off: more trades hit TP because the target is closer, but the winner-to-loser ratio is also lower. Both prices are normalized through NormalizePrice_EX02021 so they survive the digit transformation on ECN quotes with nonstandard ticks.

EX02021 does not use a classic trailing stop. Instead it runs ManageProfitLock() on every tick. The function walks PositionsTotal() and, for any position that has moved at least InpLockTrigger (300) points into profit, it ratchets the stop loss in InpLockStep (50) point increments toward a target of InpLockDistance (100) points behind current price. The SL only ever moves in the favorable direction (it is a one-way ratchet), and the TP is left untouched. This is sometimes called a 'profit lock' or 'chandelier' style trailing — the position's stop loss does not catch up to a fast move until 300 points of profit are already on the table, then it advances slowly enough that a normal pullback won't tag it.

Pyramiding is on by default. CheckPyramidSafety() enforces a 'secured profit' rule: an additional position in the same direction is only allowed if the existing position's stop loss is already in front of its open price by at least InpPyramidMinProfit (100) points. That is, the previous position must be in a 'locked-in' state where even a worst-case stop-out produces a profit. The cap is InpMaxPyramidTrades (3). If InpUsePyramid is off, only one position can be open at a time. The function also blocks adding against existing opposite-side positions, so the EA cannot be net-long and net-short simultaneously.

The hardening layer is more lenient than the more recent breakout EAs in the family. There is no regime classifier and no ADX threshold. The trade filter instead relies on: InpMaxSpread (200 points — generous for XAUUSD on most brokers), InpUseTimeFilter with InpStartTime/InpEndTime (08:00–20:00 server time by default), and an InpCloseAtEnd switch that flattens everything at end-time if true. On top of that, the EA runs the full safety stack: IsMarketOpen_EX02021() blocks weekends plus the Sunday 22:00 GMT open and Friday 22:00 GMT close, InActiveSession_EX02021() gates London (7-16 GMT) and New York (12-21 GMT) but is off by default (InpUseGMTSessions = false), IsNewsTime_EX02021() calls CalendarValueHistory for ±30 min around high-impact events, and a 5% daily realized-loss limit (InpDailyLossLimitPct) plus a 5-consecutive-loss → 4-hour cooldown via OnTradeTransaction. A capital cap (InpCapAmount) and minimum equity guard (InpMinEquityPct = 70%) are present but disabled by default; enable them by setting InpCapAmount above zero and the EA treats the lesser of cap and equity as the effective balance.

Order execution goes through three retry passes with 200 ms Sleep() on TRADE_RETCODE_REQUOTE, _TIMEOUT, _PRICE_OFF, and _PRICE_CHANGED. A pre-trade OrderCalcMargin check rejects entries that would exceed free margin. Lot clamping normalizes the result to the symbol's step, minimum, maximum, and InpMaxLotSize. The OnTester formula is profit * profitFactor / maxDD with a 10-trade gate — lighter than the 30-trade gate used elsewhere in the family, which means short out-of-sample windows can still produce a non-zero custom criterion.

For deployment, the EA assumes a broker with reliable price feeds on XAUUSD M5: spreads up to 200 points (2.0 pips on a 5-digit gold quote) and slippage tolerance up to 30 points will pass. Because the strategy is purely bar-driven with no indicator state across symbols, the chart can be detached and the EA will run. The minimum deposit on the listing is $100, but with a 1% risk default and 0.01-lot minimum the practical floor is closer to $500–$1,000 for the EA to function at its full sizing — accounts under $200 will see most orders clamped to the minimum lot and the profit-lock ratchet doing most of the work.