Pipsgrowth EX03024 Grid

Pipsgrowth EX03024 is a one-direction-at-a-time averaging grid for XAUUSD on the M5 chart. The whole point of the EA is the way it opens its first trade: a single 14-period RSI on close price decides whether the first leg is a buy or a sell, and the EA then commits to averaging into that direction with up to four follow-up entries at fixed 300-point spacing. The grid grows one rung at a time as price moves against the initial entry, and each rung carries its own 500-point take-profit. There is no martingale by default, no hedge grid, no pending-order ladder, no per-trade stop-loss, and no basket profit target — the architecture is closer to a disciplined DCA plus target than a true market-maker grid.

The decision tree starts in OnTick(). Every tick first calls UpdateDailyCounters_EX03024() to roll the realized P&L buckets when the day changes, then refreshes the symbol object with mysymbol.RefreshRates(). Two hard gates fire before any strategy logic: a spread filter at mysymbol.Spread() > InpMaxSpread (250 points) and the consolidated IsSafeToTrade_EX03024() function. The safety function checks the GMT session, the cooldown timer, an optional equity floor (InpEquityFloor), an optional realized-daily loss cap (InpDailyLossLimit), an optional realized-weekly loss cap (InpWeeklyLossLimit), an optional balance cap (InpCapitalCap), the max-trades-per-day limit, and the optional max-consecutive-losses gate. If any of these blocks, the tick returns silently and the EA does nothing — no entries, no adjustments, no grid add-ons.

With the safety gates clear, the EA runs ManageGrid() unconditionally on every tick, then CheckSignal() only on the open of a fresh M5 bar (guarded by a static lastBar == currentBar comparison). CheckSignal() ignores the grid and fires only when PositionsTotal() == 0 — once any position exists, the grid owns the runtime and the RSI entry is dormant. When the position book is empty, CheckSignal() copies the previous-bar RSI value with CopyBuffer(hRSI, 0, 1, 1, bufRSI) and branches on the level: a value below 30 calls OpenBuy(InpStartLot), a value above 70 calls OpenSell(InpStartLot). The InpUseRSIFilter input lets the user disable the filter, but the source still calls the same CheckSignal() function — disabling it just leaves the position book to be filled by the grid logic instead, so the EA will still only ever build from the first RSI-driven anchor.

The grid logic is the heart of the EA. ManageGrid() walks PositionsTotal() from the most recent ticket to the oldest, filtering on position.Symbol() == _Symbol && position.Magic() == InpMagicNumber (the magic default is 22203024). For every buy position it tracks the count and the lowest open price, and for every sell position it tracks the count and the highest open price. Then it compares the current mysymbol.Bid() against the last buy open price: if the bid is at least InpGridStep * mysymbol.Point() (300 points) below the lowest buy, it opens another buy sized at lastBuyLot * InpLotMultiplier. The mirror logic applies to sells — if the ask is at least 300 points above the highest sell, it opens another sell at the multiplied lot. The cap is InpMaxLevels (5 by default), so the grid tops out at 5 buys in a row or 5 sells in a row. With InpLotMultiplier = 1.0 the lot is fixed at InpStartLot (0.01) and every rung is identical, so the EA is a true averaging-in grid rather than a martingale. Setting the multiplier above 1.0 turns it into a soft martingale: 0.01, 0.015, 0.0225, 0.0338, 0.0506 across five rungs at 1.5×.

Each OpenBuy() and OpenSell() call calculates a per-trade take-profit of mysymbol.Ask() + InpTakeProfit * mysymbol.Point() (or Bid - 500 points for sells) and passes a stop-loss of 0 to the trade request. There is no SL anywhere in the system — the only stop is the optional InpEquityFloor and the realized-loss caps in IsSafeToTrade_EX03024(). TryBuy_EX03024() and TrySell_EX03024() both run a pre-trade margin check via OrderCalcMargin() (the trade is refused before the request goes out if marginReq > AccountInfoDouble(ACCOUNT_MARGIN_FREE)), then retry the order three times at 200ms gaps on the transient retcodes REQUOTE, TIMEOUT, PRICE_OFF, and PRICE_CHANGED. ORDER_FILLING_FOK is set in OnInit() and the slippage tolerance is InpSlippage points.

The trailing infrastructure is the rest of the hardening layer. DetectGMTOffset_EX03024() runs a best-of sweep from -12 to +12 on the server clock to guess the broker's GMT offset, scoring each candidate by weekday and trading-hour likelihood — the winner seeds g_gmtOffset and ServerToGMT_EX03024() adjusts every session check. InActiveSession_EX03024() blocks the weekend (Saturday and Sunday return false immediately) and, when InpUseGMTSessions is true, restricts the trading window to the union of InpLondonStartGMT=7 through InpLondonEndGMT=16 and InpNewYorkStartGMT=12 through InpNewYorkEndGMT=21. OnTradeTransaction() listens for TRADE_TRANSACTION_DEAL_ADD events that match the magic and the symbol, adds the profit + swap + commission to the daily and weekly realized buckets, increments the hardcoded g_consecLossesHD on a losing deal, and resets it on a win. Three consecutive losses hardcode a 30-minute cooldown via g_cooldownUntil = TimeCurrent() + InpCooldownMinutes * 60 — this cooldown is independent of InpMaxConsecLosses, which is left at 0 (disabled) in the source.

OnTester() returns a custom fitness score of (profit / maxDD) * (pf > 1 ? pf : 0), gated on maxDD > 0 and a minimum of 10 closed trades. The strategy optimizer therefore rewards high net return relative to the worst peak-to-trough drawdown, and it cancels the score entirely if the profit factor fails to clear 1.0. Because the EA is a one-direction DCA, the optimiser finds different sweet spots from a true market-maker grid: tighter grid steps profit in shallow ranging while wider steps survive deeper pullbacks. Tuning the multiplier toward 1.0 keeps the worst-case lot exposure proportional to the account; pushing it above 1.5 turns EX03024 into a martingale where the loss of a full five-rung basket equals roughly 7.6× the risk of the first leg.