Pipsgrowth EX03023 Grid

Pipsgrowth EX03023 is a price-step grid EA built for XAUUSD on the M5 timeframe (it also works on H1). It does not read any technical indicator, does not compute a regime score, and does not wait for a moving-average cross or a momentum threshold. Every decision is mechanical: it arms a fixed ladder of pending limit orders at 50-point intervals from the entry anchor, lets the market decide which orders fill, and exits the entire basket the moment the unrealized aggregate crosses a single threshold. That narrowness is the point — there is no model to recalibrate, no volatility filter to fight, and no signal ambiguity. Either the grid is armed and waiting or it is closed and idle.

The arming logic lives in OnTick(). On every tick the EA first runs a hard drawdown guard: it computes (balance - equity) / balance * 100 and, if the result exceeds InpMaxDrawdown (20% by default), calls CloseAllTrades() and returns. That is the EA's panic stop — once a 20% peak-to-trough excursion is observed on the live account, the basket is force-flattened and every pending order is deleted in the same pass. After the DD guard, the EA checks the live spread with SymbolInfoInteger(_Symbol, SYMBOL_SPREAD) and aborts the rest of the tick if the spread is above InpMaxSpread (50 points). Then IsSafeToTrade_EX03023() validates session, cooldown timer, equity floor, daily/weekly realized loss, capital cap, max trades per day, and consecutive loss count before any new logic runs.

If the basket is still alive, the EA evaluates the close condition. GetBasketProfit() walks PositionsTotal() from newest to oldest and sums the profit, swap, and commission of every position whose magic number is InpMagicNumber (22203023) and whose symbol matches the chart. When CountOpenTrades() > 0 and that aggregate is greater than or equal to InpTakeProfit * _Point * InpGridLevels * InpLotSize * 100, the EA calls CloseAllTrades(), flips gridInitialised to false, and waits for a fresh cycle. With the default InpTakeProfit=100, InpGridLevels=5, InpLotSize=0.01, the close threshold for XAUUSD works out to roughly $0.50 of combined floating profit across the basket — a small, repeatable target designed to be hit often rather than heroically.

If no positions are open and gridInitialised is false, the EA captures the current ask and bid, stores them in gridBuyBase and gridSellBase, and sets gridInitialised = true. It then calls OpenBuyGrid(), which places InpGridLevels BuyLimit orders stepping down from gridBuyBase at InpGridStep * _Point intervals. With the default step of 50 points on XAUUSD, that produces five BuyLimits at approximately -0.00, -0.50, -1.00, -1.50, -2.00 USD relative to the anchor. The anchor is the ask at the moment the basket opens, so the first level is at or inside the spread and typically fills immediately, turning into the seed position; the rest queue below as price pulls back. Each order is sent through TryBuyLimit_EX03023, which retries up to three times on REQUOTE or TIMEOUT with a 200ms sleep between attempts before giving up.

If InpBothDirections is set to true, the EA additionally calls OpenSellGrid(), which mirrors the same ladder upward from gridSellBase — five SellLimits stepping up at 50-point intervals. With the default of false, only the buy ladder arms. That is the EA's directional knob: a single boolean that turns a one-sided buy-grid into a symmetric two-sided grid.

There are no per-trade stop losses and no per-trade take profits. Every level enters at the same size (InpLotSize = 0.01) and exits only when the basket is force-closed. The two ways the basket can close are: (1) the profit threshold is reached — CloseAllTrades() runs, the EA cancels all pending orders, resets gridInitialised, and the cycle restarts on the next zero-position tick; or (2) the drawdown guard fires and the EA flattens everything immediately.

OnTradeTransaction() is the only piece of the EA that updates state outside OnTick(). When a deal is added with the matching magic and symbol, it bumps g_dailyTrades on entries, accumulates realized P&L into g_realizedToday and g_realizedWeek on exits, and increments g_consecLossesHD on losses. If three consecutive losses stack up, the EA sets g_cooldownUntil = TimeCurrent() + InpCooldownMinutes * 60 (default 30 minutes) and refuses new entries until the timer expires. This is a hardcoded trigger count of 3 — the user-facing InpMaxConsecLosses (default 0, disabled) is a separate guard read by IsSafeToTrade_EX03023() that triggers at whatever threshold the user sets; the 30-minute cooldown is the EA's built-in safety floor.

The session filter uses DetectGMTOffset_EX03023() to auto-pick a GMT offset between -12 and +12 by scoring candidate offsets on weekday validity and trading-hour range, then walks the local server time through that offset. InActiveSession_EX03023() blocks weekends entirely, then returns true for any hour between InpLondonStartGMT (7) and InpLondonEndGMT (16) and any hour between InpNewYorkStartGMT (12) and InpNewYorkEndGMT (21). With InpUseGMTSessions set to true, the EA only operates inside the union of those two windows — a tight, liquid band that captures the bulk of XAUUSD's volume without spending server time on dead Asian hours when gold's typical daily range is still forming.

OnTester() returns (profit / maxDD) * (pf > 1 ? pf : 0) and gates the result on maxDD > 0 and trades >= 10, so optimization prefers robust E/R profiles that survive at least ten resolved cycles and have a profit factor above 1. With the small basket-TP target of $0.50, the typical strategy tester run will close dozens of baskets across a multi-year backtest, which gives the metric enough samples to differentiate parameter sets without overfitting to a single trade.

In practice this means the EA produces frequent, small wins (closed baskets) interrupted by occasional, larger losing cycles (drawdown-driven flattenings). The InpMaxDrawdown knob is the real risk dial: lowering it to 10% shortens losing cycles but raises the risk of being whipsawed out of a recovery, while leaving it at 20% gives baskets more room to mean-revert at the cost of deeper peak-to-trough excursions. The grid step and lot size together determine how many lots are exposed if every level fills before the basket resolves — at 5 levels × 0.01, that is 0.05 lots of XAUUSD running against the trend during a sustained move, with no protective stops on individual legs. This EA is not a system for traders who require per-trade stop losses. It is a structure for traders who think in baskets, accept the risk profile of a non-stopped grid, and want every decision in the code to be visible without wading through indicator plumbing.

Two practical notes before deployment. First, the InpMaxSpread filter of 50 points is sized for XAUUSD's typical 15–30 point spread on a low-cost broker; widening it will let the EA run on noisier venues but exposes it to slippage on entry. Tightening it below 20 will block the EA from arming the grid during the volatile first minutes of the London open. Second, the basket profit threshold scales with InpTakeProfit * InpGridLevels * InpLotSize * 100, so any combination that preserves the product — for example InpTakeProfit=200, InpGridLevels=5, InpLotSize=0.01 — produces a $1.00 target basket, while halving any one of them halves the target. The four-quadrant relationship is the only dial that matters for tuning win rate vs. win size, and it is intentionally exposed.