Pipsgrowth EX12064 MultiIndicatorConfluence

Pipsgrowth EX12064 is a 9AM London GBPUSD breakout that doesn't read a single indicator. The full source is built on a 2005 community request quoted directly in the comment block: 'I looking for system automat wich will be trade for me this 9:00 GMT look for price GBP/USD then price is 1.8230 example (bid)'. The brief was to look at the price at 9 AM London, then place a Buy Stop above and a Sell Stop below that snapshot and let the market decide which side breaks first. That request is still the soul of the EA — every parameter in the file is a knob on that single idea.

The execution model is a daily pair of pending orders. On every tick, OnTick() reads the server clock and waits for STimeCurrent.hour == InpLookPriceHour (default 10 — the input comment explicitly says 'Change for your time zone (my is +1 Hour). Should be 9AM London time') and STimeCurrent.min >= InpLookPriceMiin (default 0). When that window is reached and the boolean clear_to_send is still true, the EA first calls CloseAllPositions() and DeleteAllPendingOrders() to flush the previous day's trades, then asks m_symbol.Ask() + ExtDistanceBuy and m_symbol.Bid() - ExtDistanceSell for the two trigger prices. The two pending orders are sent through PendingBuyStop() and PendingSellStop() and clear_to_send flips to false. The two orders live as a pair with ORDER_TIME_GTC, so they don't expire at end of day — they only get cleaned up by the next morning's reset or by the cancel-opposite logic when one of them fills.

The asymmetric geometry is the most distinctive design choice. InpDistanceBuy defaults to 18 pips and InpDistanceSell defaults to 22 pips, while InpStopLossBuy defaults to 22 pips and InpStopLossSell defaults to 18 pips. The two wings are deliberately not mirrored. The Buy Stop sits closer to the snapshot (18 pips) and is given a wider stop (22 pips), while the Sell Stop sits farther from the snapshot (22 pips) and is given a tighter stop (18 pips). The 2005 community request called this out explicitly: 'open long 18pips+4pips spread at this price so will be LONG 1.8252 profit 40pips stop 22pips' and 'open short 22pips-4pipsspread so 1.8230-22-4=1.8204 so 1.8204 short profit 40pips stop 18pips'. The four-pip spread on the original 2005 GBPUSD quote is what forced the asymmetric offsets, and the EA preserved that geometry as the default. The shared take-profit, InpTakeProfit = 40 pips, applies to both wings identically — it's applied as Ask + 40 pips for the Buy Stop and Bid − 40 pips for the Sell Stop, so each filled position has a roughly symmetric reward arm of 40 pips regardless of which side triggers first.

The cancel-opposite behavior is the EA's only 'management' mechanic. After the pair is placed, every subsequent tick runs CalculatePendingOrders(count_buy_stop, count_sell_stop) to count how many of each type are still pending. The condition ((count_buy_stop==0 && count_sell_stop!=0) || (count_buy_stop!=0 && count_sell_stop==0)) checks whether exactly one wing has filled and the other is still pending. When that asymmetry appears, DeleteAllPendingOrders() fires and removes the surviving pending order so the EA isn't carrying both a filled position and a stale stop in the opposite direction. The filled position itself is left to run until it hits its 40-pip TP, its per-side SL, or the daily close-hour cut.

The clear_to_send flag governs when the next morning's snapshot is allowed to fire. It only resets under two conditions: (1) the hour is no longer the look hour, a position exists, and the minute is past InpLookPriceMiin — i.e. yesterday's trade is still open and the new day is being respected; or (2) the clock is in the 10-minute window before the next look hour (STimeCurrent.hour == InpLookPriceHour-1 and MathAbs(STimeCurrent.min - InpLookPriceMiin) < 10) — i.e. we're approaching tomorrow's snapshot. Both paths are safety nets: the first prevents double-snapshotting a day that already has an open position; the second ensures a clean reset even if the previous day's pair was never triggered. clear_to_send is intentionally a bool rather than a date stamp, which is the simplest possible state machine for a once-a-day EA.

The end-of-day exit is InpUseCloseHour (default true) combined with InpCloseHour (default 18). On any tick where STimeCurrent.hour >= InpCloseHour and the close-hour feature is on, OnTick() calls CloseAllPositions() and DeleteAllPendingOrders() and returns early — no new pending orders are placed, no further logic runs. This means an untriggered Buy Stop / Sell Stop pair from the morning does not survive overnight. If a trader wants to hold the breakout overnight, the input is a one-click toggle: set InpUseCloseHour = false and the orders remain active until the next morning's reset cycle.

The trade-execution layer is built on the MT5 standard library classes — CTrade, CPositionInfo, CSymbolInfo, CDealInfo, COrderInfo — wrapped in three retry helpers, TryClose_EX12064, TryClosePartial_EX12064, and TryModify_EX12064. Each helper attempts the trade up to three times, sleeping 200 ms between close attempts and 100 ms between modify attempts on TRADE_RETCODE_REQUOTE, TRADE_RETCODE_TIMEOUT, TRADE_RETCODE_PRICE_OFF, or TRADE_RETCODE_PRICE_CHANGED. The m_trade object is configured once in OnInit() with SetExpertMagicNumber(m_magic) (22212064), SetMarginMode(), SetTypeFillingBySymbol() to match the broker's allowed order type, and SetDeviationInPoints(10) for the 1-pip slippage tolerance. CheckVolumeValue() validates the lot size against the symbol's LotsMin, LotsMax, and LotsStep before any order is sent. The m_adjusted_point variable multiplies the MT5 point by 10 on 3- and 5-digit symbols so that the input's 'pips' (which are 1-pip units, not points) read correctly across brokers.

The 13 inputs split across two groups — === Strategy === and === Identity === — are all that the user can change. There's no spread filter input, no session input beyond the two time markers, no news filter, no indicator settings, no magic-derivatives, no money-management input other than InpLots. The InpReport input (15 seconds) drives an EventSetTimer that periodically calls ReportStrategy() from OnTimer(); ReportStrategy() walks the deal history, filters by symbol and m_magic, and writes a single Comment(...) line to the chart showing Executed X+Y trades with P+P_open = total profit and the current server and local hours. There's no graphical panel, no dashboard, no on-chart indicator drawing — the strategy is intentionally minimal because the request was for a 9AM London breakout and nothing else.

The practical constraint is spread sensitivity. With InpDistanceBuy = 18 pips, a 4-pip spread (the 2005 baseline) eats 22% of the offset to the Buy Stop; with a 6-pip spread it's 33%. The 40-pip TP is enough to absorb up to 8 pips of round-trip spread and still net 32 pips of profit, but the EA's edge is real only when the broker's typical spread is below 4-5 pips on GBPUSD. That's why the EA works best on a low-spread ECN or RAW account during the London open, when GBPUSD's typical spread is at its session minimum. Backtests on the strategy will look very different on a 3-pip broker vs a 1-pip broker — the same code, the same inputs, but the trade count and average profit-per-trade shift materially. The recommended minimum balance is $100 for a 0.01-lot micro test and roughly $1,000 for a 0.1-lot standard test, which is the file's InpLots = 0.1 default.

This EA is the right choice for a trader who wants a deterministic, time-driven breakout on a single pair, who is happy to make the entire strategy's behavior controllable from 13 inputs, and who is willing to size lots manually rather than use a percent-of-balance money manager. It's not the right choice for traders looking for a martingale, a grid, an indicator stack, a multi-symbol basket, or any of the more elaborate tactics that other EAs in the family provide. The whole file is a 2005 forum request — 'I will be glad if I see this in code' — faithfully executed.