Pipsgrowth EX12065 MultiIndicatorConfluence

EX12065 — Gen2_RSA_Gold — is a 4-handle breakout EA that adds an RSI directional filter to the Gaussian-band framework that runs across the rest of the EX12 family. The header reads "Gen2" because the previous generation in the family was the pure Gaussian/Alligator/Donchian stack; this one keeps the Gaussian band construction but drops the Alligator, Donchian, ADX, D1 and W1 optional filters and instead hard-codes RSI(14) as a confirmation layer on every entry and as the trigger on every exit. The file is short — 16 functions, 21 inputs, 0 defines — and the architecture is small enough to walk through end to end.

The band construction starts with a 10-period EMA on the working timeframe, which the code calls the GaussianFilter. Two bands are drawn around that midline at ±DistanceMultiplier × ATR(14), with DistanceMultiplier = 0.85 by default. So if the midline sits at 1900.00 and ATR is 4.50, the upper band is at 1903.83 and the lower band at 1896.18. The midline itself, not the close, is the reference: the EA tracks gaussNow and gaussPrev to confirm slope direction.

Entry conditions are stacked, not OR'd. The first gate is the strong-candle filter — body must be at least 60% of the candle's full range. The second is the band cross with a 2-bar confirmation: today's close must be above the upper band while the previous two closes were both below it (mirror for sells). The third is the trend gate — close must be on the correct side of a 50-period EMA on the higher timeframe, defaulting to H4 (TrendFilterTF = 6). The fourth is slope — gaussNow > gaussPrev for buys. The fifth is break distance — price − gaussNow > 0.3 × ATR, which prevents the EA from buying noise that barely pokes through the band. The sixth and decisive gate is the RSI confirmation. With RSI_Period = 14 and UseRSIForEntries = true, a long requires rsi < 70 AND rsiPrev < rsi — that is, RSI is below overbought AND rising. A short requires rsi > 30 AND rsiPrev > rsi. That "rising RSI on a bullish break" filter is what makes the Gen2_RSA label earn its name: the RSI is not decoration, it gates every trade.

Before sending the order, CloseOpposite() walks the open positions and closes any that are on the wrong side. This is a directional pre-filter, not a hedge: if the EA is already long and a short signal prints, the long is closed first and then the short opens. The trade objects use trade.Buy(Lots, _Symbol, ask, slBuy, tpBuy, InpTradeComment) with Lots = 0.1 fixed, slBuy = ask − 1.2 × ATR, tpBuy = ask + 160 × _Point, and the Psgrowth.com Expert_12065 comment. HardSL_Points = 120 is the fallback used when ATR returns 0; the live SL scales with volatility.

Position management runs on every tick after a position exists. The ATR trailing stop is the first layer: if the new computed stop is on the right side of the current stop, TryModify_EX12065() pushes it. The breakeven layer is the second: once bid − entry > 1.2 × ATR (the same multiple as the trailing ATR), the stop is ratcheted to entry. The dynamic lock-profit layer is the third: with LockProfitEvery_X_Dollars = 2.0 and LockMinusBuffer = 1.0, every $2 of unrealized profit steps the lock up by $1, so at $4.20 of profit the lock sits $3 above entry (2 steps × $2 − $1 buffer). Each layer is forward-only — none of them widen the stop.

The exit system is where the Gen2_RSA architecture pulls its weight. After the position-management block, the EA reads RSI on bars 0, 1, and 2 and computes rsiTrendUp = rsi > rsiPrev > rsiPrev2 (and the mirror for down). With UseRSIEarlyExit = true and RSI_ExitConfirmationPeriods = 2, the EA counts how many of the last 2 RSI prints are overbought (for a long position) or oversold (for a short). If both bars qualify and UseRSITrendForExit = true, the EA also requires rsiTrendDown to confirm a long exit. There is a fourth, optional exit — UseRSIDivergenceExit = false by default — that fires when price makes a higher high but RSI makes a lower high (or the mirror) within the last 3 bars. None of these exits are TP-replacement: the fixed 160-point TP is still on the ticket and will close the trade naturally.

The capital cap and the session cap close out the system. CheckMaxDrawdown() returns false once 100 × (balance − equity) / balance > MaxDrawdownPercent (10.0), and OnTick exits early when that flips. The Friday machinery is built around a 17:00 server-time market close: between 15:00 and 17:00 on Friday, IsNearFridayClose() returns true, and if CloseTrades2HoursBeforeFriday = true the EA calls CloseAllPositions() and returns. If AvoidTrading2HoursBeforeFriday = true, no new entries can fire in that window either. The AutoTuneFilters() function sniffs the live ASK-BID spread and toggles pipFactor between 0.0001 and 0.01, which is what lets the same MinATR = 0.03 default work on both 4-digit FX pairs and 2-digit XAUUSD. UseAutoFilter = true flips this on by default; setting it to false makes the EA use the raw MinATR and MinBreakDistance values from the inputs.

There are a few other inputs worth knowing. OneTradeOnly = true means the EA is effectively single-position — even though MaxOpenTrades = 5 is exposed, the canTrade clause only allows an entry when no position exists. AllowOnlyProfitableAdditions = true plus MinProfitPerTradeToAdd = 5.0 is a redundant guard: with OneTradeOnly active, every existing position has to be sitting on at least $5 of unrealized profit before another can open. UseRSIForEntries, UseRSIEarlyExit and UseRSITrendForExit are all on by default; UseRSIDivergenceExit is the only RSI input that ships off. The three retry wrappers — TryClose_EX12065, TryClosePartial_EX12065 and TryModify_EX12065 — handle REQUOTE, TIMEOUT, PRICE_OFF and PRICE_CHANGED with 3 attempts and 200ms (or 100ms for modify) sleeps.

What to expect in a backtest: a 4-handle EA with single-position discipline and 6-gate entries should print a low-frequency, high-conviction equity curve. The 2-bar band confirmation and the 0.3×ATR break distance will keep win rate elevated in trending gold sessions but will cause the EA to miss the first 1-2 bars of every impulse — the trade-off is fewer whipsaw losses in exchange for entering late. The dynamic lock-profit is a meaningful contributor: the $1 buffer is what makes the system give back some open profit on pullbacks rather than getting stopped at breakeven, so expect realized R to be slightly worse than the gross R the trade would have printed at peak. The 10% drawdown cap is the only account-level kill switch — when the EA hits it, it stops opening new trades for the day but does not close existing ones. Pair this EA with an ECN-style account so the 0.85×ATR bands and the 1.2×ATR trailing stop do not get eaten by spread on every re-quote.