Pipsgrowth EX16073 Trend

Pipsgrowth EX16073 is a dual-moving-average trend EA that hunts for the first moment a fast exponential average crosses a slow one — and only takes the trade when the market is calm enough to mean it. Two EMAs do the talking (a slow 65-period and a fast 14-period, both on close, both smoothed with the same MODE_EMA method), but the trigger is not the crossing itself. The trigger is the moment the two lines are still hugging each other — within 2 pips of one another — with the fast line on the right side. That narrow corridor is where the EA wants to be, because that is when a new impulse has just nudged a slow trend out of equilibrium and a directional move is most likely to extend.

Two volatility filters sit on top of that crossover zone. The first is a 26-period ATR — both a buy-side and a sell-side handle, with their own thresholds (InpATR_Level_Buy and InpATR_Level_Sell, default 0.0016). The ATR must be below that level for a buy to fire, and the same threshold governs sells. The second is a 54-period StdDev with a default ceiling of 0.0051. The two filters are not redundant: ATR measures range expansion over a fixed window, while StdDev measures the dispersion of closes around their own mean. A market can pass one and fail the other. The EA demands both pass.

The entry logic is symmetrical but the parameters are not. Each side carries its own ATR period, ATR level, StdDev period, StdDev level, MA-distance floor, stop-loss and take-profit. In practice that lets a trader dial buys for, say, a faster MA pair and sells for a slower one without re-compiling. The buy branch in OnTick reads: fast 14-EMA is above the slow 65-EMA, the gap between them is less than InpMAs_Delta_Buy pips, the 26-bar ATR is below 0.0016, and the 54-bar StdDev is below 0.0051. If the count of open positions (filtered by symbol + magic) is already at InpMaxPositions (default 1), nothing fires. The sell branch mirrors the conditions with > and a delta-floor for the opposite side. Note: setting either delta to 0 disables that side's gap filter entirely — useful for traders who want a pure cross without proximity gating.

The bar-gate logic is conservative. OnTick first calls Trailing() on every tick so that the ratchet always runs, but the signal evaluation only happens at the open of a new bar (a static PrevBars variable compared against the current iTime(0)). One decision per bar, no jitter trades. Once a position is open, m_last_deal_IN_buy and m_last_deal_IN_sell (updated by the OnTradeTransaction callback whenever a DEAL_ENTRY_IN with the right magic lands in history) prevent the same-side condition from re-firing on the next bar of the same trend.

Position management is where the strategy earns its keep. The defaults are 50-pip stop, 50-pip take-profit, 5-pip trailing stop, 5-pip trailing step — i.e. a 1:1 R:R on entry, rescued by a tight trail. The Trailing() function only moves the stop in one direction (up for longs, down for shorts) and only when price has moved further than TrailingStop + TrailingStep beyond the entry (default 10 pips). It refuses to lower an existing SL on a long, refuses to raise an existing SL on a short, and leaves a missing SL (i.e. one that has never been installed) untouched. The m_adjusted_point multiplier handles 3- and 5-digit brokers (e.g. gold at 2 decimals or 3, JPY pairs) so that the 50-pip figure stays a true 50 pips on every symbol.

The size is fixed. InpLots is the only sizing input — there is no risk-percentage formula, no balance curve, no progressive lot multiplier. The trade opens with a single call to m_trade.Buy or m_trade.Sell after a m_trade.CheckVolume pass that returns the broker-permitted lot for the requested price. The OpenBuy and OpenSell functions both call m_trade.SetExpertMagicNumber(m_magic) in OnInit, drop the comment Psgrowth.com Expert_16073 on every deal, and tolerate a 10-point slippage. Three retry helpers (TryClose_EX16073, TryClosePartial_EX16073, TryModify_EX16073) each loop up to three times on requote, timeout, price-changed, or price-off errors with a 100–200 ms back-off, but in this build the Trailing() function is the only place where TryModify_EX16073 is actually called — the close helpers are wired in but never invoked from OnTick, so position exits are left to SL, TP, or manual action.

What the EA does not do is worth listing. There is no news filter, no session gate, no time-of-day restriction — it will take a signal at the open of the New York evening or the Asian session open if the four-bar pattern aligns. There is no drawdown cap, no daily-loss circuit breaker, no consecutive-loss cooldown, no equity-floor. There is no spread filter; on a widened-spread candle the trade can still fire. There is no opposite-signal close — once a long is open, an emerging short signal will not flatten it. Those features are deliberate omissions: the EA is built as a pure trend-emergence engine and assumes the operator (or a wrapper EA) is managing account-level risk separately.

For backtesting, the expected behaviour is that most months will print a small handful of trades — the four-bar gate (cross + proximity + ATR + StdDev) is narrow and only fires during specific quiet-then-impulse transitions on the chosen timeframe (M5 or H1 per the family default). The 1:1 R:R means breakeven expectancy requires a win rate above 50%, but the 5+5 trailing converts many of the would-be scratch losses into small wins and many of the would-be 1:1 wins into extended runners, so live equity curves tend to be more profitable than the static backtest would suggest. The asymmetry between InpStopLossBuy and InpStopLossSell, and between the two ATR/StdDev levels, lets a trader who has profiled one symbol tune buys and sells independently — for example, cutting the sell-SL on a market that historically reverses sharply from short setups while leaving the buy-SL wider for trend days.

A reasonable first test plan: run the EA on XAUUSD M5 with the published defaults (0.1 lot, 50/50, 5/5 trail, MA 14/65 EMA, ATR 26 @ 0.0016, StdDev 54 @ 0.0051) over a 12-month window using every tick from a real-history-quality source. Look for two things — frequency of signals (should be sparse, not every day) and the ratio of trailing-stop exits vs TP hits vs SL hits. If trailing-stop exits dominate, the 5-pip step is working; if SL hits dominate, the 50-pip stop is too tight for the symbol's noise floor and should be widened. Because the EA has no internal P&L guard, every test should be wrapped in a parent script that enforces a maximum drawdown and closes the chart on breach.