Pipsgrowth EX10015 Momentum-Scalper

Pipsgrowth EX10015 is a single-indicator momentum scalper that listens to the Commodity Channel Index on a 15-minute chart and trades the moment the oscillator crosses the zero line. There is no EMA, no RSI confirmation, no ADX regime gate, no higher-timeframe filter — the entire decision tree reduces to one question: did CCI just flip sign? That minimalism is the EA's defining choice, and it shapes every other design decision that follows.

The signal itself is produced inside GetCCISignal(). The EA pulls two CCI values — the most recent completed bar (currentCCI, taken from bar index 1) and the bar before that (prevCCI) — and applies a strict two-condition rule. A long entry fires only when prevCCI <= 0 and currentCCI > 0 — the previous bar's momentum was non-positive and the new bar just pushed above the line. A short entry fires only when prevCCI >= 0 and currentCCI < 0. There is no smoothing, no buffer, no overbought or oversold gating on entry; if CCI isn't in the act of crossing, no signal. The CCI is calculated on PRICE_TYPICAL with the standard 14-period lookback — nothing exotic, no deviation multiplier overrides, just the textbook formula running on M15 candles.

ATR(14) on the same M15 timeframe is the only other indicator and is used solely for stop placement, not signal generation. Inside OpenPosition(), the EA computes slDist = max(currentATR * 1.5, StopLossPips * pipSize, minDist * 1.5) — a hybrid floor where the larger of ATR-1.5×, the 15-pip user input, and 1.5× the broker's SYMBOL_TRADE_STOPS_LEVEL wins. This means on quiet gold days the stop holds at the ATR value, on volatile days it widens with the range, and on broker-restricted symbols it never sits closer than the minimum distance the venue enforces. The take-profit distance is fixed at slDist * 2, giving a 1:2 reward-to-risk placement on every trade. The combination of zero-line cross entry and 1:2 R-mult TP means the strategy is a momentum catcher — it pays a wide stop so that two-thirds of the time the trade probably loses, and the rare CCI continuations it catches are expected to pay double the risk.

Position management runs on every tick through ManageOpenPositions(). Two exit paths are active in parallel. The first is a CCI-exit: if a long position is open and the most recent currentCCI reaches the overbought level of +100, the EA closes the position immediately; symmetrically, shorts close when CCI reaches -100. The currentCCI variable is the same one refreshed in RefreshIndicators() for the new bar, so this exit can fire as the trade's momentum collapses. The second exit is a trailing stop, controlled by the UseTrailingStop, TrailingStartPips (10) and TrailingStepPips (5) inputs. The trail activates only after unrealized profit exceeds TrailingStartPips * pipSize, and once active it ratchets the stop in TrailingStepPips-sized increments via SafePositionModify(). The safe-modify wrapper validates the proposed stop against SYMBOL_TRADE_STOPS_LEVEL before sending the modify, falling back to the minimum distance if the proposed ratchet would sit too close to the current bid/ask.

A single hard position-size cap governs entries: CountPositions() >= 1 returns true and blocks any new signal. The EA does not pyramid, does not scale into winners, and does not hold a long and a short at the same time. The next signal waits until the current position is closed by TP, SL, CCI exit, or trailing stop. This single-position discipline is what makes the trailing stop meaningful — the EA never has to choose which of several open positions to ratchet.

Lot sizing runs through CalculatePositionSize() with a 2% risk budget (RiskPercent = 2.0) and a FixedLotSize = 0.01 fallback when UseMoneyManagement is off. The risk lot formula is the standard (balance * riskPct) / (slPips * pipValuePerLot), floored to the symbol's SYMBOL_VOLUME_STEP and clamped to SYMBOL_VOLUME_MIN and SYMBOL_VOLUME_MAX. A separate hard ceiling InpMaxLot = 5.0 blocks any single trade from exceeding five lots regardless of account size. Order execution uses CTrade with GetAllowedFilling() probing SYMBOL_FILLING_MODE and selecting FOK > IOC > RETURN in that priority. OpenPosition() runs three attempts with a 500-millisecond sleep and a slippage ramp that grows by 5 points per attempt (3, 8, 13); only REQUOTE/TIMEOUT/PRICE_OFF/PRICE_CHANGED class failures are caught and the dev value is widened accordingly.

Risk hardening is split between the deal-stream watcher and the per-tick safety gate. OnTradeTransaction() listens for TRADE_TRANSACTION_DEAL_ADD events, looks up the new deal in a 90-day HistorySelect() window, filters to deals whose DEAL_MAGIC matches the EA and whose entry is DEAL_ENTRY_OUT, and accumulates DEAL_PROFIT + DEAL_SWAP + DEAL_COMMISSION into g_realizedToday and g_realizedWeek. Consecutive losses increment g_consecLosses and set a g_cooldownUntil timestamp 30 minutes in the future; a single win resets the counter. IsSafeToTrade_EX10015() runs on each new bar and applies eight independent gates in this order: capital cap opt-in (if InpCapAmount > 0 and equity exceeds the cap, no new trades), capital floor (InpCapFloor = 50 blocks entries if equity drops below), equity floor (InpMinEquityPercent = 95% of initial balance), daily loss limit (3% of effective capital), weekly loss limit (twice the daily limit, i.e. 6%), consecutive-loss cooldown (3 losses → 30-minute lockout), per-day trade limit (50), market-open check (closes for Friday 22:00+ GMT, Saturday, and Sunday before 22:00 GMT), session filter (London 7-16 GMT or New York 12-21 GMT, with Asia 0-7 GMT avoided by default), and an optional news-time filter (15-minute window around the London or NY open, disabled by default). The GMT offset is auto-detected on OnInit by scanning PERIOD_H1 for the first weekend gap and computing hour - 22, with InpServerGMTOffset available to override.

The EA's OnTester() criterion is (netProfit * profitFactor) / (1 + relativeDrawdownPct) with a 30-trade minimum; backtests with fewer trades return 0.0, which means the strategy tester will rank the EA as having zero edge when the sample is too small. The two function TryClose_EX10015() and TryModify_EX10015() handle close and modify with three attempts at 200ms and 100ms sleeps respectively, retrying only on the four transient retcodes (REQUOTE, TIMEOUT, PRICE_OFF, PRICE_CHANGED) and breaking on permanent rejections.

In backtest, expect a low win rate — CCI zero-line crosses on M15 are common but most fail to extend into 2× ATR of follow-through. The trades that do extend pay for the rest. Position duration is short, typically a few M15 bars before either the trailing stop activates or the CCI opposite extreme fires. The 1:2 reward-to-risk setup means the strategy needs roughly a 35-40% win rate to break even before spread and slippage, and the trailing stop is what keeps the early-momentum winners from giving back their gains. A $100 minimum deposit is feasible for micro-lot testing, but a $1,000+ balance is more honest for live use given the 2% per-trade risk and the EA's willingness to fire multiple signals per session.