Building Your Own Expert Advisor (EA): The Complete MQL5 Algorithmic Development Guide for MetaTrader 5

What Is The Architecture of MQL5 Expert Advisors?

MetaTrader 5 (MT5) relies entirely on MQL5 (MetaQuotes Language 5). Think of it as a super-fast, event-driven cousin of C++. Unlike standard scripts that just run top-to-bottom, an Expert Advisor (EA) sits there waiting. A new tick arrives? An order book shifts? The EA instantly fires off the right event handler.

MQL5 actually compiles straight to machine code. That means you get sub-millisecond execution times, dodging the sluggish latency you typically get with interpreted languages like Python. Professional quants live and breathe MQL5. It lets them spin up fully autonomous bots that manage state and blast off trades instantly.

You generally construct an EA around three main functions: OnInit(): This fires the moment you drop the EA onto a chart. You use it to allocate memory, fire up indicators, and check account specs. OnDeinit(const int reason): Pull the EA off the chart? This runs. It cleans up the mess, tossing out old indicator handles and wiping visual chart objects. * OnTick(): This is the beating heart of your EA. It runs every single time a fresh price tick rolls in. It evaluates the logic and pulls the trigger.

How Does Mathematical Structure of Moving Average EA Crosses Work?

If you want a basic trend-following rig, look at the Dual Moving Average Crossover. We track a fast-moving average (Fastt) and a slow one (Slowt). When the fast line jumps over the slow line, boom. Momentum shift.

But we don't want the bot blindly firing trades while the averages are simply far apart. We wait for the exact moment of crossing by checking the prior complete bar (t-1) and the one right before it (t-2):

Math-Backed Risk Management & Position Sizing Stop using fixed lot sizes! Real EAs calculate volume on the fly based on account equity and your risk tolerance:

📓 Position Lot Size Formula Lot Size = (Account Balance × Risk Percent)/(Stop Loss in Points × Tick Value)

• •Risk Percent: The slice of your account you're willing to lose (e.g., 0.01 or 1.0%).
• •Stop Loss in Points: Distance from your entry to the stop loss (e.g., 500 points for a 50-pip Gold stop loss).
• •Tick Value: How much a single point shift is worth (your broker feeds you this).
• •Step-by-Step Example: Imagine you have a \50,000 account, risking 1.0\% (0.01). You set a 500 point stop loss. Gold's tick value is \0.10 per point:
• •Risk Amount = \50,000 × 0.01 = \500
• •Lot Size = (\500)/(500 × 0.10) = (\500)/(50) = 10.0 Lots

How Does High-Quality MQL5 Expert Advisor Script Work?

Here's a remarkably clean MQL5 script. It handles the dual SMA logic, dynamically scales your lot sizing, and uses standard libraries to keep execution lightning fast:

#property copyright "AlphaFinance EA Desk"
#property link      "https://alphafinancehub.app"
#property version   "2.00"

// Import Native Trade Library #include <TradeTrade.mqh> CTrade trade;

// Global Inputs (Customizable parameters in MT5 properties) input int FastPeriod = 12; // Fast moving average period input int SlowPeriod = 26; // Slow moving average period input double RiskPercent = 1.0; // Percentage of account balance to risk per trade input double StopLossPoints = 500; // Stop Loss in points (e.g. 50 pips) input double TakeProfitPoints = 1500; // Take Profit in points (e.g. 150 pips)

// Global Indicator Handles int fastmahandle; int slowmahandle;

// Initialization int OnInit() { // Configure Trade Request parameters trade.SetExpertMagicNumber(102930); trade.SetDeviationInPoints(20); // Create Moving Average Indicator Handles fastmahandle = iMA(Symbol, Period, FastPeriod, 0, MODESMA, PRICECLOSE); slowmahandle = iMA(Symbol, Period, SlowPeriod, 0, MODESMA, PRICECLOSE); if(fastmahandle == INVALIDHANDLE || slowmahandle == INVALIDHANDLE) { Print("Failed to initialize indicator handles."); return(INITFAILED); } return(INITSUCCEEDED); }

// Cleanup void OnDeinit(const int reason) { // Release indicator handles back to the terminal engine memory IndicatorRelease(fastmahandle); IndicatorRelease(slowmahandle); }

// Tick Engine void OnTick() { // Check if we have open positions already on this symbol to prevent multiple entries if(PositionsTotal() > 0) { for(int i = 0; i < PositionsTotal(); i++) { if(PositionGetSymbol(i) == Symbol && PositionGetInteger(POSITIONMAGIC) == 102930) { return; // Active trade exists, yield execution } } }

double fastmavalues[]; double slowmavalues[]; // Set indexing direction from current bar backward ArraySetAsSeries(fastmavalues, true); ArraySetAsSeries(slowmavalues, true); // Copy the last 3 values of the buffers (bar 0, bar 1, bar 2) if(CopyBuffer(fastmahandle, 0, 0, 3, fastmavalues) < 3 || CopyBuffer(slowmahandle, 0, 0, 3, slowmavalues) < 3) { Print("Error copying indicator buffers."); return; } // Check crossover conditions on complete bars bool buycondition = (fastmavalues[1] > slowmavalues[1]) && (fastmavalues[2] <= slowmavalues[2]); bool sellcondition = (fastmavalues[1] < slowmavalues[1]) && (fastmavalues[2] >= slowmavalues[2]); // Calculate dynamic position size based on SL points and Account Balance double balance = AccountInfoDouble(ACCOUNTBALANCE); double tickvalue = SymbolInfoDouble(Symbol, SYMBOLTRADETICKVALUE); double riskamount = balance (RiskPercent / 100.0); double lotsize = NormalizeDouble(riskamount / (StopLossPoints tickvalue), 2); // Clamp lot size limits to broker specifications double minlot = SymbolInfoDouble(Symbol, SYMBOLVOLUMEMIN); double maxlot = SymbolInfoDouble(Symbol, SYMBOLVOLUMEMAX); if(lotsize < minlot) lotsize = minlot; if(lotsize > maxlot) lotsize = maxlot; // Trade Execution if(buycondition) { double ask = SymbolInfoDouble(Symbol, SYMBOLASK); double sl = ask - (StopLossPoints Point); double tp = ask + (TakeProfitPoints Point); trade.Buy(lotsize, Symbol, ask, sl, tp, "MQL5 SMA Cross Buy"); } else if(sellcondition) { double bid = SymbolInfoDouble(Symbol, SYMBOLBID); double sl = bid + (StopLossPoints Point); double tp = bid - (TakeProfitPoints Point); trade.Sell(lotsize, Symbol, bid, sl, tp, "MQL5 SMA Cross Sell"); } } ```

How Does MQL5 Expert Advisor Optimization Matrix Work?

Want to know if you should stick to MQL5 or route things through a Python API? Here is a quick breakdown of what you actually get: