Articles on trading system automation in MQL5

This article introduces the DA-CG-LSTM algorithm, which offers new approaches to time series analysis and forecasting. It explains how innovative attention mechanisms and model flexibility can improve forecast accuracy.

The article explores the revolutionary Mamba/SSM neural network architecture for financial time series forecasting. We will consider a complete MQL5 implementation of a modern alternative to Transformer with linear complexity O(N) instead of quadratic O(N²). Selective State Space Models, hardware-aware optimizations, patching techniques, and advanced AdamW training methods are covered in detail. Practical test results showing an increase in accuracy from 62% to 71% while reducing training time from 45 to 8 minutes are included. A ready-made trading EA with auto learning and adaptive risk management for MetaTrader 5 is presented.

This article presents CTradeJournal, a self-contained MQL5 class for live tracking of open positions at tick frequency. It maintains MAE, MFE, and initial risk in money, calculates the R-multiple when a position closes, and writes a complete CSV record. The text explains the design choices, provides the implementation, and shows simple EA integration so you can analyze entries, stop placement, and outcome distribution.

We invite you to explore the practical implementation of a sparse mixture of experts block for time series in the OpenCL computing environment. This article provides a step-by-step explanation of how masked multi-window convolution works, as well as how gradient-based training is organized in the presence of multiple information streams.

This article shows how to configure a black-box model to automatically uncover strong trading strategies using a data-driven approach. By using Mutual Information to prioritize the most learnable signals, we can build smarter and more adaptive models that outperform conventional methods. Readers will also learn to avoid common pitfalls like overreliance on surface-level metrics, and instead develop strategies rooted in meaningful statistical insight.

The article delivers an object-based MQL5 implementation that detects weekend gaps from time discontinuities and renders them directly on the chart. It manages graphical objects, tracks state transitions (fresh, partial, reaction, filled), and preserves completed gaps as historical zones. The result is a reproducible framework for monitoring how price revisits and fills weekend gap structures.

We demonstrate a practical customization path for a built-in MetaTrader 5 EA using BullishBearish MeetingLines Stoch. The workflow covers baseline testing in the Strategy Tester, parameter optimization, and code-level changes. Two modifications are implemented: exposing Stochastic thresholds as inputs and adding an optional Moving Average filter to limit counter‑trend signals. The article includes the full modified code for replication.

Today we will learn a technique that can help us a lot in different stages of our professional life as a programmer. Often it is not the platform itself that is limited, but the knowledge of the person who talks about the limitations. This article will tell you that with common sense and creativity you can make the MetaTrader 5 platform much more interesting and versatile without resorting to creating crazy programs or anything like that, and create simple yet safe and reliable code. We will use our creativity to modify existing code without deleting or adding a single line to the source code.

Our next focus in these series on ideas that can be rapidly prototyped with the MQL5 Wizard, is a Custom Trailing class that uses the Blooming Filter. Trailing Stop systems are an optional but very resourceful part to any trading system that we want to explore more in these series besides the traditional Entry Signals.

The Actor–Director–Critic framework is an evolution of the classic agent learning architecture. The article presents practical experience of its implementation and adaptation to financial market conditions.

The article presents a production-oriented Order Block engine for MQL5 packaged as an include class, it validates zones via displacement and market structure break, maintains mitigation state only on closed bars, and avoids heavy copies by passing data by reference. A diagnostic indicator plots zones, and an EA gates logic to new bars for stable performance and reproducible tests.

We add a descriptor-driven property stack and a floating ribbon that binds to the current selection on the drawing layer. The article covers the descriptor list for each tool, the engine get/set API with snapshot-and-restore live preview, and widget renderers for color, opacity, line width, line style, fonts, and level visibility. You get in-place, real-time editing of object appearance via a compact, draggable panel.

We implement a five-stage MQL5 pipeline that quantifies market structure, liquidity interaction, and price behavior on four timeframes, then resolves them into a 0–100 Market Intent Score. Decision states (WAIT/WATCH/ACTION) are driven by explicit weights plus hard gates. The analytical core feeds a concise dashboard and, when AutoTrade is on, an execution layer with entry zones, invalidation and liquidity‑based targets.

Preprocessing is a powerful yet quickly overlooked tuning parameter. It lives in the shadows of its bigger brothers: optimizers and shiny model architectures. Small percentage improvements here can have disproportionately large, compounding effects on profitability and risk. Too often, this largely unexplored science is boiled down to a simple routine, seen only as a means to an end, when in reality it is where signal can be directly amplified, or just as easily destroyed.

We add a tabbed settings window opened from the ribbon and bound to the selected object. The tabs — Style, Text, Coordinates, and Visibility — are built from the same descriptor system, with scrolling, per-level rows, and shared color/width/style popovers. The article covers layout, rendering, interaction, and inline price/time and numeric editing. You get one place to edit every property with live preview and commit-or-discard on close.

In this article, we will get acquainted with an algorithm that uses closed-form policy improvement operators to optimize Agent actions in offline mode.

In this article, we explore the butterfly curve, a parametric mathematical equation, and render it visually on a MQL5 canvas. We build an interactive display with a draggable, resizable canvas window, supersampled curve rendering, gradient backgrounds, and a color-segmented legend. By the end, we have a fully functional visual tool that plots the butterfly curve directly on the MetaTrader 5 chart.

One point to note: although the service code is not included in this article and will only be provided in the next one, I'll explain it since we'll be using that same code as a springboard for what we're actually developing. So, be attentive and patient. Wait for the next article, because every day everything becomes more interesting.

Your Expert Advisor runs clean on your demo, then throws errors on a client's broker and quietly stops trading - and the code never changed. What changed is the broker's rulebook. This first article of the Broker Reality Check series builds a diagnostic EA that reads every relevant symbol trading condition - filling policy, stops and freeze levels, volume step, trade mode, swap and the triple-swap day - and flags the ones that silently break EAs, in plain language. It shows a green/amber/red panel, prints a report and dumps every Market Watch symbol to CSV, so you see why an OrderSend fails (10030, invalid stops, invalid volume) before it costs you a trade.

Learn how to create an EX5 module of exportable functions that seamlessly query and save data for the most recently filled pending order. In this comprehensive step-by-step guide, we will enhance the History Management EX5 library by developing dedicated and compartmentalized functions to retrieve essential properties of the last filled pending order. These properties include the order type, setup time, execution time, filling type, and other critical details necessary for effective pending orders trade history management and analysis.

We expand the capabilities of the MetaTrader 5 butterfly curve canvas by adding multi-layered wing fills, vein lines, scale dots, and a full body (abdomen, thorax, head, eyes, antennae). This article implements polygon fills with vertical and radial gradients, as well as filled circles and ellipses, all using supersampling antialiasing. You will also receive reusable MQL5 helper functions and a rendering order that transforms a simple curve into a customizable, detailed chart illustration.

The article describes an approach to trade labeling using oscillators for machine learning models. This eliminates look-ahead bias. It has been shown that this type of labeling does not lead to model overfitting, and the strategies continue to perform well over the long term.

The article provides production-ready entropy estimators (Shannon, plug-in, Lempel–Ziv, Kontoyiannis) operating on tick-rule–encoded sequences. It resolves three correctness and performance issues in the original code, verifies outputs against chapter references, and extends encoding with quantile and sigma options. Users gain reproducible results and markedly improved computation speed for large bar sets.

This article presents a circuit breaker for MQL5 that monitors combined daily P&L (realized plus floating) on every tick and compares it to a configured loss limit. On breach, it closes positions, cancels pending orders, and activates a HALTED state that blocks further order submission in the EA until server‑time midnight. The package provides a chart dashboard, a demo Expert Advisor, a verification script, and notes on extending the halt signal across EAs.

Breadth First Search (BFS) uses level-order traversal to model market structure as a directed graph of price swings evolving through time. By analyzing historical bars or sessions layer by layer, BFS prioritizes recent price behavior while still respecting deeper market memory.

A hybrid architecture based on Llama 3.2 and SEAL is being tested on eight currency pairs (M15), with forward-period data isolation and information leakage control. The methodology combines adversarial self-play, curriculum learning, and class balancing to ensure stable training. The experiments confirm the gap between forecast accuracy and actual returns, providing readers with practical guidelines for testing strategies and accurately assessing their generalizability.

In offline learning, we use a fixed dataset, which limits the coverage of environmental diversity. During the learning process, our Agent can generate actions beyond this dataset. If there is no feedback from the environment, how can we be sure that the assessments of such actions are correct? Maintaining the Agent's policy within the training dataset becomes an important aspect to ensure the reliability of training. This is what we will talk about in this article.

We invite you to explore the original implementation of the K²VAE framework — a flexible model capable of linearly approximating complex dynamics in latent space. This article demonstrates how to implement key components in MQL5, including parameterized matrices and how to manage them outside standard neural network layers. This material will be useful for anyone looking for a practical approach to building interpretable time-series models.