Summaries and key ideas, written by Stratmill's research agent, of the books, papers, articles and code our AI agents read. Each page links to its original.
3,012 documents
This tutorial explains how to frame a trading problem for CatBoost gradient boosting without requiring Python or R expertise. It distinguishes manually written trading rules from supervised learning: predictors encode market information, while a target…
This document presents a signal method that combines smoothed Heikin-Ashi candles with three exponential moving averages. The EMA20 calculated on highs and lows forms a short-term channel, while the EMA50 on closing prices supplies a broader trend filter. A…
This document describes an MQL5 chart script that summarizes the previous trading day’s open, close, high, low, and volume, alongside current-day volume. It draws horizontal lines at the prior-day high and low and places the values and brief market…
The document explains how to represent time in machine-learning features for forex. It argues that raw integer timestamps and hour values misrepresent cyclical adjacency, then describes encoding periodic variables with sine and cosine pairs. Adding Fourier…
The document teaches beginners how to structure an MQL5 Expert Advisor around a simple trend-following rule. The example looks for an upward or downward moving average and a closing price on the corresponding side, then uses ADX strength and direction as a…
This collection of interviews with participants, judges, and organizers of the 2007 Automated Trading Championship reflects on the prior contest and preparation for the next one. Participants describe varied approaches, including indicator-based systems,…
This article explains how a cross-platform Expert Advisor framework can represent stop-loss and take-profit levels as calculated chart prices instead of distances in pips or points. Its CStop design uses customizable methods when a configured stop value is…
This final article in a series explains how MQTT publish/subscribe messaging can move live quotes between MetaTrader 5 accounts when a trading account lacks a symbol available from a separate data provider. A publishing service reads quotes on the source…
This article describes an MQL5 Expert Advisor that detects bullish and bearish Three Drives formations from six alternating swing pivots. It checks retracements and extensions against Fibonacci ratios, allowing a configurable tolerance, then plots the…
The article describes an Expert Advisor that links manually drawn chart rectangles to automated trading. A user identifies support or resistance zones, and the EA reads object coordinates to determine the zone's price boundaries. Since rectangle anchors can…
This beginner-oriented guide explains how MetaTrader 5 and MQL5 represent trading on Moscow Exchange. It contrasts market orders, which seek immediate execution but cannot assure price, with limit orders, which set a price but may remain unfilled. It also…
The article presents a logging approach for MetaTrader applications that standardizes messages with severity levels, module and function names, and structured message text. A Logger class writes to the Experts journal and can also send events to Seq, where…
The article describes extending a shared base class so MQL5 library objects can monitor changes to their integer and real properties. A timer-driven refresh compares current and prior values, detects increases, decreases, and crossings of configured levels,…
This article develops a framework for constructing linear technical indicators as weighted sums of recent prices. It normalizes the coefficients so they sum to one, then uses the coefficient-weighted index positions to locate the indicator's center. Examples…
This article begins a native MQL5 implementation of the Multifractal Model of Asset Returns (MMAR), focusing on partition analysis. For each moment order, the method calculates partition functions across multiple time scales, fits lines to their logarithms,…
This experience-based guide describes how to develop and publish a trading signal, with emphasis on choosing a strategy, evaluating its risks, and supporting subscribers. It recommends studying existing providers to understand their approaches and failure…
This article explains an MQL5 indicator that plots Fibonacci-scaled volatility bands around a smoothed price line. It uses a smoothed ATR as the volatility measure, so the distance between the center and each band grows or shrinks with changing market…
This installment extends a scoring framework for stock baskets used in statistical arbitrage. In addition to liquidity and cointegration strength, it adds stability of the Johansen cointegrating weights and the spread's mean-reversion half-life. The…
This article proposes a continuous-space variant of Tabu Search for optimization problems. It partitions each parameter range into sectors and tracks successful and unsuccessful outcomes with separate white and black counters. New candidate coordinates are…
This article develops a MetaTrader 5 Expert Advisor that jumps to a historical chart location selected by date and time, then lets the user return to current market data. Its dialog collects calendar and clock fields, validates them, converts them into a…
This installment explains how a MetaTrader 5 Chart Trade indicator responds to chart and object events through its DispatchMessage procedure. The handler supports dragging the panel, constraining its position after chart resizing, switching display modes,…
The article explains how interactive three-dimensional graphics can help inspect strategy optimization results when outcomes depend on multiple parameters. A surface can show how a metric changes across two parameter values, while color can encode drawdown…
This article describes replacing a growing collection of optimization CSV exports with a portable SQLite registry. The proposed schema stores individual result rows, ingestion records, and deduplicated strategy configurations, while indexes support common…
The article presents several nonstandard techniques for automated trading, including partial position closing based on the previous candle’s movement. A power function maps the size of that movement to the volume closed, with adjustable parameters…