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 article presents MQL5 measures for studying intraday futures volatility: realized and duration-based volatility, fractional and FIGARCH-inspired volatility, clustering, leverage asymmetry, jump intensity, and multifractal spectrum width. It revises an…
In this interview, trader Leonid Velichkovsky reflects on using neural networks in automated trading and challenges the idea that they produce exceptional profits by themselves. He describes them as trading systems whose adaptive capacity can help detect…
This article completes an MQL5 implementation of a hyperbolic latent diffusion framework for financial time series. It describes projecting inputs into hyperbolic space, mapping embeddings onto tangent planes, and generating centroids and curvature…
This article describes an analyzer for judging whether a backtested strategy’s profits are broadly distributed or depend on a few exceptional trades or days. It defines top-N trade shares against net and gross profit, the Gini coefficient across winners, an…
The article introduces Chaikin Volatility (CHV) as a measure of changes in the high-low price range. It explains that rising readings indicate faster price changes, while low readings indicate quieter conditions; CHV measures volatility rather than market…
This article explains BFO-GA, a population optimization method combining bacterial foraging with genetic operators. Bacteria represent candidate solutions and move through the search space using gradient-guided and random movements. Selection, crossover, and…
This article explains HiVT, a hierarchical Transformer approach originally designed to forecast trajectories of multiple road users in autonomous-driving scenes. It represents agent paths and lane segments as vectors, using relative positions to preserve…
The article presents a chart-based interpretation of the Murray system, derived from Gann-related ideas. It divides a chosen price range into eight equal parts, called an octave, then subdivides those intervals into smaller octaves for closer analysis. The…
This installment of a multi-currency Expert Advisor series outlines how to add a new strategy to an automated optimization workflow. It introduces SimpleCandles, a contrarian idea: after a configurable run of same-direction closed candles, open a position in…
This installment explains the internal design of an application for continuous walk-forward optimization. It uses the Model–View–ViewModel pattern to separate the graphical interface from application logic, with interfaces between the view model and model…
This article proposes balancing position sizes across instruments using each symbol’s tick value and average daily volatility. It combines those inputs into an estimate of risk per standard contract, then scales position volume relative to the highest…
The article evaluates a multi-currency Expert Advisor that applies a long-only averaging grid to US stocks. RSI readings trigger entries, positions are added as price moves against the trade, and the number of grid steps is capped; at the limit, existing…
This document describes a reusable MQL5 engine for managing volatility-based trailing stops. It derives stop distance from a simple average of True Range over completed bars, multiplied by a configurable factor. For long positions the stop can ratchet…
The article reviews alternatives to the Sharpe ratio and implements them for hypothetical equity curves generated from S&P 500 returns. Its drawdown measures include two versions of the Burke ratio, using either net profit or mean returns over an aggregate…
The article tests three simple forecasts for the next candle: repeat the previous candle’s direction, predict the opposite direction, or switch between those rules after a forecast error. Candle direction is defined by whether its close is above, below, or…
This article presents an integration pattern for storing and analyzing trading data from MQL4 or MQL5 Expert Advisors in Microsoft SQL Server. Since MetaTrader does not provide direct database-server access, the approach uses a C# class library that exports…
The article explains how to define bullish and bearish divergence and convergence by relating price and oscillator swings to possible reversal or continuation signals. It notes that common labels can be ambiguous, then proposes a more systematic…
The article addresses a failure mode in Expert Advisors that keep stop-loss and take-profit levels only in runtime memory: after a terminal shutdown, the EA stops monitoring, while the broker has no knowledge of those virtual levels. It extends an earlier…
The article measures how MetaTrader 5 tester runtime and memory use change as the number of strategy instances in a multi-currency Expert Advisor grows. It compares several tick simulation modes, estimates cost per instance per test day, and separates EA…
The article describes connecting a Python forecasting model to an MQL5 Expert Advisor through a TCP socket, so the model can be used in MetaTrader 5 historical tests. The EA sends recent bar data to a Python server, which runs inference and returns a…
The article describes how to reconstruct closed positions in MetaTrader 5, where historical positions are not directly exposed through the same functions used for open positions. It uses deal history as the source: each deal carries a position identifier,…
This installment explains a specialized way to compute the pseudoinverse used to estimate a two-parameter linear relationship for a single neuron. Rather than building a fully general matrix inversion routine, it exploits the structure of the input: a…
The article introduces fractional differentiation as a way to make financial price series more suitable for machine learning while retaining some information about past price levels. Ordinary returns are presented as stationary but memory-poor, while raw…
The article explains how MQL5 services run independently of charts and how to manage them in MetaTrader 5. A service starts through its OnStart handler; it can perform a one-time task and exit or remain active in a loop until stopped. The examples show…