MQL5 Algo Trading

L1 trend filtering targets trend extraction while suppressing noise and automatically detecting slope breakpoints. The output is piecewise linear, which avoids the lag and slope blurring typical for moving averages and the Hodrick–Prescott filter. The method is formulated as convex optimization with an L1 penalty on second differences. Sparsity in Dx forces most curvatures to zero, leaving a small set of structural breakpoints. The regularization parameter λ controls segment count; at λ≥λmax the result becomes strictly linear. Practical MQL5 implementation includes λmax computation, relative tuning via λ=coef·λmax, and linear-time filtering. Delivered components cover trend and slope indicators, multiple residual-volatility variants, and testing hooks for MA, MACD, ADX, and EMA signal filtering. 👉 Read | Calendar | @mql5dev #MQL5 #MT5 #Indicator

The December 2021 TASC article “The DMH: An Improved Directional Movement Indicator” outlines John Ehlers’ update to classic directional movement by applying Hann windowing. Directional movement has been used in technical analysis for decades. The original formulation by J. Welles Wilder was constrained by the compute limits of its era, which favored simpler calculations. Ehlers argues that current tooling makes a more modern implementation practical. DMH typically defaults to Wilder’s 14-bar setting, but the period is treated as a parameter that can be selected by the trader or optimized when embedded in a systematic strategy. This version adds optional visual states: color shifts on slope change, on zero-line crossover, or no color changes to match the classic presentation. Period sensitivity testing is recommended before using DMH-derived signals. 👉 Read | CodeBase | @mql5dev #MQL4 #MT4 #Indicator

This article extends the DoEasy MQL5 DOM toolkit by introducing two core abstractions: a Depth of Market snapshot object and a per-symbol snapshot series. Each OnBookEvent() call pulls MqlBookInfo via MarketBookGet(), converts entries into typed order objects, and stores them inside a snapshot container built on CArrayObj. A key enhancement is adding a millisecond timestamp property to every DOM order, enabling consistent time-based search and sorting across snapshots even though MqlBookInfo has no native time field. The Select service is updated to support these new criteria, and order descriptions gain optional symbol output to avoid redundancy inside snapshot logs. The snapshot series class acts like a real-time timeseries: it only accumulates live data, enforces a configurable max size (default 200k), and supports retrieving snapshots by times... 👉 Read | AlgoBook | @mql5dev #MQL5 #MT5 #AlgoTrading

PrecisionSniper is an MT5 indicator that generates long/short signals from a weighted, multi-condition scoring model. Up to eight checks run per bar, and a signal prints only when the total meets a configurable minimum. Outputs are graded A+, A, B, or C, with filters to suppress lower grades and a cooldown to reduce clustering. Core inputs combine three EMAs (fast/slow/trend) with RSI momentum, MACD histogram expansion, ADX strength with DI alignment, VWAP position, tick-volume surge, and an optional higher-timeframe EMA bias. HTF bias carries the highest single weight and can be used to force alignment with the larger trend. Trade visualization is built in: entry, structure-based or ATR-based stop, three risk-reward targets, and a ratcheting trailing stop that advances at TP milestones. An on-chart dashboard shows current regime and a historical sig... 👉 Read | Signals | @mql5dev #MQL5 #MT5 #Indicator

Directional retail systems often concentrate on forecasting price moves, leaving exposure to macro surprises. A common workaround is using correlation between pairs, but correlation is not tradeable by itself. Assets can remain highly correlated while the spread keeps widening, producing persistent drawdowns. Institutional stat-arb focuses on cointegration and market-neutral positioning, where the spread between linked assets is expected to mean-revert. A practical implementation is tracking a spread Z-score rather than predicting direction. Key mechanics: logarithmic spread via ln(A) minus ln(B) to normalize scale and volatility differences, rolling Z-score to quantify deviation in standard deviations, and time-series synchronization to align both instruments even with gaps in broker data. Execution: select a cointegrated pair, wait for Z-score extr... 👉 Read | VPS | @mql5dev #MQL5 #MT5 #StatArb

DoEasy library update adds two blocks: a DOM snapshot series collection and initial support for MQL5.com Signals objects. DOM work is consolidated via CMBookSeriesCollection, storing per-symbol CMBookSeries lists with bounded history, auto refresh on BookEvent, lookup by symbol, and access to snapshots by index or millisecond timestamp. Engine.mqh is extended with the new collection, helpers, and an OnBookEvent handler that routes updates to the correct symbol series. A new “other library classes” section starts with CMQLSignal, modeling a single signal source. It defines integer/double/string property enums, sorting criteria, and wrappers over SignalBaseGet* and SignalInfoGet* for retrieval, comparison, and selection workflows. 👉 Read | Quotes | @mql5dev #MQL5 #MT5 #AlgoTrading

XANDER Pulse Candles colors price bars using an internal RSI-based momentum oscillator (period 14 by default), removing the need for a separate oscillator pane and keeping bias visible on the main chart. Color states map momentum and candle direction: Spring Green signals strong bullish bias (momentum above threshold with a bullish close). Deep Sky Blue marks weak bullish bias (momentum bullish with a bearish close). Orange Red signals strong bearish bias (momentum below threshold with a bearish close). Orchid marks weak bearish bias (momentum bearish with a bullish close). Dim Gray indicates neutral conditions. Three modes are provided: center-line (above/below 50), extreme zones (overbought/oversold activation), and direction (rising/falling momentum). Usage notes include tracking consecutive strong-bias candles, treating mixed colors as fade warnings, ... 👉 Read | AlgoBook | @mql5dev #MQL5 #MT5 #Indicator

A common rule-based approach in indicator-driven trading uses arrow signals for entries and a central reference line for exits on an existing position. An arrow is treated as a directional trigger to open a long or short trade. The middle line is then used as the close condition, typically when price crosses it or when the indicator state returns to neutral. This setup separates entry logic from exit logic and can reduce churn from frequent signal flips. Practical use still requires clear definitions for confirmation, stop-loss placement, and behavior during ranging markets, since both arrows and midline crosses can lag or repaint depending on implementation. 👉 Read | AppStore | @mql5dev #MQL5 #MT5 #Strategy

Fractal-based Algorithm (FBA, 2017) tackles continuous optimization by building a self-similar, hierarchical partition of the search space. Instead of sampling uniformly, it repeatedly splits regions and allocates work where good solutions cluster, balancing global exploration with local refinement. The workflow: generate an initial population, keep the best fraction of points, estimate each subspace’s “potential” by counting how many strong points fall inside, then mark the top-ranked regions and subdivide them. A small share of mutated points injects randomness to reduce stagnation. The MT5 implementation (C_AO_FBA) models subspaces with bounds, hierarchy level, parent links, and a normalized rank. New populations are generated proportionally to subspace ranks, with safeguards for high dimensions and a hard cap on subspace count to control resources. 👉 Read | Freelance | @mql5dev #MQL5 #MT5 #algorithm

Repeated high/low scans inside trailing stops turn many MT5 EAs into O(N) workloads, creating latency and exit drift on VPS deployments. The article switches to a data-first design: deterministic, cleaned history plus constant-time extrema queries. A Sparse Table precomputes min/max over power-of-two ranges, built once at init (or refresh) and queried in O(1) via two overlapping blocks. This replaces per-bar lookback loops with indexed range lookups suited to frequent trailing updates. For reproducible inputs, a Python Polars pipeline cleans broker data (gap handling, interpolation, Hampel outlier filtering, log-return validation) and writes to SQLite. MQL5 reads timestamped, targeted SELECT results to build the table and feed trailing logic without terminal-history inconsistencies. Trailing decisions add an excursion validator: stops update only wh... 👉 Read | Signals | @mql5dev #MQL5 #MT5 #AlgoTrading

GoertzelBrain tackles the main weakness of classic cycle detection: knowing a dominant period and amplitude still doesn’t answer “long or short” in non-stationary markets. It keeps Goertzel’s efficient single-frequency DFT, but adds context with a self-training neural ensemble. Per bar, it builds a 7-feature vector from spectrum state (dominant period, amplitude, normalized confidence), spectral dynamics (period/amplitude slopes), plus price slope and volatility. Ten small MLPs (7-12-1) train online via single-sample SGD and are averaged to reduce instability. The output becomes a directional filter: long only when the ensemble is above zero and rising; short when below zero and falling. A hidden buffer is exposed via iCustom() for EA integration, with careful heap allocation to avoid GoertzelCycle pointer lifetime bugs. 👉 Read | Quotes | @mql5dev #MQL5 #MT5 #AITrading

The EA maintains a daily session schedule by calculating range start, range end, and trading close times, then resetting state for the next cycle. During the range window it aggregates minute highs and lows, tracks the session maximum and minimum, and updates a chart rectangle to reflect the current consolidation boundaries. After the range window closes, it evaluates the close of the latest candle (not necessarily aligned with the range end time) against those boundaries. On a confirmed break, it sends a buy or sell order with take-profit set to the measured range size and stop-loss placed at the opposite boundary. Example setup: EURUSD, 2026.01.01–2026.03.03; Range Start 05:00, Range End 11:00, Trading End 22:00. 👉 Read | AlgoBook | @mql5dev #MQL5 #MT5 #EA

QuantumAlgo adds execution and risk controls aimed at reducing manual inputs and sizing errors. Automatic position sizing calculates volume from stop loss distance and a predefined risk percentage, with real-time display of total capital at risk. Twin order splitting sends two 50/50 orders, supporting a 1:1 RRR target on the first and an open-ended runner on the second. Margin metrics show free margin and required margin before order placement. Strategy filters and alerts include Bill Williams Wise Man 1–3 signals and Ichimoku logic. A symmetrical pair filter suppresses correlated-pair alerts when exposure already exists. Additional filters use Alligator, Kijunsen, or Senkou Span A, plus line/MA/Ichimoku cross notifications. Trade handling includes global break-even, optional trailing/exit management, session range boxes, an Ichimoku market scanner, bulk c... 👉 Read | CodeBase | @mql5dev #MQL5 #MT5 #EA

ASQ Session Manager v1.2 is a free, open-source MQL5 session library aimed at adding session awareness to EAs. It detects Sydney, Tokyo, London, and New York with automatic broker GMT offset, tracks session phases, and resolves overlaps by priority. Core logic includes kill zone identification (London/NY opens, Asian, London close, overlap), Asian range breakout detection for 00:00–06:00 GMT with false-break filtering and 1.5x expansion targets, plus ATR-relative volatility scoring versus a 20-day baseline. Safety gates cover weekends, late Friday, early Monday, and NFP first-Friday blocking. Integration is a small include-based API (Initialize/Update/IsTradingAllowed/HasSignal). A demo indicator visualizes sessions, kill zones, Asian range state, volatility, countdowns, and entry/SL/TP for three built-in signal types. Files: .mqh library and .mq5... 👉 Read | VPS | @mql5dev #MQL5 #MT5 #AlgoTrading

XANDER Grid XAUUSD is a bidirectional grid EA for gold trading on MetaTrader 5. It places buy orders after a bullish candle and sell orders after a bearish candle. New grid entries are added only after price moves a configured distance from the last entry, combining candle direction with spacing to limit overtrading in range conditions. It runs two independent grids (buy and sell). When multiple positions exist on one side, it calculates a weighted average entry and applies a shared take profit to the basket. Closing options include an average TP for the group or partial reduction of the worst position to manage exposure. Risk controls include a daily profit target that flattens all positions and a max floating drawdown filter that pauses new entries. Configuration covers TP, grid step, basket profit threshold, starting and max lot caps, slippage, and magic n... 👉 Read | Freelance | @mql5dev #MQL5 #MT5 #EA

Fractional differentiation solves a core preprocessing trade-off in financial ML: returns enforce stationarity but destroy price-level memory, while raw prices preserve context but break stationarity. By differencing with a real-valued d between 0 and 1, the series can pass stationarity tests while retaining more of the original structure needed for mean-reversion and trend models. The method uses binomial-series weights generated efficiently via a recurrence. Two implementations matter: expanding windows can introduce artificial drift as the lookback grows; fixed-width fractional differentiation (FFD) truncates weights by a threshold, applies a consistent window, and avoids drift. The practical workflow searches for the smallest d that achieves stationarity using the ADF test, typically via binary search, while monitoring correlation to quantify ret... 👉 Read | AlgoBook | @mql5dev #MQL5 #MT5 #Strategy

This MT5 approach rotates risk capital by time of day instead of treating all hours equally. It assigns smaller budgets to Asia, more to London breakouts, and the largest share to New York volatility, focusing each session on its historically strongest symbols and avoiding off-hours overtrading. The core is a session-aware allocation engine separated from entry logic. A daily risk cap (percent of balance) is split per session (equal or manual). Each trade sizes lots from remaining session budget using stop-loss distance and pip value, enforcing daily and session limits. Execution uses a volatility-confirmed breakout model. The EA tracks prior session highs/lows, validates breakouts with ATR “volatility stops,” and switches between momentum trades and fades. It resets budgets daily, preserves winning positions across sessions, and closes only losers. 👉 Read | Freelance | @mql5dev #MQL5 #MT5 #AlgoTrading

Tree-based classifiers often output overconfident probabilities, especially in noisy financial labels, so accuracy can look acceptable while probability estimates are systematically too extreme. That miscalibration feeds directly into bet sizing and Kelly-style sizing, creating oversized positions, deeper drawdowns, and weaker geometric growth. Calibration is evaluated with reliability diagrams plus Brier score (overall usefulness), ECE (average gap), and MCE (worst-case tail risk), with bootstrap confidence bands to reflect limited independent samples. Two practical calibrators are compared: isotonic regression (non-parametric, rank-preserving, best with enough data) and Platt scaling (sigmoid, steadier with small samples but less flexible). The key engineering constraint is avoiding temporal leakage. Calibration is fit on out-of-fold predictions p... 👉 Read | NeuroBook | @mql5dev #MQL5 #MT5 #AITrading

Matrix arbitrage reframes FX as a consistency problem across 8 majors. The EA builds an 8x8 currency relationship matrix and derives each pair’s fair value from all available cross-rates, then trades when the live quote deviates beyond a configurable threshold. Implementation details focus on performance and robustness: the matrix is stored in a 1D array, refined iteratively, and blends direct quotes with implied cross rates using tested weights to balance stability vs sensitivity. Signals are simple: undervaluation triggers buys, overvaluation triggers sells. Risk is normalized per symbol via pip value, volatility, and broker lot limits, plus an optional global profit target that closes all positions and halts trading. A real-time imbalance matrix on-chart helps validate opportunities quickly. 👉 Read | VPS | @mql5dev #MQL5 #MT5 #AlgoTrading

BEC Trade Manager is an EA focused on active position control for the current symbol, with one-click actions for breakeven, trailing, partial close, stop-loss removal, profit-side and loss-side liquidation, plus quick scalp entries. Chart overlays include equity, floating P/L, daily profit, SL labels, and a basket breakeven preview. Core logic covers basket-level account breakeven with a configurable lock-profit offset, classic price-distance trailing, per-trade breakeven lock, and step-based position trailing from the current SL anchor. Execution tools include close-profit only, close-loss only, 50% close with breakeven protection, cancel stop losses, and close-all with pending-order cleanup. Filters support order type, magic number, and comment. Operational note: this is management automation, not a signal system. Outcomes depend on spread, execution, brok... 👉 Read | Signals | @mql5dev #MQL5 #MT5 #EA