VR RSI Robot is an MT5 expert advisor using RSI with multi-timeframe confirmation. Signals from H1 are validated against D1 to reduce false entries and align trades with stronger directional bias.

Entry rules are strictly defined and based on completed candles to avoid redraw. Buy requires RSI on both H1 and D1 to be above the oversold level (default 20), with the current RSI higher than the previous value. Sell requires RSI on both timeframes to be below the overbought level (default 80), with the current RSI lower than the previous value.

Position handling follows a one-position-per-symbol model using a magic number. If an opposite signal appears, the existing position is closed and a new one is opened in the latest direction.

Lot sizing is normalized at initialization by rounding to the broker’s volume step and enforcing the minimum volume, supporting consis...

👉 Read | Calendar | @mql5dev

#MQL4 #MT4 #EA

ExMachina Supply & Demand Zones v2.0 updates an MT5 indicator for automatic supply/demand zone mapping using impulse detection filtered by ATR. The scan covers the last 1000 bars and marks the base candle preceding a qualified directional move, then rates each zone by strength, freshness, and touch count.

Impulse qualification now supports multi-candle windows (default 3 bars). Range or body can be summed across 2–3 candles and compared to ATR; moves at or above 1.0x ATR are flagged, capturing sequences missed by single-candle logic. Nearby same-type zones can be merged to reduce overlapping rectangles.

Operational features include touch tracking, nearest fresh-zone distance in a dashboard, and two alert types: touch (entry into a fresh zone) and proximity (default 50 points). An optional dark chart theme and visual controls (midline, extension, fres...

👉 Read | AlgoBook | @mql5dev

#MQL5 #MT5 #Indicator

VR Rsi Robot for MetaTrader 5 trades RSI with multi-timeframe confirmation: H1 signals are validated by D1 using closed candles to avoid redraw. It holds one position per symbol, flips on opposite signals, and auto-adjusts lot size to broker limits.

👉 Read | Quotes | @mql5dev

#MQL5 #MT5 #EA

As codebases grow, call chains become harder to track, changes start touching multiple places, and maintenance becomes risky when context is missing. Structured separation is one practical response, and MVC remains a baseline approach.

Classical MVC splits responsibilities into Model (data and rules), View (presentation), and Controller (events and orchestration). Components stay decoupled via interfaces, limiting direct state changes and keeping responsibilities explicit.

For indicators, View maps cleanly to buffers and rendering, Model to calculation, Controller to handlers and inputs. For Expert Advisors, View aligns with order/position execution and reporting, while Model covers signals, sizing, and risk logic.

MVC adds overhead in trivial scripts, but pays off once projects span many files, multiple UI outputs, external data sources, or parallel deve...

👉 Read | Signals | @mql5dev

#MQL5 #MT5 #MVC

High-Performance JSON v3.5.0 targets low-latency JSON workloads in MT5 environments running LLM function-calling flows. Community JSON libraries showed two recurring constraints: frequent allocations and high serialisation latency, with recursion and temporary strings causing terminal stalls under load.

The library is rebuilt around a zero-allocation approach: parsing uses a tape backed by a contiguous long[] buffer, with direct serialisation into uchar[] to reduce intermediates. It adds hybrid numeric parsing using long accumulation and static Exp10 lookup tables, an iterative state machine to avoid recursion and stack overflow, plus SWAR scanning to process 8 bytes per cycle when skipping whitespace and long strings.

On x64 hardware with a 50,000-node payload, benchmark results report 137 ms parse vs 1540 ms, 264 ms serialisation vs 568 ms, and 401 ms ...

👉 Read | Docs | @mql5dev

#MQL5 #MT5 #AI

Part 25 extends an MQL5 3D binomial viewer into a multi-distribution plotting module. Supported models include Poisson, normal, Weibull, gamma, and additional discrete and continuous families, with a header icon used to cycle types at runtime.

The design relies on a distribution enum and a single dispatch function that routes sample generation, histogram building, density computation, and UI labeling. Adding a new model becomes an enum entry plus a loader function.

Implementation details include per-distribution input groups, dedicated loaders that standardize seeding, sampling, binning, scaling to theoretical peaks, and statistics. Continuous histograms have forced-range variants for heavy tails (notably Cauchy). Titles, axes, and parameter panels update based on the active type.

👉 Read | AppStore | @mql5dev

#MQL5 #MT5 #AlgoTrading

Automated S/R tools often treat zones as static objects, creating analysis gaps after breakouts. A common case is resistance breaking on a strong bullish close, followed by a retest where the same level behaves as support. This is role reversal and it is typically handled manually.

A dynamic implementation in MQL5 can detect HTF base-impulse zones, track their upper/lower bounds, and evaluate each close for a confirmed breakout plus an impulse-range threshold scaled by zone height. When conditions match, the rectangle is replaced with the opposite role, color updated, and expiry extended without duplicating historical zones.

The result is an indicator that maintains a live map of supply/demand, supports retest workflows, and reduces subjective reclassification after structural breaks.

👉 Read | Forum | @mql5dev

#MQL5 #MT5 #Indicator

Part 9 builds a footprint order-flow indicator in MQL5 to expose what candlesticks hide: traded volume at each price level inside every bar, split into bid (selling) and ask (buying), plus per-level delta.

The design quantizes tick prices into configurable level bins, accumulates volumes per bar, computes max delta/total/bid/ask for normalization, then renders color-scaled volume text on a CCanvas overlay. Two views are supported: bid-vs-ask and delta+total, with optional diagonal imbalance highlighting to spot stacked aggression.

Rendering stays aligned with MT5 chart zoom/scroll by converting bar indexes and prices to pixels, while trendline-based candles update in real time. Practical use: identify absorption, low-volume nodes, POC-style levels, and delta divergence for trade context and strategy logic.

👉 Read | VPS | @mql5dev

#MQL5 #MT5 #AlgoTrading

Battle Royale Optimizer (BRO) is a game-inspired metaheuristic that reframes population search as competition rather than cooperation. Each candidate solution “duels” its nearest neighbor; the better one resets damage, the worse one gains damage and shifts toward the current global best.

Exploration is enforced by elimination: once damage crosses a threshold, the solution is discarded and replaced with a new random candidate, preventing stagnation. Exploitation is strengthened by periodically shrinking the feasible bounds around the best solution using per-dimension standard deviation, with a delta schedule that controls how often contraction occurs.

The MQL5 implementation organizes this as a C_AO_BRO class with neighbor search, Euclidean distance, damage tracking, and boundary shrinking. Tests show strong performance on continuous landscapes (Hil...

👉 Read | Forum | @mql5dev

#MQL5 #MT5 #AlgoTrading

Part 8 folds the Optuna HPO system into the production training stack, resulting in one ModelDevelopmentPipeline that can run either sklearn search (Grid/RandomizedSearchCV) or Optuna via a single model_params flag.

The Optuna path jointly samples hyperparameters and sample-weight design (scheme, decay, linearity), adds early stopping with Hyperband pruning, and persists studies in SQLite for crash-safe resume. Primary vs secondary (meta-labeling) training is detected by the presence of side in events, which controls meta-feature generation and artifact naming.

Inference reliability is improved by storing the fitted column-dropping preprocessor (constant/duplicate removal) as step 0 of the final sklearn Pipeline, preventing silent feature misalignment.

Post-HPO ensembling supports sequential bootstrapped bagging to reduce correlated bags under overlapping tr...

👉 Read | AlgoBook | @mql5dev

#MQL5 #MT5 #Optuna

DADA targets time-series anomalies in markets without retraining per instrument by avoiding fixed autoencoder compression and learning robust structure via masked reconstruction.

Its core is an adaptive bottleneck: a router selects among multiple latent sizes so clean signals keep detail while noisy regimes are compressed harder. Two decoders split responsibilities: one reconstructs normal behavior; the other is trained adversarially on synthetic anomalies injected into data, improving separation and reducing false alerts. In production, only the normal decoder runs; large reconstruction error flags anomalies.

An MQL5 implementation focuses on an efficient adaptive module using a multi-window convolution layer, pushed into OpenCL kernels for forward pass, backprop, and Adam updates with shared weights across multivariate series.

👉 Read | CodeBase | @mql5dev

#MQL5 #MT5 #AI

A price action Expert Advisor built without lagging indicators, using raw OHLC across multiple timeframes to derive entries from market structure. Logic aligns lower-timeframe momentum with higher-timeframe highs/lows, requiring multi-timeframe confirmation before execution, such as H1 vs H4 structure checks combined with M5 breakout validation.

Core components include MTF structure alignment via previous highs/lows (H1, H4, D1), role-reversal detection for support/resistance flips, and breakout/continuation rules based on candle closes through key levels. Execution favors pending limit/stop orders placed at structural levels rather than market orders.

Designed for liquid, directional markets such as XAUUSD and BTCUSD, typically attached to H1 while sourcing M5, M30, H4, and D1 internally. Parameters are modular: structure scan depth, boolean MTF conditions...

👉 Read | Signals | @mql5dev

#MQL5 #MT5 #EA

NeuroPro Verbalisation Converter for MQL5 automates porting trained neural networks from the 1997 NeuroPro package into MetaTrader 4/5 code. It targets typical incompatibilities in verbalised output: missing type declarations, extra brackets, absent semicolons, nonstandard array indexes, and parsing issues where “--” is treated as a decrement operator. It also addresses ANSI CP1251 encoding, preventing Cyrillic identifiers from being corrupted.

Conversion is done by direct byte reading via FILE_BIN, avoiding clipboard-related distortions. Naming is preserved as provided in the source, including case. Substitutions are limited to required elements, such as mapping SigmoidX to SiX while keeping indices, and generating double declarations for intermediate neurons.

A bracket-balance pass trims redundant characters and normalizes line endings with semicolons. Inpu...

👉 Read | Docs | @mql5dev

#MQL5 #MT5 #script

ExMachina Smart Money Concepts v1.0 is a free Smart Money Concepts (ICT) indicator for MetaTrader 5, implemented in native MQL5 with no external dependencies and no iCustom usage. It targets multi-timeframe use across any symbol, providing structure tooling plus on-chart state reporting.

Structure detection covers internal (pivot length 5) and swing (pivot length 50) with BOS/CHoCH options, independent bull/bear filters, confluence filtering, and distinct line styles with midpoint labels. Order blocks are generated on structure breaks (internal and swing), track up to 100 zones, render live rectangles, and remove on mitigation by Close or High/Low. Fair Value Gaps use 3-candle rules with an auto body-percentage threshold and mitigation handling.

Additional modules include ATR-based EQH/EQL, strong/weak highs and lows with forward extension, premium/discou...

👉 Read | Signals | @mql5dev

#MQL5 #MT5 #Indicator

ExMachina Heikin Ashi Enhanced v3.0 is a free Heikin Ashi indicator for the MQL5 Code Base, focused on completeness and configuration. It supports two render modes: color-coded HA candles with correct High/Low, or a smoothed color line overlay.

Smoothing options include SMA, EMA, double-smoothed EMA (Vervoort-style DEMA), SMMA, and LWMA with adjustable periods. A pip-based step filter suppresses minor HA flips, and Wingdings arrows can be plotted on each color change with configurable codes and offsets. MTF mode maps higher-timeframe HA onto the current chart.

The indicator includes recursive HA OHLC, a consecutive bar counter, a strength metric (body-to-range) with percentage and bar display, HA delta, and a 9-row dashboard. Alerts are available via popup, sound, push, and email. Implementation notes include 14 buffers, modular calculation paths, 82...

👉 Read | Docs | @mql5dev

#MQL5 #MT5 #Indicator

A regime-aware, restartable grid-trading EA implements constrained grid cycles based on Bi-Directional Grid Constrained (BGC) stochastic process research (USQ, 2020–2022). Instead of running an open-ended grid, it operates in finite cycles with defined exits (profit target, drawdown limit, max age, weekend flat), then restarts to reset variance and exposure.

Three algorithms switch per cycle: BGT for ranging (both sides per level), TGT for trends (directional stops with basket exit), and MGT for post-trend mean reversion (counter-trend limits back to anchor). Regime selection combines an ATR-to-drift sigma/mu gate with cooldown and a frozen-baseline CUSUM structural break detector, plus deployment blocks near extremes.

Infrastructure includes ATR-dynamic spacing, equity-based dynamic lot sizing, tick-level drawdown kill, and extensive CSV diagnostics wit...

👉 Read | Calendar | @mql5dev

#MQL5 #MT5 #AlgoTrading

This article reframes grid EAs as an engineerable stochastic process, not a martingale. Loss grows quadratically with adverse grid levels (triangular numbers), buying time for detection and intervention, but indefinite operation still leads to almost sure ruin.

The core production insight is “finite, restartable cycles”: close baskets in profit to reset both variance (time) and exposure (grid depth), and enforce exits via an equity-based kill switch, a max cycle age, and weekend flattening. A loss-exposure formula using session high/low provides O(1) diagnostics, while live risk control uses real equity to handle re-arming, costs, and path effects.

The EA becomes a regime-adaptive cycle manager with three grid modes (range, trend, post-trend mean reversion), ATR-driven spacing, equity-based sizing, structural-break detection (CUSUM/volatility gatin...

👉 Read | Signals | @mql5dev

#MQL5 #MT5 #AlgoTrading

MT5 EAs lose all in-memory state on terminal close, chart refresh, or removal, making multi-terminal coordination hard. A practical fix is persisting a single parameter/state record in a shared binary file so restarts resume cleanly.

The approach packs EA settings into an MQL5 struct and writes/reads it as raw bytes using FileWriteStruct/FileReadStruct with FILE_BIN, stored under FILE_COMMON for cross-terminal access on the same PC.

Key constraint: only simple, fixed-layout structs (int/double/bool, optionally nested) are safe for binary persistence. Strings, dynamic arrays, pointers, and objects require extra handling because their memory layout isn’t stable.

Saving typically happens in OnDeinit(); loading on start restores the exact parameter set without text conversion, reducing code and type/order errors compared to CSV.

👉 Read | Docs | @mql5dev

#MQL5 #MT5 #EA

A Market Entropy Indicator formalizes market randomness using Shannon entropy on discretized price states (up/down/flat). A normalized rolling score classifies regimes: TREND (<0.35), TRANSITION, and CHAOTIC (>0.65), reducing reliance on price-only heuristics.

Implementation in MQL5 typically uses separate-window plots with fast/slow horizons (e.g., 20/100), momentum, divergence, and compression/decompression detection. Buffers and flags drive a color histogram plus markers for regime shifts and 25%+ “information shocks.”

Signal logic is rule-based: buys align with compression to decompression, fast entropy crossing above slow, and improving momentum outside chaotic zones. Sells reverse those conditions, with rising disorder and negative momentum. Visual outputs are spaced to prevent clustering.

👉 Read | CodeBase | @mql5dev

#MQL5 #MT5 #Indicator