An indicator is available to chart the spread between two instruments as either a difference or a ratio.

Configuration includes Instrument1 and Instrument2, which must match the broker’s symbol strings exactly. Multiplier1 and Multiplier2 apply scaling factors to each instrument to normalize contract sizes or pricing units.

The ChartDifernce boolean selects the output mode. When enabled, the indicator plots the difference between the two instruments. When disabled, it plots the ratio.

The spread series is rendered in a separate subwindow. Once plotted, standard technical indicators can be applied to the spread line for additional analysis.

👉 Read | Forum | @mql5dev

#MQL4 #MT4 #Indicator

MetaTrader 5 indicators rely on event-driven flow: timeframe changes trigger OnDeinit, then OnInit after the chart rebuild. During this cycle, indicator memory is cleared, so runtime counters reset unless state is stored externally.

One approach is to handle OnDeinit(reason) and preserve the counter only when reason equals REASON_CHARTCHANGE; otherwise reset to zero. This keeps logic local, but file-based persistence can introduce race conditions when multiple charts or rapid period changes contend for the same state.

An alternative is terminal global variables. They persist across reinitializations while the terminal is running, store only double, and can be adapted via unions for packed data. GlobalVariableTemp enables session-scoped state without filesystem dependency, but shifts control to terminal-level storage semantics.

👉 Read | CodeBase | @mql5dev

#MQL5 #MT5 #AlgoTrading

Unidirectional ML trading is often a better fit for instruments with persistent trend, where buy/sell classes are not separable and bidirectional classifiers generate systematic errors.

A one-direction sampler replaces buy/sell markup with a single label: 1 means “open trade in selected direction”, 0 means “skip”. Trade duration is randomized between min/max bars.

The tester is updated for one-direction logic with Numba-accelerated processing, supporting exits by model signal, stop loss, or take profit, whichever occurs first.

Meta learners are used as preprocessing, not as final models. Cross-validated or bootstrapped “heads” identify consistently misclassified rows and set meta_labels and labels to 0 to filter low-quality entries. This simplifies learning under trend non-stationarity and reduces false entries for the chosen direction.

👉 Read | AlgoBook | @mql5dev

#MQL5 #MT5 #AlgoTrading

Classic Ilan popularized grid averaging with Martingale-like sizing: add positions against price to improve the average entry. It can look robust in range markets, but it predictably fails in sustained trends as exposure grows exponentially and margin collapses.

Ilan 2.0 added volatility-based grid spacing, multi-symbol correlation checks, and extra filters, yet remained a static rule engine with no way to learn when to stop averaging.

Ilan 3.0 AI reframes the EA as a reinforcement learning agent. Q-learning drives decisions via a dynamic Q-table over discretized market/position states, updated with the Bellman equation and an ε-greedy explore/exploit schedule. Rewards favor profitable closes and penalize averaging, with persistence by saving/loading the Q-table.

Tests still show drawdown risk typical of Martingale, motivating a next step: neural...

👉 Read | AppStore | @mql5dev

#MQL5 #MT5 #AlgoTrading

Price Action Day Trader EA executes on each new bar using three setups: pin-bar rejection with configurable wick-to-body thresholds, engulfing candles for reversal momentum, and inside-bar breakouts from a defined mother bar.

Entries are gated by a dual EMA trend filter (20/50 by default) and a support/resistance lookback that validates signals against recent structural levels.

Risk is position-sized by balance percentage (default 1.5%) with a max daily loss cap (3%). Trade management includes break-even automation and trailing stops, plus a session window that can block entries and force end-of-day flat.

Execution includes automatic detection of broker filling mode (FOK, IOC, Return). Key inputs cover RiskPercent, StopLossPips, TakeProfitRatio, PinBarRatio, SRLookback, EMA periods, and trading hours. Recommended for M15/H1/H4 on majors and XAUUSD, with SL...

👉 Read | AppStore | @mql5dev

#MQL5 #MT5 #EA

A computer-vision pipeline is built around MetaTrader 5 data to predict EURUSD direction without hand-crafted indicators. The system pulls ~2,000 H1 bars, converts rolling 48-candle OHLC windows into normalized multi-channel “images,” and labels each sample by the move 24 bars ahead.

A Keras Functional CNN (two Conv blocks with batch normalization, then dense layers) is trained with early stopping to reduce overfitting. Functional design also enables introspection.

Interpretability is treated as a core feature: intermediate feature maps are visualized alongside candlesticks, and an attention-style heatmap highlights the chart regions driving the decision, often aligning with reversals and key levels.

Outputs are presented as both direction and probability, supporting confidence-based position sizing and server-side execution via headless plotting.

👉 Read | Quotes | @mql5dev

#MQL5 #MT5 #AI

Market reversals are often driven by crowd positioning, not clean trend logic. Larry Williams’ smash day reversal targets breakouts that fail fast and trap late participants.

A smash day is a bar that closes beyond recent structure: below prior lows for buy reversals, above prior highs for sell reversals. Outside bars are excluded to reduce ambiguous range expansion.

The EA framework makes the pattern testable in MQL5. It supports buy-only, sell-only, or both, with a single active position to keep attribution clean.

Entries trigger on a break of the smash bar level, either immediately on price cross or after bar-close confirmation. Stops sit at the smash bar extreme; take profit uses a configurable risk-reward ratio. Position sizing can be fixed-lot or percent-risk.

👉 Read | NeuroBook | @mql5dev

#MQL5 #MT5 #EA

This article builds an on-chart position labeler for MetaTrader 5 to close the usability gap between the Trade tab and live price action. Each open position is tied to a chart “bubble” anchored at entry time/price, showing ticket and real-time P/L with a clickable close area.

The implementation uses MQL5 Standard Library components: CCanvas for ARGB overlay rendering, CPositionInfo for stateful access to live position fields, CTrade for safe programmatic closes, and CiMA (Trend) to generate simple test positions.

Key engineering points include custom hitboxes for mouse interaction, OnInit bootstrapping of indicator handles and chart event routing, and a throttled redraw loop (~30 FPS) to keep UI responsive without slowing trading logic.

👉 Read | AppStore | @mql5dev

#MQL5 #MT5 #EA

A new option has been added to select the timeframe used when drawing Bobokus Fibonacci levels. This allows Fibonacci calculations and placement to be based on a chosen chart period rather than the currently active view, improving consistency when switching timeframes.

Support is also included for custom Fibonacci levels. Any additional ratios can be defined and displayed alongside the default set, making it easier to align levels with a specific trading model or internal standards.

The update focuses on configurability and repeatable layout, reducing manual rework when analyzing multiple symbols or sessions.

👉 Read | NeuroBook | @mql5dev

#MQL5 #MT5 #Indicator

Better MQL5 code often comes down to habits, not years of experience.

Strong forum culture matters: critique code only after understanding intent, and provide explanations that teach. Dropping unclear snippets encourages copy-paste fixes, spreads bad patterns, and leads to wrong answers and unfair reviews.

Adopt a growth mindset. Complex EAs, advanced math, and adaptive models become approachable through steady practice and reading real source code.

Keep code lean and intentional. Remove unused blocks and variables to reduce bugs, speed reviews, and make debugging easier.

Treat mistakes as training data. Try multiple approaches before posting, document what was attempted, then understand the fix to prevent repeats. Progress in algo trading development is incremental, built through iteration, experiments, and helping others.

👉 Read | Docs | @mql5dev

#MQL5 #MT5 #Strategy

An equidistant channel tool is used to mark rectangular consolidation by placing two parallel boundaries around price. The upper line tracks repeated highs, the lower line tracks repeated lows, while the midline helps assess balance and potential mean reversion.

Key checks include multiple clean touches on both sides, stable channel width, and limited slope to confirm a range rather than a trend. When volatility contracts inside the channel, breakouts can be evaluated by close outside the boundary, follow-through, and a retest of the broken line as former resistance or support.

👉 Read | Quotes | @mql5dev

#MQL5 #MT5 #Indicator

GUI animation on CCanvas requires deterministic redraw. Overpainting frames accumulates pixels, so state must be restored before drawing the next frame.

A minimal sprite workflow is implemented by caching the background under a moving element, drawing the element, then restoring the cached pixels on the next update. This scales better than full form redraw for multiple independent elements.

GCnvElement is extended to store an initial pixel snapshot, add resource-to-array copying, and track last text coordinates plus anchor-based X/Y offsets for correct background capture.

ShadowObj is simplified by reusing the new base snapshot and Width/Height accessors. Form.mqh adds a pixel-copier class with ID-based instances, storing rect coordinates, effective clipped size, and save/restore methods for partial canvas regions.

👉 Read | Calendar | @mql5dev

#MQL5 #MT5 #algorithm

A new technical indicator marks values when they enter oversold and overbought zones, with an arrow signaling when the value returns from an extreme level.

Configuration options cover colors, line width, zone levels, and arrow display, allowing quick adjustment to different chart templates and instruments.

For more stable outcomes, the indicator is designed to be used alongside other signals and broader market analysis rather than as a standalone trigger.

Optional alerts are available to notify on key zone events and return signals, supporting monitoring without constant chart focus.

👉 Read | AlgoBook | @mql5dev

#MQL4 #MT4 #Indicator

Geometric Asymmetry (GA) Breakout frames breakouts as a structural condition, not a volatility event. Consolidations are qualified via fractal swing sequences, ATR-normalized range rules, and internal leg proportion checks to detect directional pressure before any boundary breach.

Decision flow is staged: fractal-based swing extraction, consolidation validation, geometric voting on distance/slope/time asymmetry, then pattern locking to prevent repainting. Signals trigger only on candle close beyond range high/low with a configurable buffer and cooldown.

MQL5 implementation follows a modular layout: fractal detection, range boundary calculation, qualification filters, asymmetry validation, breakout scanning, and pattern lifecycle management. OnTick runs a bar-close pipeline and emits chart objects plus multi-channel alerts without repeated triggers on the sa...

👉 Read | NeuroBook | @mql5dev

#MQL5 #MT5 #EA

Markets often look random, but chaos theory frames them as systems with hidden structure: mean reversion, sensitivity to initial conditions, and fractal self-similarity across timeframes.

The article builds a MetaTrader-style Neural Attractor Oscillator: a small feedforward network trained on price history, using tanh for bipolar dynamics and normalization for stable learning. It predicts a moving “attractor” level, then outputs an oscillator as percent deviation from that attractor to flag overextension and potential snap-back.

Signal quality is gated by a Lyapunov exponent to quantify how chaotic the market is, with optional fractal noise (Midpoint Displacement) and momentum optimization to improve training. Practical use includes adaptive overbought/oversold bands, divergence detection, and smoothing choices (EMA/SMA) based on regime.

👉 Read | AppStore | @mql5dev

#MQL5 #MT5 #AITrading

MetaTrader 5 indicators and EAs ship with a standard settings dialog built from default tabs. Most UI additions seen in custom tools duplicate what the platform already provides, and increase maintenance without adding capability.

The first tab can be adjusted via #property directives, including icon, purpose text, and a URL. These values are compiled into the binary and rendered by the terminal, with link details available via hover and click.

Indicator execution hinges on OnInit and OnCalculate. MQL5 provides overloaded OnCalculate signatures; the chosen signature affects whether specific configuration tabs are shown. A minimal indicator can be defined statically with #property, one plot buffer, and SetIndexBuffer, relying on platform-managed styling options such as color, line style, thickness, and calculation mode.

👉 Read | NeuroBook | @mql5dev

#MQL5 #MT5 #Indicator

Moving Average Cloud renders four moving averages on the same chart using a histogram-style band. It combines slow SMA(200) and EMA(200) with fast SMA(50) and EMA(50).

The space between the paired slow lines and the paired fast lines is plotted as boundary histograms, forming a cloud that highlights MA separation and compression. This makes it easier to track trend structure and potential support/resistance zones derived from MA alignment.

Histogram color changes are used to flag momentum shifts, helping detect transitions where fast/slow relationships start to rotate or converge.

👉 Read | Quotes | @mql5dev

#MQL4 #MT4 #Indicator

A lightweight EMA in MQL5 can be implemented with a single change in the calculation line, relying on MetaTrader 5 event handling and buffer management. The EMA formula only needs the current price, prior EMA value, and period, with an initialization workaround for the first value.

A key follow-up is parameterization. Using an input constant allows period changes without recompilation, and a string comment improves the settings UI by providing a clear label instead of the raw identifier.

The article also contrasts the two OnCalculate signatures. The extended version exposes more series data but typically fixes the applied price (for example, Close), which removes UI options and shifts responsibility to the implementation.

👉 Read | Signals | @mql5dev

#MQL5 #MT5 #Indicator

Manual Supertrend trading is deterministic on paper, but live execution often adds latency and decision noise. An MQL5 Expert Advisor can keep analysis separate from execution by acting only on confirmed trend flips.

Supertrend stays in two states. A buy triggers when a completed candle closes above the upper band and the state turns bullish. A sell triggers when a completed candle closes below the lower band and the state turns bearish. Signals are evaluated only on closed bars to avoid unstable reads.

EA rules: close any opposite position first, then open in the new direction if enabled. Direction can be long-only, short-only, or both. Stop loss is optional, using last candle high/low or the Supertrend band. Lot sizing supports fixed volume or risk-percent auto sizing.

👉 Read | VPS | @mql5dev

#MQL5 #MT5 #Indicator

StrategyTester5 adds multi-symbol, multi-timeframe history handling to a MetaTrader 5–Python simulator via a HistoryManager that fetches bars, ticks, and synthetic ticks, then syncs timeframe data for later copy_rates/copy_ticks access.

A naive multi-timeframe loop is a major bottleneck. Real-ticks mode can drop to ~13 ticks/s when repeatedly calling copy_rates_from_pos across symbols and timeframes. Switching to direct MT5 terminal reads can lift throughput to ~600 ticks/s, but long ranges still remain costly.

Parallel per-symbol OnTick execution improves total wall time by isolating trade operations per instrument while sharing account state. Additional gains come from new-bar event gating to avoid repeated history reads, and choosing faster modelling (M1 OHLC/new-bar) when tick-level accuracy is not required.

👉 Read | VPS | @mql5dev

#MQL5 #MT5 #AlgoTrading

MetaTrader 5 already supports alerts, push, email, logs, and simple sounds, but typical notifications lack context. A tone does not identify BUY vs SELL, symbol, timeframe, or signal conditions.

Semi-automated trading tools amplify this gap because they depend on human confirmation while communicating mainly through chart visuals. Accessibility limits come from UI assumptions, not from MQL5 capabilities.

A practical approach is an MQL5 MA crossover indicator that triggers structured text plus embedded WAV resources. On each confirmed crossover it reports symbol, timeframe, MA values, and price, plays buy/sell audio, logs events for later review, and avoids duplicate alerts via new-bar gating.

The design can be extended with richer voice feedback, voice commands, and future AI voice synthesis via external APIs, without changing the core signal logic.

👉 Read | Quotes | @mql5dev

#MQL5 #MT5 #Indicator