Forex Next Candle Forecast 3D MT5

A forecast at the start of a candle, a result check after it closes. Forex Next Candle Forecast 3D for MetaTrader 5 estimates the direction and range of a new candle on its first tick. The forecast is displayed on a three-dimensional panel and recorded for comparison with actual price movement.

The indicator uses the sequence of quotes, completed market episodes, and current context. Once recorded, a forecast is not changed using subsequent data. Adaptive memory updates after the candle closes and applies to future forecasts.

Main features

• Tick analysis. Deep Flow assesses the direction and sequence of price changes, momentum persistence, bursts of activity, pauses, reversals, and spread behavior.
• Scenario comparison. Scenario Engine and Context Layer compare current conditions with completed episodes, accounting for trend, volatility, and trading time.
• Anomaly checks. Shock Engine compares price jumps, spread, and tick activity with historical baselines. Anomalies detected when calculating a new forecast may widen its range.
• Logging and evaluation. The forecast is written to a log and a CSV file before the outcome is known. After the candle closes, Scoring Engine compares it with the actual outcome.
• Three-dimensional panel. Displays the forecast, direction, time until the candle closes, outcome, and cumulative statistics.

Lumi is the character in the visual scene. The setting changes with the time of day and season, and her comments reflect the forecast state.

Usage and limitations

For testing, use the visual Strategy Tester with the "Every tick based on real ticks" modeling mode. Observe consecutive forecasts from startup: calculation at the beginning of a new candle, followed by evaluation after it closes. Results depend on the availability and quality of historical ticks.

Forex Next Candle Forecast 3D is an indicator, not an Expert Advisor. It does not open, modify, or close trades. The statistics describe forecasts, not executed trades. The internal probability estimate should not be equated with the probability of a profitable trade. The forecast range and past results do not guarantee future price movement or profits. Quotes, spread, data quality, and broker conditions affect the calculation; the user determines execution, position size, and risk.

Settings

Eight parameters are available. They do not disable data checks, Shock Engine, forecast logging, or outcome evaluation.

Memory and built-in model

• Analog memory. Default: 96 bars, range 24 to 256. Sets the depth of market-analog memory: lower values use a shorter history, while higher values use a longer one.
• Built-in model weight. Default: 1.00, range 0 to 2. Adjusts the contribution of the built-in model. A value of 0 removes its contribution, 1 preserves the original weight, and 2 doubles it.

Context adjustments

• Short impulse weight. Default: 0, range -3 to 3. Positive values favor continuation of short-term momentum; negative values favor a pullback.
• Six-bar trend weight. Default: 0, range -3 to 3. Accounts for the direction of the last six closed bars. A positive weight favors trend continuation; a negative weight favors movement against it.
• Range-location weight. Default: 0, range -3 to 3. Accounts for the position of price within the observed range. Positive values favor a breakout; negative values favor a return into the range.
• Opening-gap weight. Default: 0, range -3 to 3. Accounts for the gap between the previous close and the new open. Positive values favor continuation in the gap direction; negative values favor a gap fill.

Direction selection

• UP probability threshold. Default: 0.50, range 0.30 to 0.70. Sets the threshold for the internal estimate used to select an upward forecast. Lowering the threshold produces more upward forecasts; raising it makes this condition stricter.
• Signal polarity. Default: Normal, which preserves the model's direction. Inverted reverses the direction of every forecast.

Start with the default settings. Change one parameter at a time and check the result on a later period of historical data that was not used for tuning. Assess consecutive forecasts, not just individual matches.