MSRLab Strategy Engine

MSRLab Strategy Engine is a strategy-development and testing EA not one robot/EA. Build, test and combine your own MT5 strategies from a modular library of breakout, trend, mean-reversion, volatility, structural and price-action mechanisms.

The engine allows users to:

• Load individual strategy profiles through CSV. It is shipped with some examples.
• Test different entry modules, risk models, stops, targets and management rules.
• Enable or disable profiles.
• Combine multiple profiles into one collection.
• Run backtests and inspect results.
• Start with supplied examples, then create their own variations.

The customer workflow should be:

1. Buy or rent the EA through MQL5 Market.
2. Install the EX5 into the customer’s MT5 MQL5\Experts directory.
3. Attach it to a chart.
4. Start with no-order/Discovery mode.
5. Copy a supplied CSV profile into the documented MT5 files location.
6. Select the CSV/profile input in the EA. Example CSVs are available.
7. Run a Strategy Tester backtest.
8. Review trades, drawdown, costs and sample size.
9. Test on demo before enabling trading.
10. Only then consider live use.

What the engine contains

The engine currently accepts 23 entry models.

Breakout and structure

• DONCHIAN_BREAKOUT
• RANGE_BREAKOUT_BUILDER
• SESSION_RANGE_BREAKOUT
• VOLATILITY_EXPANSION_BREAKOUT
• NR7_REGIME_BREAKOUT
• SR_BREAKOUT
• SR_RETEST
• TURTLE_SOUP
• WEEKLY_DIRECTION_BREAKOUT

These cover:

• Donchian/channel breaks.
• Asian/session range breaks.
• Volatility contraction and expansion.
• Narrow-range/NR7 breakouts.
• Structural support/resistance pivot breaks.
• Retest and stop-limit concepts.
• Turtle-style breakout logic.
• Weekly directional breakouts.
• Stateful range-builder logic.

The range builder itself supports different breakout behaviours such as:

• STOP_TOUCH
• CLOSE_BEYOND
• RETEST
• FADE_EDGE
• FAKEOUT
• QUALITY_CLOSE
• FOLLOW
• STATEFUL_RETEST

Trend-following and continuation

• DONCHIAN_TREND
• DONCHIAN_HOLD_TREND
• MA_CROSS
• MA_PULLBACK
• MACD_REGIME_CONTINUATION
• DECYCLER_TREND_TRANSITION
• QSTICK_REGIME_CROSS
• RSI2_TREND_FADE
• WEEKLY_RECOVERY

These can express:

• Moving-average trend transitions.
• Pullbacks within a trend.
• MACD regime continuation.
• Qstick/candle-pressure continuation.
• Donchian trend following and hold logic.
• Recovery or directional weekly conditions.
• RSI2 trend fade behaviour.

Mean reversion and exhaustion

• BOLLINGER_RSI_FADE
• ZSCORE_FADE
• KALMAN_RANGE_REJECT
• UPPER_CLOSE_EXHAUSTION_FADE

These support ideas based on:

• Bollinger-band extension plus RSI re-entry.
• Z-score extremes.
• Kalman range rejection.
• Exhaustion after strong closes.

Price-action patterns

• ENGULFING
• PIN_BAR
• INSIDE_BAR_BREAK

These are candle-structure examples for:

• Engulfing reversals or continuation.
• Pin-bar rejection.
• Inside-bar contraction and breakout.

Entry timing and order types

The engine is not limited to “enter at market.”

It supports:

• Market entries.
• Stop entries.
• Stop-limit entries for the retest model.
• Entry expiry.
• Setup-gap requirements.
• Breakout confirmation.
• Retest offsets.
• Session and event-time restrictions.
• One-event-per-day controls for session/range models.
• Weekday and clock filters.

The engine deliberately uses completed-bar logic. The forming candle is not treated as a confirmed signal.

Stop-loss models

Available protection models include:

• ATR-based stops.
• Fixed-point stops.
• Previous-day/H1 ATR stops.
• Price-percentage stops.
• Session-range stops.
• Swing-based stops.
• ATR-plus-structure stops.
• Opposite-range stops.
• Range-width stops.
• Mid-range stops.

This allows the same entry idea to be tested with different definitions of invalidation.

Target and exit models

Targets include:

• Risk/reward multiple.
• ATR multiple.
• Fixed points.
• Price percentage.
• Range-based target.
• No fixed target.

Exit and management options include:

• Stop-loss and take-profit.
• Maximum holding bars.
• Exit-on-signal.
• Break-even.
• ATR trailing.
• Bar-based trailing.
• Chandelier-style trailing.
• Target trailing/step logic.
• Pending-order expiry.
• Automatic cancellation.
• Risk-breach cancellation.

Partial exits are not part of the current public candidate.

Risk controls

The risk layer supports:

• Risk-percentage sizing.
• ATR-budget sizing.
• Fixed-lot mode where permitted by the profile contract.
• Maximum lots.
• Maximum positions.
• Family risk caps.
• Portfolio/bundle risk caps.
• Tranche weights.
• Staged entries.
• Profit guards for staged tranches.
• Daily drawdown block.
• Peak drawdown block.
• Optional close-on-breach.
• Spread limits.
• Cooldown bars.
• Minimum and maximum range filters.
• Minimum breakout strength.
• Structural stop maximums.
• Symbol-specific minimum-volume fail-closed handling.

The engine calculates volume from stop distance and symbol contract information rather than assuming every broker uses the same lot geometry.

Portfolio and collection controls

A user can:

• Run one profile.
• Enable several profiles.
• Assign each profile a unique magic number.
• Combine profiles into one collection.
• Set profile-level risk.
• Set family-level risk.
• Set bundle-level risk.
• Assign staged tranche roles and weights.
• Disable individual profiles.
• Use one enabled profile per symbol for safer account-neutral operation.

The supplied Collection B is an example of how two individual profiles can be combined:

• Bollinger EURUSD H1.
• Bollinger AUDCAD H1.

CSV profile system

The profile format is deliberately extensive. The current collection schema has approximately 99 columns covering:

• Identity and labels.
• Symbol and timeframe.
• Entry model.
• Long/short permissions.
• Indicator periods.
• Regime filters.
• Stop and target models.
• Trailing settings.
• Breakout settings.
• Session/event times.
• Risk settings.
• Position limits.
• Portfolio/bundle metadata.
• Entry order type.
• Expiry and retest settings.
• Structural S/R parameters.
• Range-builder parameters.
• Kalman parameters.
• Weekday and clock filters.
• Magic number and comments.

This means a user is not buying a fixed robot with one hardcoded strategy. They are buying a configurable research engine.

Example profiles and backtests

These are historical examples, not promises of future performance.

BOLLINGER_RSI_FADE — EURUSD H1

• Broker: IC Trading
• Period: 2026-04-01 to 2026-08-31
• Model: 100% real ticks
• Net profit: +$300.11
• Profit factor: 1.31
• Trades: 72

BOLLINGER_RSI_FADE — AUDCAD H1

• Broker: IC Trading
• Same period and model
• Net profit: +$78.89
• Profit factor: 1.09
• Trades: 60

PIN_BAR — EURUSD H1

• Broker: IC Trading
• Same period and model
• Net profit: +$71.68
• Profit factor: 1.03
• Trades: 144

Collection B — EURUSD + AUDCAD

• Broker: IC Trading
• Same period and model
• Net profit: +$787.40
• Profit factor: 1.40
• Trades: 132
• Maximum equity drawdown: approximately 3.05%

These are supplied example portfolio to start with back testing and understanding the product, not a recommended live portfolio.