When you evaluate Expert Advisors for MetaTrader 5, two patterns dominate the marketplace: broad “all-pairs” robots with generic rules, and narrowly focused systems that specialise in a single market micro-structure. Prodigy FX EA V9.105 (MT5) belongs to the second camp. It’s built specifically for CAD/CHF, applies Bollinger Bands as its primary signal context, and layers exposure with a controlled averaging methodology supported by adaptive take-profit and stop-loss per position. The vendor positions the current build as version 9.105 with a list price of 299 USD, and notes that tuning was carried out over six recent years of history (2020–2025) in a Standard, no-commission environment—details that help you replicate conditions before committing capital.


Why specialise in CAD/CHF?
CAD/CHF tends to oscillate for long periods, then break into directional moves when macro drivers (oil sensitivity for CAD, risk sentiment and SNB tone for CHF) take the wheel. That profile is a natural match for mean-reversion tactics during range phases, provided that risk is tightly capped during trends. By anchoring entries to Bollinger Bands—a moving-average centerline with standard-deviation envelopes—the EA reacts to volatility in real time: bands expand when markets are active and contract during compression. This volatility-aware geometry is exactly why Bollinger-based systems often serve as a backbone for controlled fading and range re-engagement. 


Core design in plain English
The EA watches CAD/CHF relative to its Bollinger structure and initiates positions when the price becomes statistically “stretched.” If the move continues, it adds layers selectively (averaging) to improve the blended entry—never a free pass to unlimited exposure, but a structured way to normalise price if the market drifts before snapping back. Crucially, each position carries its own TP/SL rather than a single global exit, so risk and targets flex with context. In practice, adaptive exits can reduce the “all-or-nothing” outcomes you see with fixed pip targets in choppy sessions. These ingredients—pair specificity, Bollinger context, per-position TP/SL—are stated explicitly in the product listing and new-experts catalogue entry. 


Vendor-stated facts you should mirror
From the launch page and catalog: Pair: CADCHF; Version: 9.105; Price: 299 USD; Strategy: averaging guided by Bollinger Bands with adaptive TP/SL per position; Optimization window: 2020–2025 on a Standard account; Timeframe: “Any” (handled internally); and a publication timestamp of 7 October 2025 visible on the product pages. If you want results that rhyme with the presentation, reproduce these conditions in your testing stack. 



Setup a blueprint for MT5



  1. Install and attach: Place the EA in the Experts folder, restart MT5, enable algo trading, and attach it to a CAD/CHF chart. Because timing is handled internally, the listing indicates no strict timeframe requirement.

  2. Match the cost model: The vendor emphasises a Standard (spread-only) environment as the preferred profile. Averaging increases trade count; commission on a Raw/ECN feed can subtly shift breakeven math for tighter exits, so confirm expectancy if you insist on a different account type. 

  3. Start conservative: Run with small lots, limit the maximum number of layers, and phase in size only after you validate behaviour across sessions and news cycles.

  4. Run 24/5 on stable hosting: If the EA is part of your core stack, keep it on reliable hosting to avoid missed updates or disconnects.


Strengths that traders can harness
• Focused micro-structure: Specialising in CAD/CHF reduces parameter sprawl and makes monitoring easier than multi-symbol robots. 
• Volatility-aware triggers: Bollinger Bands scale with standard deviation, so signal thresholds evolve with market state rather than staying static. 
• Per-position adaptivity: Independent TP/SL assignments reduce herd behaviour across layers and can smooth the equity line during chop. 
• Recent-era tuning: Running optimisation over 2020–2025 captures pandemic-era turmoil, inflation cycles, and policy pivots—a realistic stress palette for a modern EA. Always verify locally. 


Limitations you must respect
No averaging system is immune to path dependency. When CAD/CHF trends hard—on a surprise SNB remark, a persistent risk-off, or an oil-driven CAD swing—layers can stack in an adverse direction. Adaptive stops help, but they don’t erase tail risk. Your capital policy—not clever entry logic—decides survivability. That means hard caps on layers, session filters, and explicit drawdown brakes.


A reproducible testing ladder (use this order)



  1. Replication backtest: CAD/CHF only, Standard-like spread model, historical span 2020–2025. You’re checking the shape of outcomes under vendor-stated assumptions. 

  2. Walk-forward: Segment data into rolling windows, avoid constant re-fitting, and observe stability through regime shifts.

  3. Monte Carlo: Shuffle trade order and volatility to gauge sensitivity. Averaging thrives in chop; these tests reveal how quickly risk compounds when the path deviates.

  4. Forward test: Demo or micro-live for several weeks. Track realised spread, slippage, frequency, and average exposure time per layer before scaling.


A risk-first operating playbook
• Max layers: Pre-declare a ceiling. No exceptions under emotion or FOMO.
• Daily loss & equity guard: Pause opens at a daily loss limit or at a portfolio drawdown threshold.
• Event calendar: CAD/CHF reacts to BoC, SNB, CPI and risk shifts. Reduce size or pause 30–60 minutes around tier-1 releases.
• Session hygiene: Spreads often widen at rollover or during illiquid hours; avoid fresh layers when costs spike.
• Weekly debrief: Review profit factor, MAE/MFE per layer, median hold time, and variance in realised vs. quoted spread.


FAQ—quick answers for common decision points
Does it need a specific timeframe?
The listing indicates “timeframe: any” because the EA orchestrates timing internally. Attach to CAD/CHF and let it run. 


Why prefer a Standard account?
With averaging, more trades = more costs. Spread-only pricing avoids per-turn commission drag that can distort tight targets. If you operate on a Raw account, recompute expectancy with commission included before scaling.


Is this a martingale?
It’s averaging with adaptive exits, not an infinite doubling scheme. Still, exposure grows with layers. The guardrails—max layers and daily loss—are non-negotiable.


Can Bollinger Bands handle trends?
Bands are volatility-aware, not trend-immune. They can frame trend-pullbacks and breakouts, but using them for mean-reversion demands strict stop logic when price walks the band. 



Editorial verdict
Prodigy FX EA V9.105 is a coherent, single-pair play for traders who prefer structured mean-reversion on CAD/CHF with Bollinger-adaptive entries and per-position TP/SL. It’s not a “set-and-forget” martingale—nor should you treat it like one. Approach it as a specialist component inside a diversified portfolio: mirror the vendor’s Standard cost assumptions, verify over the 2020–2025 window, enforce hard exposure caps, and keep an event-aware schedule. Do that, and you’ll have a repeatable, defensible process for deciding whether Prodigy FX earns a permanent slot in your MT5 lineup.