GBP Machine Pro MT5 Review: 70 Features, Regimes & Test Limits
GBP Machine Pro EA MT5 trades GBPUSD on M5 using 70 market features and four range/trend and volatility regimes. Separate buy and sell models select entries and regime-specific stop-loss, target and maximum holding-time plans. Its appeal is a selective, observable decision process with one position and explicit exits. The main unresolved question is forward robustness: the supplied long-range simulation is not a dedicated live account.
View GBP Machine Pro features and pricing
- Developer source link: GBP Machine Pro EA MT5 on MQL5
- Live signal link: No dedicated public account verified for this product.
- Public manual: GBP Machine Pro PDF in ZIP archive
Source-based review checked 16 September 2026. We read the current listing, the 31-page developer manual and original gallery, and checked customer feedback. We have not independently traded the EA or reproduced its tests.
How the 70-feature model makes a decision
A completed M5 candle becomes a feature vector describing market conditions. The system divides those conditions into range or trend, each at low or high volatility. Four regimes and separate directional forests produce eight buy/sell model paths. The active path has its own acceptance threshold and trade-management profile. A high scanner score alone does not order a trade; the quality and execution gates can still reject it.
What distinguishes the GBP model
The GBP feature vector covers returns, candle anatomy, rolling ranges, mean-distance measures, volatility, volume, directional balance, RSI/slopes, time and spread. Those inputs are combined inside forest models; an RSI value is not a standalone trigger. The modern GBP branch adds a normalized confidence-edge check.
Training is performed offline and the model is embedded in the EA. During trading it classifies conditions using that stored logic; it does not learn a new strategy on the customer’s account. No cloud-model API is required for the trading decision.
Stop, target and maximum holding time
The entry regime is retained with the position so a later change in market regime does not silently replace its original plan. If neither stop nor target is reached, the maximum-hold rule exits after the configured number of M5 bars. Automatic sizing uses the actual stop distance; fixed lots retain the same entry and exit logic but change how risk scales with the account.
One simultaneous position prevents a basket of accumulated open trades. It does not by itself establish every detail of post-loss sizing. The manuals show a Recovery status without fully documenting its behaviour; this review therefore does not turn the one-position statement into an unsupported blanket no-recovery claim. Verify the installed inputs and intended risk before trading live.
The historical test has a separate early-data branch
The GBP manual explicitly describes one historical forest set and execution calibration for 2000–2007, and a modern branch from 2008 onward. The earlier branch does not operate in current live trading. A continuous curve labelled 2000–2026 therefore must not be presented as twenty-six years of one unchanged current model, or as an untouched forward test.
At 0.5% Auto Risk the developer’s control run reports 2,682 trades, profit factor 13.22, 94.74% winners and 1.62% relative equity drawdown. Relative balance drawdown is 0.75%, with 98% reported history quality. These unusually strong figures warrant reproduction on independent data and realistic costs. The public material does not establish a clean held-out validation period for the current trained model.
Reading the interface without confusing it with evidence
The main panel shows model paths, entry gates and the active trade plan. The analytics panel summarises closed results for the EA’s Magic Number. The bottom scanner helps check server time, sessions and the next M5 bar. A displayed confidence percentage or winning trade card is interface information, not a probability guarantee or an independently verified account.
Setup checks that matter
Attach to GBPUSD M5, enable algorithmic trading and start with the documented strategy defaults. Choose Auto Risk or fixed lots deliberately, then confirm the winter broker-server GMT offset and DST. With a UTC+2 winter/UTC+3 summer broker, the base offset stays +2 and seasonal adjustment is handled separately. The developer recommends a VPS; uninterrupted operation matters for time exits and trade management.
The listed $100 minimum is an operating starting point, not a capital-safety threshold. Minimum lot size, stop distance, leverage and account currency can constrain the percentage risk a small account can actually express. Test the intended broker and deposit rather than assuming every nominal percentage is feasible.
Live signals, customer reviews and purchase scope
No dedicated public GBP Machine Pro signal and no customer reviews were available at this check. The seller’s Aurum and ORIX accounts belong to different systems, so their returns are not borrowed as proof here. There is no customer-review screenshot to invent.
The MQL5 listing advertises the other Machine Pro as a developer-direct bonus obtained by contacting the author. These CheaperForex listings are separate offers; that promotion is not represented as an included entitlement.
Assessment
The useful distinction is the complete chain from market classification to separate directional models, execution filters and a retained exit plan. The detailed interface also makes waiting states easier to understand. The evidence gap remains forward verification under realistic broker conditions. Treat the historic curve as a test to scrutinise, not a forecast, and compare the companion system by its model design rather than by the shared branding.
More EAs from Leonid Arkhipov
Compare the developer’s gold, GBPUSD and EURUSD systems. Their markets, models and evidence differ; a result from one is not proof of another.