A Multi-Account Crypto CFD System with Fixed Risk and Regime Filters
Summary
The article describes a proposed automated system for cryptocurrency CFDs across separate MetaTrader 5 accounts. Its architecture assigns each account an independent Python process and terminal, with shared configuration values and trade execution handled by an MQL5 Expert Advisor. The strategy combines an LSTM direction signal, technical features, and a compression-based market regime filter that permits trading only when price data is classified as structured. It also outlines partial profit taking, a breakeven stop adjustment, and account drawdown limits.
Risk sizing is framed around a fixed dollar loss, while take profit is set as a multiple of stop distance. The author reports that the compression filter improved backtested win rate by about 12 percentage points and specifies a multi-objective walk-forward training approach. These figures and design claims are presented without enough test details in the supplied text to assess robustness, costs, or live behavior. Fixed dollar sizing also cannot guarantee a maximum realized loss through gaps or execution slippage, and the proposed model and thresholds require independent validation.
Key ideas
- Run a separate Python process and MetaTrader terminal for each account to isolate account connections.
- The proposed system combines LSTM predictions, engineered indicators, and a compression-based regime classification.
- Trade sizing targets fixed dollar risk and uses partial profit taking with a move to breakeven.
- The article reports a backtest win-rate improvement from the regime filter but gives limited validation detail.
- Execution gaps and slippage can make realized losses exceed the intended fixed-dollar amount.
Tags
This summary was written by Stratmill's research agent from the original; it is not a copy of the source.