Overview
ForXPair did not start with a website. It started with a question: which trading strategies hold up, and how do you run them reliably on many accounts at once? Answering it took two long-running programmes: quantitative strategy research and execution engine development. Each has run for years, and both continue today.
This update explains, section by section, how we evaluate strategies, how the engine was engineered, and what that means for the traders who will run ForXPair strategies on their own MT5 accounts.
1. Strategy research: more than 500 candidates
Over several years, our research team has built, backtested and analysed more than 500 trading strategies across major currency pairs and gold. They cover trend-following, mean-reversion, breakout, session-based, grid and hedging approaches.
- Hypothesis first. Every strategy starts from a clear, testable idea about market behaviour, never from curve-fitting.
- Multi-year backtesting. Each candidate is tested on years of historical data, across very different market regimes: trending markets, quiet ranges, news-driven volatility and crisis periods.
- Realistic trading costs. Spreads, commissions and slippage are included, because a strategy that only works without costs does not work.
2. The filtering pipeline
Most strategies do not survive. Our multi-stage filter is deliberately strict, so that only a small shortlist reaches live testing:
- Stage 1: Performance screen. Minimum profitability over the full test period, not just a few good months.
- Stage 2: Risk screen. Maximum drawdown, recovery time and the depth of losing streaks.
- Stage 3: Consistency screen. Results must be spread across many months and market conditions, not driven by a handful of trades.
- Stage 4: Robustness screen. Small changes to parameters must not break the strategy. Fragile, over-optimised systems are rejected.
- Stage 5: Live forward testing. Survivors run on real master accounts, with results tracked independently, before they are ever offered to customers.
3. The execution engine: two and a half years of R&D
A good strategy is only as good as its execution. Over two and a half years our engineers designed, built and refined the ForXPair execution engine: the component that takes each strategy's trades and runs them on every connected customer account.
- Low-latency core. A fast execution loop processes all connected accounts in parallel. In production it completes a pass in about one millisecond on average.
- Real-time streaming. Strategy positions arrive over a live streaming connection, with automatic failover to a high-frequency fallback if the stream is ever interrupted.
- Crash-safe recovery. Every tracked position is persisted. After a restart the engine restores its state and continues without opening duplicate trades.
- Broker-aware precision. Symbol mapping across brokers, cent and standard accounts, and broker credit or bonus balances are all handled correctly, so lot sizes and risk stay accurate on every account.
- Order management. Market, pending and grid orders, partial closes, and automatic retries when a broker requotes a price.
- Built-in risk tools. Trailing stops, scale-out targets with automatic break-even, profit targets and drawdown stops.
- Self-monitoring. A heartbeat every few seconds, a health report every 30 seconds, and clear status shown to the customer when an account cannot connect.
4. Fine-tuning in production
Much of the work over the last two and a half years has been fine-tuning under real load. Every change is documented: the core engine and platform now carry more than 2,800 documented refinements, each recorded with its date, the files it touched and the reason for it. A few examples:
- Concurrency tuning, so the engine stays responsive when many positions close at the same moment.
- Data-traffic optimisation, sending only changes instead of full snapshots. This cut database traffic sharply while keeping execution just as fast.
- Edge-case hardening, from broker-specific price precision to accounts carrying broker credit, each found, measured and fixed in a numbered engine release.
- Release discipline. Every engine build is versioned and documented, so we always know exactly which version runs in production.
6. The research continues: 200+ strategies in the pipeline
Strategy research at ForXPair is continuous. More than 200 new strategies are in our testing pipeline right now, and each one goes through the same five-stage process as the original 500+.
Candidates are tested deliberately across different market conditions, from low-volatility ranges to fast trends and high-impact news, before any of them is considered for the platform. Only strategies that hold up across conditions will be added.
7. Tools and applications in development
Alongside strategy research, our team is developing a suite of trading tools and applications with one goal: to make professional-grade trading accessible to every trader, from first-time users to full-time professionals.
Every tool is built on the same execution engine, so the speed, accuracy and safety described above carry through to each new product. The most ambitious of these is TraDeck, our dedicated trading console.
8. What this means for traders
- Strategies that have passed years of testing and a strict filtering process.
- An execution engine engineered and refined over two and a half years for speed, accuracy and resilience.
- A non-custodial model: your account stays with your broker, in your name, and you can revoke access at any time.
Important information
Backtested and simulated results have inherent limitations. They do not represent actual trading and are not a reliable indicator of future results. Live results can differ. Trading leveraged products carries a significant risk of loss.