Neural Grid FX Portfolio 2026

=== Latest Report: Q1 Review, June - August 2026 (Weeks 1-13) ===

The first quarter of this signal is complete. Numbers first, then the one event that matters.

Net growth: +13.3 % (avg. ~4.1 % per month)

Closed trades: 147 (76.9 % win rate)

Profit factor: 2.28

Recovery factor: 4.53

Max equity DD: 23.8 % (late August)

Realized balance DD: 2.8 %

Max deposit load: 8.3 %

The August drawdown - what happened and what changed

In late August, a GBPNZD grid escalated against a sustained one-directional move. Peak equity drawdown reached 23.8 %. The risk framework worked as designed: the maximum grid depth cap contained the exposure and no forced stop was triggered. As an additional risk-control measure, the opening of new positions in the other pairs was manually suspended until the basket resolved, preserving margin headroom. No positions were manually opened or closed. The basket resolved, and the account recovered fully to a new equity high within days. No deposits were added at any point.

Consequence: GBPNZD has been removed from the portfolio. It produced the deepest grid escalation and the worst pip balance of all pairs (-1.5K pips gross); its volatility profile fits this grid configuration worse than any other pair. The portfolio now runs on nine pairs. Expect somewhat lower trade frequency going forward; the risk parameters (35 % hard equity stop, capped grid depth) are unchanged.

This is a grid-based system. Drawdowns of this kind are part of its profile and will occur again. What this quarter demonstrated is how the system behaves when they do.

This report is updated at the start of each month.

------------------------------------------------------------

Fully automated multi-currency grid strategy on 9 diversified forex pairs: AUDCAD, AUDNZD, NZDCAD, NZDUSD, EURCHF, EURGBP, USDCAD, EURJPY, USDJPY. Entries are filtered by a trained neural network that evaluates M5 market conditions and skips low-probability setups.

Core logic:

- Grid-based mean reversion with predefined spacing

- Progressive volume scaling across grid levels

- Neural network filter trained on historical tick data to reject weak entry signals

- Diversified across 9 low-correlation currency pairs to reduce single-pair risk

- Position sizing scaled to account equity

Execution:

- Fully algorithmic, runs 24/5 on dedicated VPS

- No discretionary trading. Execution is fully automated; as a risk-control measure, the opening of new positions may be manually suspended during extreme drawdown events.

- No new trades are opened on Fridays to avoid weekend gap risk

- Designed for ECN / Raw Spread accounts

What this strategy does not do:

- No news trading

- No hedging of opposing positions

- No high-frequency scalping

This is a patient mean-reversion system that profits from range-bound market behavior across multiple pairs.

Risk profile:

This is a grid strategy with progressive lot sizing. Floating drawdowns are expected and are part of normal system behavior. Positions are managed algorithmically until the grid cycle closes in profit or the equity stop is triggered. A hard equity stop is set at 35% account drawdown to protect capital.

Important: Grid systems can experience prolonged drawdown phases during strong directional trends. This strategy is not suitable for low-risk or capital-preservation accounts.

Recommended subscriber setup:

- Leverage: 1:200 minimum, 1:500 optimal

- Balance: minimum 1000 USD to replicate lot proportions accurately

- Account type: Raw Spread or ECN

Updates and communication:

A monthly performance report is maintained at the top of this description. Questions about strategy behavior, risk settings, or subscriber setup are welcome anytime.

All performance data on this page is tracked and verified by MQL5 in real time.