Grid trading is one of the most misunderstood strategies in retail forex. It looks simple and its backtest results are often impressive. But the risk structure is fundamentally different from what the equity curve suggests — grid EAs have a specific failure mode that can eliminate an account with no warning. This guide explains how grid trading actually works, where it genuinely excels, and what you must understand before running a grid EA.

Risk warning: Grid trading carries the risk of large drawdowns during trending markets. Accounts can sustain losses significantly larger than initial deposits with some configurations. See our risk disclosure.

How Grid Trading Works

A grid EA places orders at regular intervals above and below the current market price — creating a "grid" of orders that covers a price range.

Basic structure:

• Current price: 1.0850 EUR/USD
• Grid step: 20 pips
• Number of levels: 5 above, 5 below
• Buy orders at: 1.0850, 1.0830, 1.0810, 1.0790, 1.0770
• Sell orders at: 1.0850, 1.0870, 1.0890, 1.0910, 1.0930

When price oscillates: As price moves up and down within the grid, orders fill and close at profit. Each round trip earns the grid step (20 pips in this example) minus spread and commission. In a ranging market, the EA collects small profits continuously.

The appeal: In sideways, oscillating markets — which constitute a significant portion of forex market behavior — grid EAs can produce very consistent, smooth equity curves.

Grid Trading Variants

Standard Grid (Fixed Grid)

Orders are placed symmetrically around the current price at fixed intervals. All orders are the same lot size. The grid is reset when all orders complete or when price exits the grid range.

Best for: Highly ranging, low-directional instruments in quiet market conditions.

Asymmetric Grid

More buy levels than sell levels (or vice versa), positioned to reflect a directional bias. If you expect EUR/USD to trend slightly upward, you'd place more buy levels below and fewer sell levels above.

Risk: If the directional bias is wrong, the asymmetric grid accumulates larger losses in the unexpected direction.

Martingale Grid

Lot size increases with each grid level. Level 1 might trade 0.01 lots, Level 2 at 0.02, Level 3 at 0.04, etc. As price moves against the initial direction, each level's increasing lot size means that when price reverses, the total recovery is faster.

This is the most dangerous variant. The geometric increase in lot sizes means a sustained adverse move (before reversal) can accumulate enormous unrealized losses. The account blows up when margin is insufficient to hold all open positions through the trend.

Hedged Grid

Both buy and sell orders are placed simultaneously at each level. In theory, the net directional exposure is zero. In practice, execution timing differences and spread costs mean the grid has small net exposure.

Advantage: Hedged grids have lower directional risk. Disadvantage: More complex to manage, and hedging in the same account doesn't reduce margin requirement at many brokers.

The Grid's Fatal Flaw: Trending Markets

The grid strategy works on the assumption that price oscillates within a range. It fails when price trends strongly in one direction.

What happens during a trend:

1. EUR/USD starts at 1.0850
2. Grid places sell orders at 1.0870, 1.0890, 1.0910, 1.0930, 1.0950
3. EUR/USD trends upward to 1.1050 (+200 pips)
4. All sell orders have filled — the grid has 5 open losing sell positions
5. Total unrealized loss: 5 positions × average ~100-pip loss × lot size
6. If lot size was 0.10 lots per level, total loss = 5 × 100 × $10/pip × 0.1 = $500 unrealized
7. On a $1,000 account: -50% drawdown while the positions are open

If EUR/USD reverses to 1.0850: all positions close at net profit (each closed at a 20-pip step above its entry). The drawdown was temporary — the grid recovered.

If EUR/USD continues to 1.1200 without reversing: the unrealized loss grows, margin is eventually insufficient, and the broker force-closes all positions. The account is blown.

The critical asymmetry: Grid EAs show good performance during range-bound periods (most of the time) but catastrophic losses during sustained trends (less frequent but inevitable). Backtests over periods that happened to be ranging look excellent. The same EA run over 2022 (strong USD trend) would have blown many accounts.

Why Grid EA Backtests Are Misleading

Most backtest periods are favorable: Forex markets spend roughly 60–70% of time ranging. A 2–3 year backtest in a predominately ranging period shows the grid at its best, not its worst.

Drawdown in backtests is "paper" drawdown: The grid accumulates floating (unrealized) losses during trending periods. In backtests, these appear as drawdown that eventually recovers. In live trading, a sustained trend without sufficient margin causes actual account closure — not recovery.

Backtests hide the catastrophic path: The equity curve shows the final result. It doesn't show the moments during backtesting where the strategy was underwater by 60% — the position it was in before the market reversed. Seeing "20% max drawdown" in the report doesn't reveal the path: 60% down → reversal → ended up at 20%.

Defensive Configuration for Grid EAs

If you choose to run a grid EA, these settings reduce (but don't eliminate) catastrophic risk:

1. Maximum open trades limit

Set a hard limit on the number of simultaneous open positions in the grid. Example: maximum 5 positions. When the grid has 5 open positions in one direction, no new levels are added — effectively capping the accumulating loss.

2. Basket stop loss

A total stop loss for the grid basket. If total unrealized loss exceeds X% of account, close all positions regardless of individual position status. This converts the "unlimited loss potential" of an uncapped grid into a defined maximum loss.

// Simple basket stop loss implementation
double totalFloatingPnL = 0;
for(int i = 0; i < PositionsTotal(); i++)
{
   if(PositionGetInteger(POSITION_MAGIC) == GRID_MAGIC)
      totalFloatingPnL += PositionGetDouble(POSITION_PROFIT);
}

double accountEquity = AccountInfoDouble(ACCOUNT_EQUITY);
double initialBalance = AccountInfoDouble(ACCOUNT_BALANCE) - totalFloatingPnL;

if(totalFloatingPnL < -(initialBalance * BasketStopLossPct / 100))
{
   CloseAllGridPositions();
   Print("Basket stop loss triggered at ", totalFloatingPnL);
}

3. Trend filter

Stop adding new grid levels when price is strongly trending (e.g., ADX above 30). This prevents the grid from accumulating positions during its worst-case condition.

4. Adequate capitalization

Grid EAs require significant account buffers. As a rule: your account should be able to sustain a basket stop loss at 20–30% drawdown multiple times before the grid's expected annual return justifies the risk. For most small retail accounts ($500–$2,000), grid strategies are poorly suited to the account size.

Grid Trading vs. Standard Strategies: The Honest Comparison

Frequently Asked Questions

Are grid EAs legal?

Yes. There are no regulatory restrictions on strategy type for personal trading (some brokers prohibit specific strategies in their terms, but grid trading is not broadly restricted).

What's the minimum capital for running a grid EA safely?

Varies by EA configuration, but $5,000–$10,000 is a commonly cited minimum for standard grid configurations. This allows enough buffer to survive a 500-pip trend before the basket stop triggers. Running a grid on $500 is very high risk.

Can I identify if a vendor's EA is running a hidden grid or martingale?

Yes. In any Myfxbook-verified account, open the trade history and look for: lot sizes that increase after consecutive losses, many simultaneously open trades in the same direction, or trades that all close at the same profit target. These are characteristic of grid/martingale behavior.

How does grid trading perform in crypto markets?

Crypto markets trend more aggressively and for longer periods than forex. Grid trading in crypto faces more frequent catastrophic scenarios. Some traders run grid strategies in narrow price ranges on crypto pairs with defined range expectations — but the trending risk is more acute than in forex.

What happens to a grid EA during a major news event?

If the news event causes a sustained move outside the grid range, the grid accumulates losses as described. Some grid EAs include news filters that pause trading 30–60 minutes before and after high-impact events. This reduces (but doesn't eliminate) news-induced catastrophic trends.

Grid trading carries risks not fully reflected in typical backtest reports. Understand the catastrophic trend risk before deploying any grid EA with real capital. All trading involves risk of capital loss.