blog · ~6 min read

Grid Trading in Forex and Crypto: Parameter Design

Design a grid trading system for forex and crypto with spacing, lot sizing, and drawdown caps that survive ranging markets without blowing up on trends.

T By tradernewbie · Curated for beginners
#strategies#systems
Эта статья на английском. Открыть на вашем языке? Google Translate →

Интерактивные инструменты могут не работать в переведённом виде.

Grid Trading in Forex and Crypto: Parameter Design

Grid trading places a ladder of buy orders below price and sell orders above, profiting in ranging markets as price oscillates through the grid. It prints steady small gains for weeks, then loses catastrophically in a single trend. The strategy looks like free money until it isn't. The entire edge is parameter design that survives the trend — everything else is just decoration.

Core Concept: Harvesting Volatility With Bounded Risk

A grid is a volatility-harvesting structure, not an alpha source. It works because markets range roughly 70% of the time, and within a range price crosses the same levels repeatedly. Each crossing captures a small profit. The strategy fails when price trends in one direction and the grid keeps adding losing positions — that is the known, fatal failure mode, and every parameter exists to bound it.

The four core parameters define whether your grid survives. Grid spacing is the distance between orders — use 0.25–0.5x ATR(14) on the timeframe you run, because fixed-pip spacing (e.g. every 20 pips on EURUSD) breaks when volatility shifts while ATR adapts. Grid range is the upper and lower bound — a grid placed without bounds is a martingale that keeps adding into an infinite move, so set bounds at ±3 to ±5 ATR from the midpoint and stop adding beyond them. Order type is buy-limit below and sell-limit above for the profit-taking grid (the common ranging version), or stop orders for a trend-following grid. Lot sizing must be fixed per level, never martingale doubling — doubling turns a 10-level grid into a 1024x position at the bottom, which is exactly how grids blow accounts.

Worked example: EURUSD 1H, ATR(14) = 15 pips. Spacing = 0.3x ATR ≈ 5 pips. Range = ±4 ATR ≈ 120 pips total (24 levels). Fixed 0.01 lot per level. A 50-pip oscillation fills roughly 10 levels and closes them for profit as price reverts. The same grid in a 150-pip one-way trend hits all 24 levels and stops adding at the bound.

Practical Application: Parameter Tables and Drawdown Control

The parameter set differs by asset class because volatility and trending behavior differ sharply.

Forex vs crypto parameters:

Parameter Forex (EURUSD) Crypto (BTC)
Grid spacing 0.25x ATR(14) 0.5–1x ATR(14)
Grid bounds ±5 ATR from midpoint ±3 ATR from midpoint
Max open levels 8–10 6–8
Lot sizing Fixed, never martingale Fixed, never martingale
Hard equity stop 12% drawdown 10% drawdown
Trend filter Disable beyond 2 ATR of 50-mean Disable beyond 2 ATR of 50-mean
Event guard No grid 2h around BoJ/NFP No grid around CPI / FOMC

Daily operation checklist:

  • Confirm price is inside the defined grid range; no new levels beyond bounds.
  • Check the 50-SMA slope and ATR — if slope exceeds threshold or price is beyond 2 ATR of the mean, disable new placements (trend guard).
  • Count open levels; if max open levels (8–10) is reached, no new orders until some close.
  • Verify unrealized drawdown is below the equity stop (10–15%); if breached, close the entire grid and reset.
  • Confirm no high-impact news in the next 2 hours.

Management rules:

  • Max open levels: cap at 8–10. Beyond that, cumulative drawdown exceeds the grid's profit buffer.
  • Equity stop: if unrealized drawdown reaches 10–15% of account, close the entire grid and reset. Do not "wait for it to come back" — that is the trap that ends accounts.
  • Trend filter: disable new grid placement when price is beyond 2 ATR of the 50-period mean AND the 50-SMA slope exceeds a threshold. Strong trends are when grids bleed.
  • Reset, do not rescue: after the equity stop triggers, restart the grid from the new midpoint with fresh parameters — never add to a trapped grid.

Common Mistakes

  1. Martingale lot doubling. Doubling the lot at each level to "average down faster" turns a 10-level grid into a 1024x position at the bottom. Correction: fixed lot per level, always. The grid's edge comes from frequency, not from size.
  2. No grid bounds. An unbounded grid keeps adding into a trend until margin runs out. Correction: set hard bounds at ±3 to ±5 ATR and stop placing orders beyond them — the grid must have a defined failure point.
  3. Removing the equity stop. When drawdown hits 12%, traders disable the stop "because it will revert." Correction: the equity stop is non-negotiable. If it triggers, close the grid and reset. The one trend you refuse to stop on is the one that liquidates you.

Advanced Tips

Run grids only on instruments with a documented ranging bias — forex majors (EURUSD, GBPUSD) and high-cap stablecoin pairs range far more than single stocks or altcoins. Combine the grid with a regime detector: disable new placements when ADX rises above 30 (trending) and re-enable when ADX falls back below 20 (ranging). Track grid performance by regime in your journal — the ranging-regime subset should be sharply profitable while the trending subset is the loss source. For logging grid cycles and drawdown events, see /journal; for the full strategy set including trend-following counterparts, see /strategies; and use the ATR and ADX screener at /tools to qualify instruments before launching a grid.

Summary

Grids are a volatility-harvesting structure with a known fatal failure mode in trends. Run them only with ATR-based spacing, bounded ranges, fixed lots, max 8–10 open levels, and a hard 10–15% equity stop. If your grid has no upper bound on size or levels, it is not a strategy — it is a delayed liquidation.

Related market data, powered by TradingView.

Educational content · Not financial advice · Trade at your own risk