Risk and range breakouts

A grid is low‑risk in sideways markets — not no‑risk. Its single most important risk is a breakout: price leaving the range you set. This page explains what happens then, and the three tools GRIDer gives you to manage it.

What a breakout does

Inside the range, a grid is balanced — buying dips, selling rips. At the edges, that balance becomes a directional position:

• Price breaks above the upper limit. The grid has been selling all the way up. A LONG grid ends flat/in cash; a SHORT grid ends fully short and keeps losing if price keeps rising.
• Price breaks below the lower limit. The grid has been buying all the way down. A LONG grid ends fully long and keeps losing if price keeps falling; a SHORT grid ends flat/in cash.

In other words: a grid quietly converts a sustained trend into the worst position to hold against that trend. This is inherent to grid trading on any platform — the skill is in choosing a range that price is likely to respect, and in setting exits.

Tool 1 — Backtest the range

The cheapest risk management is choosing a sensible range in the first place. Use backtesting to replay real recent prices and see whether your range held or got broken, and how many trades it would have produced. The Auto button proposes a range from the last 7 days as a starting point.

A wider range breaks out less often but trades less; a tighter range trades more but is more exposed. Backtesting makes that trade‑off visible before you risk anything.

For a quick boundary check outside the app, use the Range Boundaries PnL Calculator to see how a long grid behaves if price reaches either edge of the range. For SHORT grids, use the Short Grid PnL Calculator to map the upper-edge risk before deploying capital.

Tool 2 — Stop-loss

A Stop-loss is a price at which GRIDer closes the whole grid to cap a losing breakout. It is backend‑managed: GRIDer watches the price, detects the cross, and closes your position at market — it does not rely on a native exchange trigger.

• For a LONG grid, the stop‑loss sits below the lower limit.
• For a SHORT grid, it sits above the upper limit.

How the close is made robust

Closing at market on an illiquid token can fail if a naive order can't cross the spread. GRIDer closes carefully:

1. It reads your real position from the exchange (not its own bookkeeping).
2. It prices off the live mid, not a stale trigger price.
3. It sends a reduce‑only order with escalating slippage tolerance (3% → 6% → 10%), re‑checking your position after each attempt, up to three tries.

If a residue remains after all attempts (e.g. extreme illiquidity), GRIDer does not pretend the grid closed cleanly: it closes the grid with a ‑residual reason, logs it, and sends you an urgent alert so you can finish the close manually on the exchange. This exists because a silent half‑close once left an orphaned position in an illiquid token — GRIDer is built to surface that, not hide it.

Tool 3 — Close Grid Price

A Close Grid Price is the opposite of a stop‑loss: a target at which you're happy to take the grid off the table. When hit, GRIDer closes the grid and cancels remaining orders.

• For a LONG grid, it sits above the upper limit.
• For a SHORT grid, it sits below the lower limit.

Use it to lock in a result when price reaches a level you consider "done," rather than letting the grid ride into a possible reversal.

Other risks to keep in mind

• Leverage (perps). Leverage multiplies breakout losses and introduces liquidation risk. Treat a stop‑loss as mandatory when using it. See Spot vs perpetual grids.
• Fees on tight grids. Very tight spacing can let fees eat the spread. See Fees and funding.
• Funding drift (perps). A persistent adverse funding rate is a slow cost; check it on the market before committing to a direction.
• Insufficient margin. If other positions consume your margin, the grid pauses rather than misbehaving — but it isn't trading while paused. See Troubleshooting.

A practical risk checklist

• [ ] The range is one price is likely to stay inside for your horizon.
• [ ] You ran a backtest and the result looks reasonable.
• [ ] A stop-loss is set (especially with leverage).
• [ ] You can accept ending fully long or fully short if the stop doesn't trigger in time.

Next: set it all up in Create a grid.