Anello Team · Oct 25, 2026 · 6 min read

How to Choose a Grid Trading Range and Levels

Set a grid's range to where price has actually traded, then pick a level count whose spacing beats round-trip fees several times. Worked numbers inside.

To choose a grid trading range, start from where price has actually traded, not where you hope it goes. Put the lower bound where you would be comfortable holding a full position and the upper bound where you would be happy to have sold. Then pick the number of levels so the gap between rungs is several times your round-trip fee cost.

A grid bot places buy orders below the current price and sell orders above it. Each time a buy fills, it places a sell one rung higher, and the reverse. Profit comes from price crossing rungs again and again. If the idea is new, start with what grid trading is. This post covers the two settings that decide how a grid behaves: the range and the level count.

Why is the range the most important grid setting?

The range decides whether price spends its time inside your ladder, where the grid trades, or outside it, where the grid sits still. Level count, spacing and order size all follow from the two bounds. A well-tuned grid with the wrong range earns nothing.

Two mistakes bracket the choice:

  • Too narrow. The first ordinary move carries price out of the range, and the grid stops cycling.
  • Too wide. Price spends its time in the middle of the ladder and crosses few rungs, or the capital is spread so thin that each fill barely covers fees.

How do you pick the lower bound?

Pick the lower bound at the price where you would be comfortable holding a full position in the asset. When price falls to the bottom rung, every buy has filled. The grid now holds inventory bought on the way down, at an average price above the market, and that inventory is at a loss until price recovers.

So the question is not “how low could it go?” It is “where would I be fine owning all of this?” Anchor the bound to where the asset has traded over recent weeks, and leave enough room below the current price for normal dips to fill buys and get sold back. A grid does not protect you on the way down. It accumulates.

How do you pick the upper bound?

Pick the upper bound at the price where you would be happy to have sold. When price rises past the top rung, every sell has filled and the grid is fully in USDC. If the rally continues, the grid misses it. That is missed upside, not a realized loss, but it is why a range so tight that a routine rally clears it in a day is a poor fit.

A worked example

Say an asset trades at $100 and has spent the last several weeks between about $85 and $120. A range of $85 to $120 brackets that history with the current price near the middle.

Anello uses geometric spacing only, so every rung sits the same percentage above the one below it. With N levels, the step per rung is:

step = (upper / lower) ^ (1 / (N - 1)) - 1

For $85 to $120, the ratio is about 1.41:

LevelsStep per rungNotes
11about 3.51%Large, rarer fills
20about 1.83%Common middle ground
40about 0.89%Frequent, smaller fills
60about 0.59%Thin margin over fees
116about 0.30%Anello’s minimum step

Now size it. Anello takes the order size in USD per order, and every rung has the same dollar value, so the token amount varies by price. With 20 levels at $50 per order, one rung near the current price is held back (placing it would fill at once), which leaves 9 buys below $100 and 10 sells above it. The preview lists what you need in your account to start: about $450 of USDC for the buys and about 4.53 tokens (around $453 at $100) for the sells.

If price falls through the whole range to $85, all 9 buys fill. The grid then holds about 9.46 tokens, worth about $804 at $85, against roughly $903 at the start. That is the number to accept before you launch, not after.

How many levels should a grid have?

Use as many levels as you can while keeping each step at a comfortable multiple of your round-trip fee cost, about five times or more. More levels mean smaller profit per cycle but more frequent fills. Fewer levels mean the opposite. There is no single correct count.

A round trip on Hyperliquid, a decentralized exchange, pays the venue fee twice. At the base tier, spot fees are 0.040% for maker orders and 0.070% for taker orders, and tiers improve with 14-day volume (Hyperliquid fees, checked October 25, 2026). Through Anello you also pay a 0.02% builder fee on spot sell fills only; buys carry no Anello fee.

  • At a 0.04% fill rate: 0.08% + 0.02% = about 0.10% per round trip.
  • At a 0.07% fill rate: 0.14% + 0.02% = about 0.16% per round trip.

Against 0.16%, a 1.83% step keeps most of each cycle. A 0.59% step keeps under three quarters. At 0.3%, fees take more than half. The fees and risks article works through the same math.

Anello enforces two limits at preview so a grid cannot be set up to lose to fees by design:

  • Minimum step of 0.3% per rung. Below this, most fills earn less than fees, and on coarse-tick pairs adjacent rungs round onto the same price.
  • At most 900 levels. You can set density by level count or by step percentage; both are checked against the same ceiling.

How much capital does a grid need?

A grid needs USDC for every buy rung and the base token for every sell rung, each worth your per-order size. Each order must be at least $20, and after Hyperliquid’s price and size rounding every rung must stay above Hyperliquid’s $10 minimum order value. If a full sweep to the bottom would leave you with more inventory than you want to hold, lower the size per order or narrow the range.

What happens when price leaves the range?

The grid stops trading until price returns. Below the range you hold the inventory at a loss. Above it you hold USDC and miss further gains. Anello grids run on spot markets only, so there is no leverage and no liquidation price, but there is also no built-in stop-loss. Choosing a floor you can hold at is the risk control. The risks of grid trading are worth reading before you commit capital.

If price breaks out and settles into a new band, you can extend a running Anello grid by adding rungs below or above at the same spacing, without touching the resting orders. Narrowing or re-centering the range means stopping the bot and starting a new one.

Check the exact grid before you launch

On Anello’s grid bot, the preview shows every order’s exact price and size after Hyperliquid’s rounding, plus the USDC and tokens you need. Those are exactly the orders the bot places. You can also backtest the same settings over the last 7, 30, 90 or 180 days of Hyperliquid prices, with fees. A backtest shows how the range would have behaved, not how it will. The bot trades through a revocable agent wallet that can place and cancel orders but cannot withdraw. How grid trading on Hyperliquid works covers the rest of the setup, and how grid bot PnL is calculated explains what you see once it runs.

Sources

Nothing in this article is financial advice. Grid trading can lose money, including holding inventory at a loss when price falls below your range.

Frequently asked questions

What price range should I set for a grid bot?
Use the band the asset has actually traded in over the period you expect the grid to run, not a price target. Set the lower bound at a price where you would be comfortable holding a full position and the upper bound where you would be happy to have sold. Check that the current price sits well inside the range.
Should a grid range be wide or narrow?
It is a trade-off. A wide range keeps working through bigger swings, but with the same capital each rung is smaller or further apart, and price crosses rungs less often. A narrow range fills more often but breaks sooner. Match the width to how far the asset really moves.
How many grid levels should I use?
Use enough levels that price crosses rungs often, but not so many that the spacing between rungs gets close to your fees. A common rule is spacing of at least five times your round-trip fee cost. On Anello the spacing must be at least 0.3% per rung and a grid can have at most 900 levels.
What happens if price leaves my grid range?
Below the range, every buy has filled and the grid holds inventory bought at higher prices until price comes back. Above the range, the grid has sold its inventory and sits in USDC while the rally continues. On a spot grid neither case is a liquidation, but the grid stops earning until price returns to the range.
Can I change a grid's range after it starts?
On Anello you can extend a running grid by adding rungs below or above at the same spacing, and you can add funds to raise the size per order. You cannot narrow or re-center the range in place. For that, stop the bot and start a new one.