strategiesJuly 9, 2026

Grid trading bots are everywhere on Hyperliquid. Here is how they work, what decides if they make money, and the one number you should check before deploying one.

Does Grid Trading Work on Hyperliquid?

Grid trading bots are everywhere on Hyperliquid. You have seen them mentioned in threads, promoted by builders, and occasionally advertised with screenshots of equity curves that look too good to be true.

The question people actually ask is simple:

does grid trading work on Hyperliquid, or is it a trap?

The answer is not yes or no. It depends on one number you can check in thirty seconds. Here is what that number is, why it matters, and how to check it yourself.

What a grid trading bot actually does

A grid bot places staggered buy and sell orders inside a price range you define. When the price drops, it buys. When the price rises, it sells. Repeat.

If BTC trades between $62,000 and $63,000 all week, a grid with orders every $100 captures the wobble. Buy at $62,100, sell at $62,200. Buy at $61,900, sell at $62,000. Each round-trip is one fill. The profit per fill is your grid spacing minus fees.

Here is what a 10-level grid looks like with $100 spacing:

PRICE
$63,000  ═══════════════════  UPPER BOUNDARY
$62,900  │  SELL  ████████  ─┐
$62,800  │  SELL  ████████   │  price moves up
$62,700  │  SELL  ████████   │  → sells filled
$62,600  │  SELL  ████████   │
$62,500  │  SELL  ████████  ─┘
$62,400  ├──────────────────  MID-POINT (no order here)
$62,300  │  BUY   ████████  ─┐
$62,200  │  BUY   ████████   │  price moves down
$62,100  │  BUY   ████████   │  → buys filled
$62,000  ═══════════════════  LOWER BOUNDARY

Zigzag diagram showing grid trading: buy orders fill on downward price moves, sell orders fill on upward moves

When price oscillates between the boundaries, buy orders fill on the way down and sell orders fill on the way up. The bot replaces filled orders automatically to keep the grid intact. If price breaks above $63,000 or below $62,000, the grid stops. You have either sold everything (breakout up) or bought everything (breakout down).

On Hyperliquid, a grid bot runs through agent keys: API keys that can place and cancel orders but cannot withdraw funds. Your capital stays in your wallet. The bot trades on your behalf with strictly limited permissions. This is the trust architecture: non-custodial, trade-only, revocable any time.

You can run a grid on spot (you hold the asset, no leverage) or on perpetuals (you hold a contract, with leverage and funding-rate exposure). The mechanics are the same. The math is different. The rest of this article uses BTC/USDC spot because it is the deepest spot pair on Hyperliquid. The same framework applies to perps with one extra check we will cover.

The fee math: your grid spacing must clear this floor

Every fill on Hyperliquid spot costs a maker fee. At the default Tier 0 with no staking discount, that fee is 0.040% per order.

A round-trip (one buy plus one sell) costs twice that: 0.080% .

Your grid spacing must be wider than that floor. Here is what different spacing levels look like on BTC/USDC.

Spacing Net profit per fill Estimated fills per day
0.1% 0.02% ~30
0.5% 0.42% ~6
1.0% 0.92% ~3
1.5% 1.42% ~2
2.0% 1.92% ~2

Below 0.5% spacing, fees consume nearly everything. Above 1.5%, you get too few fills to matter. The sweet spot is 0.5% to 1.0%: enough profit per fill to compound, enough fills per day to stay busy.

This math assumes BTC stays inside your range. When it does not, the profit calculation stops mattering. Which brings us to the number.

The number that decides everything: ADX

ADX stands for Average Directional Index. It measures trend strength on a scale of 0 to 100. It does not tell you which direction the trend is going. It tells you how strong the trend is, period.

The thresholds are simple:

A grid profits from sideways price movement. The moment a trend forms, the grid works against you. In an uptrend, you sell too early. You exit at $62,500 while price runs to $68,000. In a downtrend, you keep buying as price falls. You fill at $62,000, then $61,500, then $60,000, holding a growing bag of an asset going the wrong way.

ADX tells you which regime you are in. You check it before deploying the same way you check the weather before sailing.

The data: when it worked, and why it does not work right now

We pulled 180 days of daily candles for BTC from the Hyperliquid API and calculated ADX for every day. The results show two clear windows where a grid would have performed well, and one window (right now) where it would not.

Window Dates Duration Avg ADX BTC range Fills/day Net/fill Daily profit
1 Mar 18 – May 4, 2026 47 days 18.8 ~$55k–$65k ~6 0.42% ~2.5%
2 May 19 – May 31, 2026 12 days 21.4 similar range ~6 ~0.42% ~2.5%

Right now (July 9, 2026): ADX is 27.6 . This is trending territory. The Directional Movement Index confirms a downtrend bias. DMI-minus at 22.2 dominates DMI-plus at 14.1. A grid deployed today would buy every dip into a falling market and sell every brief rally. Net result: negative.

The lesson is not that grids are bad. The lesson is that grids are a tool for a specific market condition. You check the condition before you pick up the tool.

Perpetual grids: the extra check

If you run a grid on perpetuals instead of spot, you add one variable: funding rates.

Funding is a periodic payment between long and short traders designed to keep the perpetual contract price close to the spot price. When funding is positive, longs pay shorts. When negative, shorts pay longs.

We pulled 500 hours of BTC perp funding history from the Hyperliquid API. The average hourly funding rate over the last 30 days was +0.00056% . That translates to 0.0135% per day paid by longs.

This is negligible compared to grid profit. At 0.5% spacing with six fills per day, the grid earns roughly 2.5% daily. Funding costs 0.0135% of that. That is about half of one percent of your profit. Even the 95th-percentile worst hour only pushed funding to +0.00125%, or 0.03% per day if sustained. Still not a dealbreaker.

Funding will not kill your perps grid. A trend will. The ADX check matters far more than the funding check.

The 30-second decision checklist

Before deploying a grid bot on Hyperliquid, verify these five things. Every check has a corresponding API call you can make directly from the Hyperliquid info endpoint.

# Check API call Threshold
1 Is the market sideways? candleSnapshot (daily), compute ADX(14) ADX below 20-25
2 Which direction is the bias? Same candles → DMI+ vs DMI− No strong bias. If DMI− dominates, do not deploy a long grid
3 Is the pair deep enough for your capital? l2Book At least 10 levels of your position size visible in the orderbook
4 Is your grid spacing profitable? userFees (or hardcode Tier 0 rates) Spacing % > 0.08% (spot) or > 0.03% (perps). Prefer 0.5%+
5 If running perps, is funding acceptable? fundingHistory Average daily funding cost well below expected daily grid profit

If all five checks pass, deploy. If any check fails, wait. The API calls are free and rate-limited generously. Run them daily if you want to catch the next sideways window.

How to pull the numbers yourself

All data in this article came from a single Hyperliquid API endpoint: POST https://api.hyperliquid.xyz/info.

To replicate the checks:

That is four API calls. The ADX calculation requires a short script (about 40 lines of Python). Everything else is a single POST request.

The bottom line

Grid trading works on Hyperliquid when the market cooperates. It does not work when the market trends.

Right now, that number says wait.

The next time you see a grid bot promoted on your timeline, check the ADX first. If it is below 20, run the rest of the checklist. If it is above 25, close the tab. The Hyperliquid API makes both of these checks free and fast. Use them.