Hyperliquid trading strategies, backtested
Keel is a strategy builder for Hyperliquid perpetuals, and its public library holds Hyperliquid trading strategies backtested on the exchange's own price and funding history, after fees and slippage. Below is what each one does, when it tends to work and when it doesn't. Open any of them to see the rules and the full results, or test your own idea the same way.
What changes when you trade on Hyperliquid
The strategies here trade Hyperliquid's perpetual futures. There is no expiry, you can short a coin as easily as you buy it, and a funding payment changes hands every hour between longs and shorts. When longs are crowded they pay shorts, and when shorts are crowded the payment flips.
That hourly funding shapes every strategy below. A long-only trend strategy pays it while the trade is popular, which quietly eats into returns. A carry strategy does the opposite and makes funding its whole source of profit. Two of the strategies on this page are built around that payment rather than around price alone.
Shorting being cheap matters too. Many of these strategies hold longs and shorts of the same size, so a broad crypto selloff hits both sides and mostly cancels out. What is left is the difference between the coins they like and the coins they don't. That is harder to build on a spot exchange, and it is why the market neutral ideas here are a natural fit for this venue.
The coin list moves fast. New listings arrive, volume shifts from one sector to another, and a coin that led last quarter can be forgotten by the next. A rule tested across many coins tells you far more than one that looked good on a single chart, so most of the strategies here trade a basket and each strategy page shows how every coin in it did.
The strategies, one by one
Each line is the strategy's equity curve over its backtest. Open a strategy for its rules, its numbers and every setting tested.
- carrymedium risk
Funding Carry (Cross-Sectional, Hyperliquid)
On a perpetual, whichever side is crowded pays the other side every hour. This strategy ranks a wide set of Hyperliquid perps by funding rate, shorts the coins where longs pay the most and goes long the coins where shorts pay. The long and short sides are the same size, so the market's direction mostly cancels out and what is left is the funding you collect.
Open the funding carry backtest and rulesTends to work
Leveraged, crowded markets where funding is actually being paid and the gap between coins is wide. That has been Hyperliquid for long stretches.
Tends to struggle
When funding drifts toward zero there is little to collect. It also gets hurt by short squeezes: the coins it is long fall while the crowded coins it is short keep going up.
- momentum-carry-combomedium risk
Momentum + Funding Strategy on Hyperliquid
Two scores blended into one. The first is recent price trend: coins that have been beating the rest tend to keep doing it for a while. The second is funding: it prefers longs where holding is cheap and shorts where the other side is paying. Strong coins with cheap funding get the biggest longs, weak coins with expensive funding get shorted.
Open the momentum with funding backtest and rulesTends to work
Markets with clear winners and losers and active funding. Because the book is balanced long and short, it does not need the whole market to go up.
Tends to struggle
Sharp reversals, when last week's leaders crash and the book is leaning the wrong way. The funding tilt softens that but does not stop it. It trades often, so fees matter.
- momentumhigh risk
Momentum Strategy on Hyperliquid
Long only. Each day it buys the Hyperliquid perps that have gained the most against their peers recently, but only if they are trading on real volume, and sizes each one so they carry similar risk. A filter on the whole market moves the book to cash when crypto as a whole turns down.
Open the momentum backtest and rulesTends to work
Bull markets where a few coins break away and lead for days or weeks, the rotation crypto tends to produce when money is flowing in.
Tends to struggle
The turn at the top. The market filter only steps aside after the turn, so the drop right at the peak is where the losses come from. Choppy markets with no clear leaders cause false starts and fees.
- trend-breakouthigh risk
Channel Breakout Strategy on Hyperliquid
A trend follower that trades both directions. It draws a channel around each coin's recent average price, sized by how much the coin usually moves. The further price pushes outside the channel, the bigger the position, long above and short below.
Open the channel breakout backtest and rulesTends to work
Strong moves that keep going, in either direction. In crypto, enough breakouts keep running to pay for the ones that fail.
Tends to struggle
Sideways markets, where price pokes out of the channel and comes straight back. Each false break costs a little, and it has no brake on risk beyond its own sizing, so drawdowns can be deep.
- mean-reversionmedium risk
Mean Reversion Strategy on Hyperliquid
Buys the bounce after a panic. On hourly bars it looks for large coins that have dropped well below their normal range and are deeply oversold, but only while the coin is still in a longer uptrend. It sells once the oversold reading clears.
Open the mean reversion backtest and rulesTends to work
Choppy markets with lots of small overshoots and quick snaps back, the kind forced liquidations produce.
Tends to struggle
One-way selloffs, where oversold keeps getting more oversold. It has also been the weakest of these strategies over the recent backtest, with the bounces barely paying for the trading costs. It is here because knowing what has not worked is useful too.
- regimemedium-high risk
Regime Rotation (BTC Trend Switch)
One decision a day. While BTC is above its long-term trend line, hold an equal mix of the major coins. When BTC closes below the line, go to cash. No shorting and no leverage. The goal is to miss the worst of the bear markets, not to time every move.
Open the regime rotation backtest and rulesTends to work
Long uptrends led by BTC. It rides them with a diversified basket and steps aside before most of a big decline.
Tends to struggle
When BTC chops around its trend line, the switch flips in and out and each round trip loses a little. In falling markets it sits in cash, which protects money but makes none.
- index-long-shorthigh risk
Long HYPE / Short Alts Index (Market-Neutral)
A single view held all the time: HYPE, Hyperliquid's own token, does better than other large alts. It is long HYPE with half the book and short an equal basket of large alts with the other half, so the result depends on the gap between them rather than on where crypto goes.
Open the long hype, short alts backtest and rulesTends to work
Any market where HYPE beats the basket, rising or falling. It can hold up in selloffs that hurt long-only books.
Tends to struggle
This is a concentrated bet on one token. HYPE moves far more than the basket, so it drives most of the risk, and nothing in the rules changes if its lead fades. When HYPE lags, losses come fast.
Funding carry, start to finish
The most Hyperliquid-specific idea on the page, with its full backtest.
Hyperliquid, Aug 2024 to Sep 2026, default settings, net of fees + slippage + funding.
Look at the shape as well as the final number. Most of the time the strategy is quietly collecting funding on both sides of its book. The dips tend to come when a crowded trade unwinds in a hurry, usually a squeeze in the coins it is short.
Its strategy page also shows a price-only result, the same trades without the funding. The gap between the two is the funding the strategy earned, and for this idea it is most of the story. If you build a version of your own, that comparison is the first thing to check.
Four things to check before you trust a result
Look at the worst drawdown first
The total return tells you what you would have made if you had held on. The worst drawdown tells you whether you would have. A strategy that fell by a third before recovering needs a size you can live with through that fall.
Check where the return came from
On Hyperliquid, funding can be most of the result. Each library page shows a price-only Sharpe ratio beside the full result, so you can see whether a strategy made money from price moves or from collecting funding.
Count the trades
A strategy with a handful of trades rests on a handful of decisions, and a few lucky calls can make it look good. One with thousands of trades pays a lot of fees, so check that the result is after costs. Every library result is.
Prefer default settings to the best run
Any idea looks great with settings tuned to the past. The library headline is the default setting, and each page shows how the other settings did, so you can see whether the idea holds up or only works at one point.
Test a variation, or something new
The library is a starting point. The same tools that made these results are open to you.
- 01
Start from the closest strategy
Most ideas are a variation of something above. Open that strategy's page, choose Fork and customize, and you get your own copy with its settings in reach, ready to change and test again.
- 02
Or start from your own words
If nothing above is close, write the rule out in the app, the way you would explain it to someone at the next desk. Keel's AI builds it from the same blocks the library strategies use, so you can read what it made before you test it.
- 03
Backtest it on Hyperliquid history
The test runs over Hyperliquid's own prices and funding and charges fees and slippage on every trade, the same way the library results were made. Compare your version against the original on the same window and see whether your change helped.
- 04
Paper trade, then go live if it holds up
Paper trading follows the market as it happens, with no real money at stake. If the results hold up, link your Hyperliquid account. Your funds never leave it: Keel is allowed to trade there and has no way to withdraw.
More on testing: how a Hyperliquid backtest works, and building a strategy with AI from a written description.
Questions about these strategies
The engine behind every backtest is explained on the how it works page.
What are the most common Hyperliquid trading strategies?
For systematic traders, the usual families are trend following (momentum and breakouts), funding carry, mean reversion, and relative value trades such as one coin against a basket. Market making is common too, but it needs fast quoting rather than a daily or hourly strategy. The library above has a tested version of each of the first four.
Is there a funding rate strategy for Hyperliquid?
Yes. Funding carry shorts the coins where longs pay the most funding and goes long where shorts pay, with both sides the same size. Keel trades only on Hyperliquid, so it does not run arbitrage between exchanges. The carry trade happens between coins on the same exchange.
Do these strategies still work?
Each result runs up to the date shown on the strategy's page and is refreshed regularly. Some ideas have done well over that window and one has barely broken even. None of it says what happens next. That is why each page shows the monthly returns and how the other settings did, not just the headline.
Can I copy one of these strategies and run it?
Yes. Every strategy page has a button to run it as it is and another to fork it and change the settings. You can paper trade it on live prices first. To trade real money you link your own Hyperliquid account. Your funds stay in it, and Keel is allowed to trade but has no way to withdraw.
How are the backtests run?
On Hyperliquid's own price and funding history, with trading fees and slippage charged on every trade and funding applied to every open position. Fills are modelled, so the result is still a simulation. The how it works page explains the engine in detail.
Can I trade stock or commodity perps on Hyperliquid with these?
Not yet. Stock, index, commodity and currency perps on Hyperliquid are HIP-3 markets, and Keel doesn't support them yet. Support is coming soon. Every strategy here trades Hyperliquid's crypto perpetuals.
What does it cost?
Reading the library is free and needs no account. Backtesting, forking and paper trading start on the free plan. Running a strategy live on the free plan carries a 5 basis point fee on trades, and paid plans from $49 a month lower it and allow more strategies at once.
Put your version to the test
Fork a strategy above or start from your own rule, and see its backtest on Hyperliquid data before any money is involved.