You can copy a trader on Hyperliquid by depositing into their vault or by mirroring their wallet with a bot. How each one works, what it costs, what Hyperliquid’s own data says about the results, and how to vet a leader before you follow one.
November 15, 2026
17
min to read
The short answer
You can copy trade on Hyperliquid three ways: deposit into a vault, where a leader trades a shared pool and keeps 10% of the profits; run a bot that mirrors a wallet’s trades in your own account; or copy a tested rule instead of a person. Most vaults with a real record are down all-time, so vet first.
Copy trading means letting someone else’s decisions drive your money. On Hyperliquid there are three ways to do it, and they differ in where your money sits and who decides each trade.
| Vault | Mirror bot | Copy a strategy | |
|---|---|---|---|
| Where your money sits | In the vault, an account the leader trades | In your own Hyperliquid account | In your own account |
| What you copy | The leader’s whole book, at the vault’s leverage | One wallet’s trades, at a size you set | A rule, with a record you can test |
| Who decides each trade | The leader | The leader, with a delay | The rule |
| What it costs | 10% of your profit to the leader of a user vault; HLP takes nothing | The tool’s price: a plan, a fee per trade or a profit share | The price of the tool that runs it |
| Getting out | Withdraw one day after your last deposit (HLP: four days) | Stop the bot and close the positions | Stop the rule and close the positions |
| Main risk | The leader’s losses are yours, and big withdrawals can force closes | Delay, slippage and the leader’s leverage | The rule stops working when the market changes |
Most vaults on Hyperliquid’s vaults page are user vaults built on the original design from 2023, which its docs now call HyperCore vaults (legacy). Their rules are built into the exchange, and the leader’s page spells them out:
The depositors’ page shows the math. You put 100 USDC into a vault that holds 900, so you own 10%. It grows to 2,000 USDC with nobody else coming or going. Your 10% is now 200 USDC, your profit is 100, and the leader takes 10% of that profit, so you withdraw 190, give or take some slippage as positions close.
Each vault’s page shows its return, total deposits, profit and loss, max drawdown, volume, open positions, trade history, and how long depositors have stayed. Those are the numbers to check, and the checklist below says how.
Hyperliquid’s docs now point builders to a newer kind of vault on HyperEVM, Hyperliquid’s smart-contract chain. Builders can tokenize them, set their own accounting and let them trade spot and HIP-3 markets, which legacy vaults can’t. The other side of that freedom: fees, lock-ups and who trades are whatever each builder writes, so read the contract and its docs. The 10%, 5% and one-day rules above apply to legacy vaults only.
HLP, the Hyperliquidity Provider, is a protocol vault: Hyperliquid’s own, and “fully community-owned”, in the docs’ words. It “provides liquidity to Hyperliquid through multiple market making strategies, performs liquidations, supplies USDC in Earn, and accrues a portion of trading fees.” Unlike a user vault, it charges no fees and takes no profit share. Its lock-up is four days: a deposit made on September 14 at 08:00 can be withdrawn on September 18 at 08:00.
A record like that is not a promise. HLP takes over liquidations that the order book can’t absorb, and as a market maker it can end up on the wrong side of a sharp move. Those are the moments it loses money. On March 12, 2025, a trader’s huge leveraged ETH long was liquidated and left HLP down about $4 million, around 1% of its $451 million at the time, The Block reported; Hyperliquid then cut max leverage on BTC and ETH to 40x and 25x. Two weeks later, a squeeze on the JELLY memecoin put HLP about $12 million under water on paper, and Hyperliquid’s validators voted to delist the market. On November 12, 2025, an alleged manipulation of POPCAT left HLP with $4.9 million of bad debt, per CoinDesk. In the API’s history, sampled about every two weeks, HLP’s two worst stretches end on the days of the ETH and POPCAT losses: a loss of $4.7 million from October 29 to November 12, 2025, and $4.0 million from March 5 to March 12, 2025. Shorter swings inside those steps don’t show. We cover these losses, and the rest of Hyperliquid’s record, in is Hyperliquid safe?
Its size also moves with money coming and going, not only with trading: the vault shrank by about three quarters from its 2025 peak while its profit kept rising. Size alone says little about how a vault trades.
Every position and fill on Hyperliquid is public. Anyone can ask Hyperliquid’s API for an address’s open positions and recent fills, or subscribe to them live. A copy bot does exactly that for the wallet you pick. When the leader trades, it places a matching order in your account. Most sign it with an API wallet you approved: a separate key that can trade for you but, per Hyperliquid’s SDK, can’t transfer or withdraw funds. Others create a wallet for you and trade from it. Our guide to Hyperliquid trading bots explains that key in full.
Most tools let you size each copy in one of three ways: a fixed amount per trade, the same share of your account that the leader puts in, or a multiple of the leader’s size. Good ones add caps on leverage and position size, and a stop that ends the copy after a set loss. These are the tools we could verify on their own pages:
| Tool | How it copies | What it charges |
|---|---|---|
| Hyperdash | Mirrors, or takes the other side of, up to 10 wallets you pick, with leverage caps and asset filters | A builder fee of up to 0.05% on copy orders |
| hypercopy.co | A Telegram bot: your size as a share of theirs, a fixed amount or a % per trade; new trades only | A 0.05% builder fee on each perp order; no subscription |
| AlphaYou | Follows the trader about every 20 seconds, with a risk multiplier and one stop for the whole copy | 0.05% of each copied trade |
| Dexly | A share of the leader’s size, with a leverage cap, slippage tolerance and a pause at a drawdown limit | No extra copy-trading fee, according to its copy guide; any builder fee shows in the app before you approve it |
| WakeTrades | Up to three leaders (five on its Advanced level) at four exposure levels; only trades opened after you start | 19.50 USDC a month at its founder price, nothing per trade |
| Hyperbot | A fixed amount or the leader’s share of capital; pauses a leader after three losing trades in a row | Not published |
| Pear Protocol | Only trades its leaders place through Pear, at the same share of equity; entries it can’t copy within 10 minutes are dropped | Free to the follower; Pear’s own trading fee applies |
| PvP.trade | Copies PvP users who opt in, from Telegram | The leader takes 10% of your profit when a position closes |
Hyperliquid’s own trading fees come on top of all of these. Read the fine print as closely as the price. What does stopping do? On Hyperdash, Stop “always closes the wallet’s open positions”; elsewhere, positions can stay open until you close them. What happens to the trader’s open positions when you start? hypercopy.co says they “are not copied. You join on their next move.” Which trades get skipped? The same FAQ lists orders under $10 and orders that arrive after the price has moved more than 1%. And check the tool still copies at all: WunderTrading, whose 2025 guide to Hyperliquid copy trading still ranks, now says “Copy trading was discontinued and no longer exists on the platform.”
For most of the people who tried it through vaults, no. Hyperliquid publishes a list of its vaults, open and closed, with each one’s profit or loss to date. We counted it on November 14, 2026:
The pattern in the chart is survivorship bias: judging by the ones still standing. Small vaults lose most often (83% of open ones under $10,000 were down) and many of them close. The vaults page, sorted by deposits or returns, puts the survivors on top, and even among open vaults with $100,000 or more, 48% were down all-time. A leaderboard tells you who has done well so far, not who will keep doing well.
One caution on reading these numbers: a vault’s all-time result is not every depositor’s result. Someone who joined after a drawdown and left at a high did better than the vault; someone who joined at the top did worse.
Research on copy trading elsewhere points the same way. In an experiment published in Management Science in 2020, people who could see others’ results took more risk, and more still when they could copy them directly; the authors conclude “that copy trading leads to excessive risk taking.” A 2024 study of two large crypto copy-trading platforms found that copiers follow the platform’s rankings, which “are often easily gameable by unscrupulous leaders who prey on novice copiers.” And a study of 1,084 wikifolios, a German social-trading product, found that “on average, wikifolios do not outperform the market.” If an offer promises steady returns from copying, our guide on how to tell a real trading tool from a scam lists the red flags.
The leader’s leverage is your leverage. Leverage lets a position be bigger than its margin, the money behind it. Copy a 20x long on ETH with isolated margin, where the position has its own margin, and a drop of about 3% liquidates it on Hyperliquid; at 10x, about 8%. The bots guide shows the formula with BTC.
Drawdowns are harder to win back than to lose. A drawdown is the fall from a peak to the next low. The bigger it is, the bigger the gain you need just to get back to even:
| Loss from the peak | Gain needed to get back |
|---|---|
| 10% | 11% |
| 20% | 25% |
| 30% | 43% |
| 50% | 100% |
| 75% | 300% |
| 90% | 900% |
Delay eats small edges. Say a leader makes 1% on an average trade and your copy fills 0.25% worse going in and 0.25% worse coming out. You keep half of the gain, and trading fees still come out of that half. The leaders who trade most often for small moves are the hardest to copy.
Check these on the vault’s page or the wallet’s history before you put money in:
A person’s trades are the output of a process you can’t see. A strategy is the process itself. When you copy a rule instead of a trader, you can test it on past data before any money goes in, size it to your account, keep it running after its author loses interest, and tell when it breaks, because you know what it is supposed to do.
The steps are simple. Write the rule in plain words: entry, exit, stop and size. Test it on past data. Look at the worst drawdown and the number of trades, not only the return. Then run it on paper or small. Our guide on how to backtest a trading strategy walks through each step.
Ethy doesn’t copy wallets, and it doesn’t deposit into Hyperliquid vaults, HLP included. It takes the strategy route:
One day after your most recent deposit for user vaults, and four days for HLP. Vaults that builders deploy on HyperEVM set their own rules, so check each one.
In a user vault on HyperCore, the leader gets 10% of the profits and must keep at least 5% of the vault in it at all times. Opening a vault takes a 10,000 USDC gas fee and a 100 USDC minimum deposit. HLP, the protocol vault, charges no fees and takes no profit share.
No vault is. HLP has made money over time, but it takes over liquidations and makes markets, so it can lose in sharp moves. Its worst stretch in the API’s history was a loss of about $4.7 million from October 29 to November 12, 2025. Money you deposit is locked for four days after each deposit.
Yes. A copy bot should only need an API wallet that your main wallet approves. Per Hyperliquid’s SDK, an API wallet can place trades but has no permission to transfer or withdraw funds. Never give any copy-trading tool your seed phrase or your main private key.
It depends on the tool. On Hyperdash, Stop always closes the copy wallet’s open positions. On some other tools, stopping only ends the copying, and your positions stay open until you close them. Find out before you need it. In a vault there is nothing to stop: you withdraw, once the lock-up has passed.
It depends on the tool. hypercopy.co says a trader’s existing positions are not copied and you join on their next move, and WakeTrades also copies only trades opened after you start; Hyperdash copies open positions too, unless you turn on its “Only copy newly opened positions” setting. A vault is different: your deposit buys a share of the whole account, open positions included.
The user vaults on HyperCore can’t: they trade only the main exchange’s perps, not spot and not HIP-3 markets such as stocks. The newer vaults that builders deploy on HyperEVM can trade spot and HIP-3. A mirror bot can copy HIP-3 trades if the tool supports those markets.
For most vault depositors on Hyperliquid it hasn’t been: of the user vaults that ever gained or lost $1,000 or more, about 77% were down all-time on November 14, 2026. Copying can work with a leader whose record is long, steady and modestly leveraged, but the past record is the only evidence, and survivors look better than the average.
No. Ethy doesn’t copy wallets and doesn’t deposit into Hyperliquid vaults. You can copy a Top Strategy, which deploys a strategy’s published settings at your size from your own wallet, or describe and backtest a rule of your own.
Trading involves risk, including the loss of your money. Ethy AI is a research and trading tool and does not give investment advice or recommendations. Backtests are simulations on historical data and do not predict future results.
Deploy your AI trading agent in one click.
