Hyperliquid Vaults & HLP (2026): Best Vaults by Risk-Adjusted Return, APR and Risks

1. What a Hyperliquid vault is

On Hyperliquid, a vault is simply an account whose balance is owned by many addresses. One address — the leader — places the trades; every depositor owns a percentage of the account equal to their share of total deposits. When the account's equity rises, each share is worth more; when it falls, each share is worth less. Because vaults live on HyperCore, the same order book everyone else trades on, every position, fill, deposit and withdrawal is public in real time. That is the core difference from copy-trading products on centralised exchanges, where you see whatever the platform chooses to show.

There are two kinds, and they behave differently enough that the rest of this guide treats them separately:

• Protocol vaults — chiefly HLP (Hyperliquidity Provider), which runs Hyperliquid's own market-making strategies, performs backstop liquidations, supplies USDC to Earn and receives a portion of trading fees. It is community-owned and charges nothing.
• User vaults — created by any trader who deposits 100 USDC and pays a 10,000 USDC creation fee. The leader runs whatever strategy they like on validator-operated perps and keeps 10% of depositors' profits. This is Hyperliquid's native form of copy trading.

One caveat for 2026: the docs now call these HyperCore vaults the "legacy" product. Builders can deploy tokenised vaults on HyperEVM (for example following ERC-4626) with full access to spot and HIP-3 markets. Everything below describes the native vaults in the app's Vaults tab, which is what "Hyperliquid vaults" means to almost everyone searching for it.

Live snapshot — November 3, 2026

Metric	Value
Total value in all open vaults (HLP + user vaults)	$240.6M
HLP (protocol vault) TVL	$150.3M
HLP APR shown in the app	-0.09%
HLP all-time PnL (since May 2023)	+$138.6M
Open user vaults	3,055
User vaults with TVL ≥ $1M / ≥ $100k	17 / 69
User vaults with TVL under $1,000	2,568

Two things stand out. HLP is about 62% of all vault capital. And the long tail of user vaults is very long: 2,568 of the 3,055 open user vaults hold under $1,000, which usually means the leader's own minimum deposit and nobody else. The vault list looks crowded; the vaults that matter fit on two pages. Source: api.hyperliquid.xyz vaultDetails and stats-data.hyperliquid.xyz vault list, read at build time.

2. Best Hyperliquid vaults by risk-adjusted return (live)

There is no official “best vaults” list, and the app’s APR column rewards whoever had the hottest month. This scoreboard is a computed ranking: user vaults ordered by how much they returned per unit of drawdown over the last 30 days, from each vault’s own history in the public API, with a cap that keeps a vault that fell more than 20% out of the ranking however high its return. HLP is shown underneath as the benchmark; it is a different product (next section). Past results, not investment advice.

High drawdown (not ranked): 19 of the 69 scored vaults had a 30-day max drawdown above 20% on November 3, 2026. They are counted here and left out of the ranking, whatever their return.

Our calculation from api.hyperliquid.xyz vaultDetails (each vault’s 30-day account-value and PnL history; leader’s stake = leaderFraction) and stats-data.hyperliquid.xyz (TVL, age), read November 3, 2026, 12:27 UTC. Each chart is the vault’s 30-day return path; the faint line is where it started. Past results, computed by the formula above; not investment advice.

What the scores say on November 3, 2026: 30 of the 69 scored vaults (ranked or above the cap) were up over 30 days. The median vault returned −0.9% with a 10.2% drawdown, against +0.03% and 0.32% for HLP, and 2 beat HLP on both counts. Thirty days is short: a high score this month says nothing about the next one, a leader can change strategy without notice, and a vault can stop taking deposits (marked above). Read the vault’s own page (positions, leverage, the longest chart) before treating a score as more than a starting point.

Next step: depositing into any of these vaults needs a funded Hyperliquid account. How to open and fund it (step 1 below); code HOLYGRAIL takes 4% off your own trading fees and does not change vault returns.

The ten largest Hyperliquid vaults by TVL

Size is not quality, but it is how the app’s list sorts. The ten largest open user vaults on November 3, 2026, with the app’s APR next to our 30-day return and the dollar results the list does not show (a leader can switch deposits off at any time):

Source: stats-data.hyperliquid.xyz vault list (TVL, APR, PnL, creation time) and api.hyperliquid.xyz vaultDetails (leader’s share, whether deposits are allowed, the 30-day series behind the 30-day return), read November 3, 2026. Ranked by TVL, not by quality.

What the table says on November 3, 2026: 8 of the ten had positive all-time PnL and 3 were up over the past 30 days in dollars. Leaders held between 5.5% and 52.3% of their own vaults; 4 of the ten sat below 10%, close to the 5% minimum, which means depositors' money carries most of the risk. 1 of the ten were not accepting new deposits, so you can only watch them. The app's APR does not document its window and does not match a simple 30-day annualisation, which is why we show a 30-day return next to it: when the two disagree, read the vault's PnL chart. A high APR next to a short age is a hot month extrapolated to twelve; section 5 explains what to check before you trust any row. This is a list of the largest vaults, not a list of recommendations.

Survivorship: how many Hyperliquid vaults are actually up

The vault list makes vaults look better than the odds, because the ones that lost and closed drop out of view. Of the 9,470 user vaults in Hyperliquid’s vault list on November 3, 2026, 6,415 (68%) are closed. The 3,055 still open, by size:

stats-data.hyperliquid.xyz vault list, read November 3, 2026, 12:27 UTC: open user vaults (HLP and its child vaults excluded) grouped by TVL; “positive” = all-time PnL above zero; APR as the app shows it. Closed vaults stay in the list with their history but are not counted in the buckets.

In the smallest bucket (under $10k), 17% of open vaults are up since they started; at $1M and more, 71% are. The median app APR is above HLP’s -0.09% in 1 of the 4 buckets. A positive all-time PnL includes the leader’s early months, so treat it as a floor test, not a quality grade.

3. Hyperliquid HLP: what it is, APR history, drawdowns

HLP is not an AMM curve; it is a set of active strategies with their own accounts that post and cancel limit orders on the book like any professional market maker. The parent HLP vault holds the aggregate; its components are child vaults you can see in the app and the API:

HLP component (November 3, 2026)	Account value	Role
HLP parent account	$10.8M	Direct perp account of the parent vault
HLP Strategy X	$100.8M	Market-making strategy
HLP Liquidator 2	$30M	Backstop liquidations
HLP Strategy B	$3.1M	Market-making strategy
HLP Strategy A	$3M	Market-making strategy
HLP Liquidator 4	$1M	Backstop liquidations
HLP Liquidator	$999,999	Backstop liquidations
HLP Liquidator 3	$644,213	Backstop liquidations

The child vaults add up to $139.5M; together with the parent's own account that is the $150.3M shown as HLP's TVL. Four income streams feed depositors:

• Market-making PnL. Spread captured across Hyperliquid's perp markets, net of adverse selection and inventory losses. This is the largest and most variable component.
• Backstop liquidations. Positions that fall below two-thirds of maintenance margin are taken over by the liquidator vault, a component strategy of HLP. The docs note these are profitable on average (the trader's maintenance margin is not returned); on most venues that flow goes to the operator. See the liquidation and insurance guide.
• A portion of trading fees. The fee docs say fees go entirely to the community — "HLP, the assistance fund, and deployers" — without publishing a split. On-chain, the assistance fund clearly takes the bulk: in the week to October 22, 2026 it spent $11.7M of USDC buying HYPE, while HLP's entire PnL for that week, fees included, was $128,646.
• Earn. The vault supplies USDC to Hyperliquid Earn.

HLP APR history: the last 90 days

The app shows only today's APR. To see how HLP's return actually moves, we annualise each window of its own account-value and PnL history (the API keeps one sample about every two weeks; the table starts at the last sample before the 90-day mark, so it covers 104 days, and the last window ends at the build):

Annualised = PnL in the window ÷ the average of opening and closing TVL × 365 ÷ days (the whole-period row uses the time-weighted average TVL). Our calculation from api.hyperliquid.xyz vaultDetails (HLP allTime series), read November 3, 2026. For comparison, the app's own APR on November 3, 2026: -0.09%.

Over these 104 days the windows ranged from −10.55% to 10.40% annualised while TVL moved down 39%, from $247.4M to $150.3M. For scale, the two crash fortnights of October 2025 and early February 2026 added $41.4M and $18.8M on their own: those two fortnights alone are 43% of HLP's all-time profit (below).

HLP vs user vaults: returns and drawdowns side by side

The same time-weighted method, applied to HLP and to all 69 user vaults scored in section 2 (open, at least $100k of TVL, November 3, 2026; the vaults above the drawdown cap are included here so the field is not flattered):

Medians and the 90th percentile are taken column by column, so a row is not one vault. ~90 days runs from the last sample of each vault’s all-time history at or before 90 days ago (104 days for HLP, whose samples are about two weeks apart); 58 of the scored vaults have that much history. Our calculation from api.hyperliquid.xyz vaultDetails, read November 3, 2026, 12:27 UTC.

Over the last 30 days HLP’s return was above the median scored vault’s, with a smaller drawdown. User vaults spread far wider in both directions: the top decile returned +11.0% in 30 days, while 39 of the 69 were down.

Upside and downside in one view. Since January 2024, HLP’s best sample window in its all-time history was Oct 1 – Oct 15, 2025 (+$41.4M, +9.70%) and its worst Oct 29 – Nov 12, 2025 (−$4.7M, −0.80%); 6 of its 78 windows lost money. The liquidation flow that hurts HLP when a position is pushed onto it at a bad price is the same flow that pays it in a cascade, so read the two extremes together (every loss, dated).

Liquid HLP (wHLP). HLP itself issues no token, but third-party protocols on HyperEVM tokenise deposits. Per HyperLend’s docs, USDC deposited into HyperLend's Liquid HLP vault goes into HLP on Hyperliquid; the vault issues ERC-20 shares (wHLP) that can be used as collateral in HyperLend's isolated pools, and redeeming burns the shares and withdraws the proportional USDC from HLP. That stacks HyperLend’s smart-contract risk on top of HLP’s, and HyperLend’s docs state no redemption time. We have not tested it; the numbers on this page are for HLP itself.

HLP rules that matter to a depositor

• No fee and no profit share — you receive 100% of your pro-rata PnL.
• 4-day lock-up from your most recent deposit. Deposit on September 14 at 08:00 and you can withdraw from September 18 at 08:00. Topping up restarts the clock for the whole balance.
• Returns are USDC-denominated. HLP is not a bet on HYPE or on crypto going up; it is a bet on Hyperliquid's trading activity and on its strategies' edge.
• Withdrawals can carry slippage if open positions have to be reduced to free your margin.
• HLP does not backstop HIP-3 markets. Each builder-deployed perp DEX has its own onchain backstop liquidator that falls back to auto-deleveraging (HIP-3 docs).

HLP APY today, and how its returns actually arrive

The number people search for as "HLP APY" is the APR on the vault page. It is not a promised rate: it is recent performance annualised, and it is the only return: HLP deposits earn no points, and the 2023–24 points program behind the Hyperliquid airdrop has ended. Hyperliquid does not document the window; as a cross-check, the past 30 days' PnL annualised on the starting balance gives 0.65%, in the same range as the displayed -0.09%. Here is what HLP actually earned, from the same API the app uses:

HLP PnL (November 3, 2026)	USDC
Past 24 hours	+$25,105
Past 7 days	−$359,282
Past 30 days	+$80,153
All time (since May 10, 2023)	+$138.6M

Three facts from HLP's full history (our aggregation of the vaultDetails all-time series, sampled about every two weeks since May 2023) explain why the APR alone misleads:

• Returns come in bursts. HLP's PnL rose $41.4M in the two weeks around the October 10, 2025 crash and $18.8M in the two weeks to February 4, 2026. Those two fortnights are 43% of everything it has made since launch. Between crashes, returns are thin: from February 4, 2026 to November 3, 2026, HLP's cumulative PnL changed by just +$557,643, with losing and winning months cancelling out. That is why the APR can sit in low single digits for months and then spike.
• TVL is well off its peak. HLP peaked at about $603.9M in early September 2025; at $150.3M it is roughly 75% lower. Deposits chase the APR after a good month and leave in quiet ones, so your share of the next burst depends on how much capital is sitting in the vault when it happens.
• Crash profits are the flip side of the risk. The same liquidation flow that made HLP money in October 2025 is what cost it money when a single account was liquidated into it at a bad price (below).

If you want an HLP APY estimate, the honest one is a range: low single digits in quiet months, much more in months with a liquidation cascade, and an occasional loss. Size the position for the quiet months.

Drawdowns: can HLP lose money? Every HLP incident, dated

Yes. "Market-neutral" does not mean "cannot lose", and Hyperliquid's own docs warn that trading is inherently risky and that vaults' past performance is no guarantee of future returns. The table below lists every known HLP incident and every large move in its PnL history. The last column is the change in HLP's cumulative PnL over the API sample window that contains the date (usually two weeks; one week in early March 2025) — so it nets the incident against everything else HLP did in that window.

Two lessons. First, the losing windows have been small next to HLP's $138.6M all-time profit — since 2024 no sampled fortnight has cost more than 1% of the TVL at the time (the worst, November 2025, about 0.8%). Second, the worst-case outcome was avoided by intervention, not by design: in the JELLY episode the unrealised loss on the inherited short grew sharply before validators delisted the market and settled it. HLP's PnL for that fortnight ended positive, but the fix was a governance decision. If you deposit, you are also trusting that process.

4. User vaults: the rules from the official docs

User vaults are where the "hyperliquid vaults explained" question gets interesting, because the leader is a stranger. Hyperliquid's rules are short and strictly enforced by the chain:

Rule	Detail
Leader's cut	10% of profits, taken on your gain when you withdraw. No management fee, nothing on losses.
Skin in the game	Leader must hold ≥ 5% of the vault at all times; a withdrawal that would break this is rejected.
Lock-up	1 day after your most recent deposit. The docs describe no leader-set or longer lock-ups.
Creation cost	10,000 USDC fee (distributed like trading fees) plus a 100 USDC minimum deposit. Name and description are permanent.
Tradeable markets	Validator-operated perps only — no spot, no HIP-3 markets.
Custody	The leader can trade the balance but cannot withdraw depositors' funds.
Closing	Leader must flatten every position, then Leader Actions → Close Vault; depositors receive their share automatically.

The 10% profit share, worked through

The docs' own example is the clearest. You deposit 100 USDC into a vault already holding 900 USDC, so you own 10% of a 1,000 USDC vault. Nobody else moves money and the vault grows to 2,000 USDC. Your share is now 200 USDC; the leader takes 10% of your 100 USDC gain, so you withdraw 190 USDC. Three consequences: the leader earns nothing on a losing period; the fee is on your gain since your deposit, so a vault up 300% lifetime but flat since you joined costs you nothing; and because the leader holds at least 5%, a 20% drawdown costs them a fifth of their own stake as well. Some third-party guides describe a high-water mark or end-of-day pricing; the official docs describe neither — the rule is simply 10% of your gain at withdrawal.

5. Hyperliquid vault APR (APY): what the number means and what it hides

The APR column on the vault list is annualised from recent performance. That makes it useful for HLP, whose returns are relatively smooth outside crash weeks, and close to meaningless for a young, leveraged user vault. The scoreboard puts a 30-day return next to its drawdown, and the largest-vaults table puts the app’s APR next to the 30-day return and the dollar PnL the list does not show.

Read the table as a warning, not a menu of the "best Hyperliquid vaults". The highest displayed APR among the ten is 255.0%, against -0.09% for HLP — that is a hot month extrapolated to twelve, and it says nothing about the drawdown risk that produced it. Across the 69 user vaults holding at least $100,000, 40 showed a negative APR and 22 had negative all-time PnL on November 3, 2026. What to look at instead:

• Account value and PnL charts on the vault page, over the longest window available, and the max drawdown the vault page reports. A vertical line up followed by a cliff is not a strategy.
• Past Month Return next to APR — the app shows both, and the gap between them tells you how much of the APR is extrapolation.
• Positions tab: how many markets, what leverage, how much unrealised loss is sitting in the book right now.
• Age and depositor tenure: the page shows how long depositors have stayed. A vault that kept its depositors for 600 days through a drawdown has been stress-tested; one that is three months old has not.
• Leader history: paste the leader address into an explorer to see their own trading and past vaults.

Is the vault's % my return?

Only if you were in for the whole window and the vault charges nothing. The APR and all-time PnL describe the account, not you. Your figure is the change in the vault's value between your deposit and your withdrawal, minus the leader's 10% of any gain. A vault that rose 20% while you were in pays you about 18%; one that rose before you arrived pays you nothing for it. The Your Performance tab on each vault page shows your own number.

6. How to deposit into a Hyperliquid vault (4 steps)

1. Open and fund a Hyperliquid account. The vault takes your deposit from this account's USDC, so open it with code HOLYGRAIL before the first deposit (a code cannot be added after $10,000 of volume), then deposit USDC from Arbitrum, Ethereum, Base or Polygon, or send BTC, ETH, SOL or another supported asset and sell it for USDC on spot. Trading and vault actions on HyperCore are gas-free; you pay only source-chain gas on the way in. Details in the Hyperliquid deposit guide.
2. Pick a vault at app.hyperliquid.xyz/vaults (Vaults in the top navigation; protocol vaults first, user vaults below, sortable by APR, TVL and age) and read its page in the order given in section 5: for HLP the account-value chart over the longest window, for a user vault positions, drawdown and age first.
3. Click Deposit, enter the USDC amount and confirm. It is a gas-less signature. Your share appears immediately under Your Performance on the vault page and in the Portfolio page's vault balance.
4. Note the unlock time, then withdraw when you want out. One day for user vaults, four for HLP, counted from this deposit and reset by any later deposit. After that, Withdraw on the vault page returns USDC to your perps account.

Withdrawing, and what happens to open positions

After the lock-up, click Withdraw on the vault page, enter the amount and confirm; USDC returns to your perps account (moving it off Hyperliquid to Arbitrum then costs 0.2 USDC plus a tiny HyperEVM gas fee). The docs spell out what happens inside the vault. If there is enough free margin to keep the open positions at their set leverage, nothing changes. If not, open orders that use margin are cancelled first, smallest margin first. If that is still not enough, 20% of positions are closed at market, repeatedly, until your withdrawal can be paid. Leaders can alternatively set the vault to always close positions proportionally so liquidation prices stay put. Either way, one large withdrawal during a drawdown forces the vault to realise losses — which is why big depositors leaving hurts the people who stay.

7. Hyperliquid vault risks, in one table

A vault page shows performance, not risk. These are the ways depositors lose money, which vaults each applies to, and what has actually happened (dates and amounts from Hyperliquid's statements and the API):

Two risks no table captures. Strategy drift: a vault's description cannot be edited, but its behaviour can change, so a "BTC swing" vault now holding dozens of altcoin positions is a different product from the one you joined. Survivorship: closed vaults drop off the list, so the vaults you see look better than the odds you face. Size any vault deposit as a leveraged trade you do not control.

8. HLP vs user vaults vs trading yourself

	HLP	User vault	Your own account
Who trades	Hyperliquid’s own market-making and liquidation strategies	The vault leader (a trader or a bot)	You
Fee / profit share	None	10% of your profits, paid to the leader	Trading fees only (0.045% / 0.015% base)
Lock-up after each deposit	4 days	1 day	None
Markets	All validator-operated perps, liquidations, USDC in Earn	Validator-operated perps only (no spot, no HIP-3)	Perps, spot, HIP-3
Leader’s own stake	Community-owned	At least 5% of the vault, enforced	100%
Referral / staking discounts	Do not apply	Do not apply	Apply — 4% off with HOLYGRAIL
Return driver	Spread capture, liquidations, fee share	Leader’s trading	Your trading

Which one should you use?

• You hold idle USDC on Hyperliquid and want exposure to the exchange's activity rather than to price direction → HLP. Accept the 4-day lock and a return that varies from month to month; size it so a losing month does not matter.
• You want a specific trader's strategy and can read their positions → a user vault, treated as one leveraged position with a fixed cap (many traders use 10–25% of trading equity across two or three vaults with different styles). Prefer vaults older than a year whose depositors have stayed.
• You want a fixed yield → neither. There is no fixed rate anywhere in the vault system. Compare HYPE staking (HYPE-denominated, 7-day unstake, fee discounts) or the wider DEX passive-income options.
• You trade actively yourself → the cheapest improvement is usually reducing your own fees before chasing vault yield: join with a referral code and stake HYPE. See the fee guide and the HOLYGRAIL referral code page.

9. HLP vs other protocol vaults: Lighter LLP, Jupiter JLP, GMX GLV

Other perp DEXs have their own "house" pools, and they are not the same product. HLP and Lighter's LLP are actively managed USDC vaults that make markets and absorb liquidations; Jupiter's JLP and GMX's GLV are liquidity pools that act as the counterparty to every trader and hold volatile tokens, so their value also moves with SOL, ETH or BTC.

If you want an HLP-style vault on a second venue, Lighter's LLP is the closest match — USDC in, market making and liquidations out, no lock-up in the docs — with the catch that you must stake LIT to get deposit capacity. Our Lighter LLP and staking guide has its dated returns and walks through it (Lighter’s code and sign-up: Lighter referral code). For JLP and GLV, the fee share is higher on paper but you also carry the pool's token exposure and trader PnL — see Hyperliquid vs Jupiter and Hyperliquid vs GMX.

10. Running your own vault

Anyone can lead a vault: choose a permanent name and description, deposit at least 100 USDC and pay the 10,000 USDC creation fee. On the trade page you switch the address dropdown from "Master" to the vault and trade on its behalf. You must keep at least 5% of the vault at all times, you can only trade validator-operated perps, and you close the vault by flattening every position and choosing Leader Actions → Close Vault. With a 10,000 USDC entry cost this is for traders who already have a track record or an audience; the 2,568 sub-$1,000 vaults on the list are what happens when it is treated as a side project. Firms that manage money for clients should read our desk and institutional setup guide first: it covers multi-sig custody, sub-accounts and the legal terms.

FAQ

Tax: see the DEX tax reporting guide. Not tax advice.

Disclosure: HOLYGRAIL is our Hyperliquid referral code and we also link to Lighter with our referral; we earn a share of referred accounts' trading fees. Vault rules (profit share, lock-ups, withdrawal handling, creation cost) are from Hyperliquid's official documentation (re-read October 19, 2026); TVL, APR, PnL and vault counts come from the public API at build time and change constantly; the scoreboard, survivorship and HLP figures are our own calculation from the API's histories, with the formulas stated next to each table; the ranking is a formula, not a recommendation. Vaults can lose money; nothing here is investment advice.