Hyperliquid HLP withdrawals have a four-day lock measured from the latest deposit. The documented legacy HyperCore user vaults have a one-day lock and a 10% leader profit share; HLP has no vault-owner profit share. In either case, recorded vault equity, withdrawal eligibility and cash received are separate checks. Custom HyperEVM vaults can use different accounting and exit rules.
Identify which kind of vault you hold
The word “vault” now covers more than one implementation. Hyperliquid's current vault overview, checked on 29 October 2026, describes builders creating customizable vaults on HyperEVM using CoreWriter and precompiles. Those facilities let contracts read HyperCore state and submit actions. The overview separately identifies the older native HyperCore vaults as legacy.
This comparison covers HLP and the documented legacy user vaults. For a builder's HyperEVM vault, identify its contract, share accounting, fee policy and redemption mechanism separately. A familiar underlying trading venue does not establish matching withdrawal terms.
HLP means Hyperliquidity Provider. It participates in market making and liquidations, supplies USDC in Earn and receives a portion of trading fees. Depositors share its strategy PnL, meaning profits and losses. The protocol-vault policy describes HLP's four-day lock. The legacy vault overview gives user-vault owners a 10% profit share and excludes protocol vaults from vault fees and profit sharing.
| Withdrawal question | HLP | Legacy HyperCore user vault |
|---|---|---|
| Which lock is documented? | Four days from the most recent deposit | One day |
| Does a leader take a profit share? | No vault-owner profit share | 10% of profit under the documented model |
| What does your balance represent? | Exposure to pooled strategy PnL | Exposure to the leader's pooled strategy PnL |
| What needs separate checking? | Personal lock, current claim and available exit | Personal lock, current claim, commission and available exit |
The legacy depositor guide confirms the one-day and four-day periods. It also warns that closing positions for a withdrawal may introduce slippage. Removing a vault-owner commission from the comparison does not remove strategy losses or execution costs.
Reconcile ownership before estimating a payout
Consider a hypothetical legacy user vault whose participants all contribute at the same valuation, with no intervening deposits or withdrawals. A depositor contributes 3,000 USDC, the leader contributes 5,000, and other participants contribute 42,000. Total starting equity is 50,000 USDC. Assume it later becomes 57,500 through strategy performance after ordinary trading costs, before any departing depositor's leader profit share.
These invented values illustrate accounting. The 15% growth is neither an observed vault result nor an expected return. Assume the depositor's lock has expired, the full exit is permitted and closing positions adds no further cost.
| Participant | Starting contribution | Ownership in this example | Gross claim at 57,500 USDC |
|---|---|---|---|
| Departing depositor | 3,000 | 6% | 3,450 |
| Leader's own capital | 5,000 | 10% | 5,750 |
| Other participants | 42,000 | 84% | 48,300 |
| Total | 50,000 | 100% | 57,500 |
The departing depositor's profit is 450 USDC. Applying the documented simple full-exit model gives a leader profit share of 45 USDC and a depositor receipt of 3,405 USDC. All amounts in this reconciliation are USDC:
3,450 = 3,405 + 45
Charging 10% against the entire 3,450 would deduct 345 USDC. That would confuse profit with the gross claim. The official depositor example applies the percentage to the gain above the original contribution. Our figures use that same limited accounting case with different inputs.
For a hypothetical HLP claim with identical starting ownership and identical gross growth, removing the leader share leaves 3,450 USDC before additional closing costs. Holding growth equal isolates the fee difference. It supplies no evidence that HLP and a user vault will earn comparable returns.
The leader's own 5,750-USDC gross interest is also distinct from the 45-USDC commission. Combining them would hide the difference between gains on contributed capital and compensation for managing other capital. This worksheet does not assert where the engine credits that commission.
Multiple deposits, partial withdrawals, prior losses and earlier commission settlements require the actual position history and applicable fee accounting. A shortcut using lifetime net deposits can fail to reproduce a current withdrawal estimate. Preserve exact decimal amounts when building that history; the API decimal-precision guide explains how rounding can change an otherwise plausible ledger.
Read the HLP clock from the latest deposit
The HLP withdrawal policy measures four days from the most recent deposit. It does not describe a separate unlock date for each contribution.
For an illustrative UTC timeline, suppose the first deposit occurs on 10 October 2026 at 08:00. Without another deposit, the four-day point is 14 October at 08:00. A further deposit on 12 October at 20:00 moves the four-day point to 16 October at 20:00. That is 60 hours later than the first date.
The top-up may be small relative to the original contribution, but the documented clock still references the latest deposit. A reader who records only the first entry can therefore predict the wrong withdrawal date. Keep the timestamp, time zone and latest deposit event together.
A one-day legacy user-vault lock is a different documented policy. Use the account's actual lock information when evaluating an exit, especially after more deposits. Neither a screenshot of vault age nor the time since someone else deposited establishes your eligibility.
Withdrawal can change the vault's orders and positions
The legacy leader guide describes a conditional process. If enough margin remains at the configured leverages, withdrawal leaves positions intact. Otherwise, margin-consuming orders are canceled in increasing order of margin used. If that still fails to free enough margin, 20% of positions are closed repeatedly until processing is possible. Leaders can also enable proportional position closing on every withdrawal.
This means two identical ownership claims can encounter different execution conditions. One exit may use available margin. Another may require trading into the current order book. Treat the displayed equity and the final proceeds as observations made at different stages, with possible price changes between them.
The cross-versus-isolated margin guide explains the broader collateral-boundary question. A vault ownership percentage alone does not describe the margin needed by its remaining exposure.
After a withdrawal changes positions or cancels orders, inspect the surviving exits too. The Hyperliquid normal TP/SL versus position TP/SL comparison explains why the type of attached exit matters. Do not assume an earlier order list still describes the vault after capital has left.
Check a leader's ownership after the proposed withdrawal
Legacy leaders must retain at least 5% ownership, and a withdrawal cannot leave them below that floor. The relevant fraction uses both remaining leader equity and remaining vault equity.
Take a separate hypothetical vault worth 100,000 USDC, with a leader interest worth 6,000. Ignore commissions, price movements, costs and all other exit restrictions. A 1,000-USDC leader withdrawal leaves 5,000 of 99,000, approximately 5.0505%. A 1,100-USDC withdrawal leaves 4,900 of 98,900, approximately 4.9545%.
The first candidate passes this ownership test; the second fails it. Subtracting 5,000 from the initial 6,000 would miss that the denominator shrinks too. Passing the fraction test still establishes only one condition. The lock, current limits and ability to execute remain separate checks.
An ordinary depositor should not apply the leader's ownership restriction to their own fraction. Record the account's role before interpreting a refusal. A restriction on the leader's remaining stake and a depositor's deposit lock have different causes and different resolutions.
Validate the claim, then confirm the completed movement
For native vaults, Hyperliquid's Info endpoint documents userVaultEquities for the owning account and vaultDetails with an optional user. The response schema includes leaderCommission, lockupUntil, maxWithdrawable, maxDistributable and alwaysCloseOnWithdraw. It also names vaultWithdrawalCanceled as an order status. These fields help investigate a proposed exit; the documentation's example values are not a current quote for your account.
The pinned official Python SDK method forwards the supplied user to userVaultEquities. It does not calculate the lock or reproduce the engine's fee accounting. Query the actual owning account address; substituting an agent signing address can return empty account data.
Use this sequence to make the evidence reviewable:
- Record the vault address, implementation class, owning account, account role and observation time. Match these across the interface and your records.
- Reconcile contributions and earlier withdrawals with current vault equity. Keep gross gain, leader profit share and possible closing costs in separate columns.
- Read the current personal lock and withdrawal estimate. Investigate the named limit fields rather than treating any one maximum as interchangeable with your own claim.
- After processing, verify the reduced vault interest, credited account balance and any order cancellations or closing fills. Explain discrepancies using actual events.
The Exchange endpoint documents native deposits and withdrawals as vaultTransfer actions. Leaving a vault and withdrawing through a bridge are separate movements. For the balance credited to a trading account, the Hyperliquid equity-versus-withdrawable-balance guide covers the next accounting question.
Keep the completion record tied to the requested amount and account. If equity falls by one amount while credited funds rise by another, reconcile commission, fills, fees and other activity at their own timestamps. A passing lock check resolves the time restriction; the final ledger establishes what actually moved.
