HLP and HyperCore Vaults Explained
How Hyperliquid vaults pool capital, how HLP differs from user-created vaults, and what depositors are exposed to.
HyperCore vaults let multiple depositors allocate capital to one trading account or strategy. The best-known example is HLP, Hyperliquid's protocol vault, but the vault system also supports user-created strategies.
A vault is pooled trading capital
A depositor receives a proportional claim on the vault's equity. If the strategy makes money, the value of that claim can rise. If the strategy loses money, the depositor shares that loss.
That makes a trading vault different from a lending pool. The return is driven by the underlying trading strategy rather than a contractual borrower interest rate.
What makes HLP different?
HLP is protocol-operated and has historically performed functions connected to Hyperliquid's market structure, including market-making and handling positions created through the liquidation system. Its exact positions change continuously with market conditions.
Users should therefore avoid treating historical HLP returns as a fixed yield. The vault takes market risk and can experience drawdowns.
User vaults add another layer of risk
HyperCore also allows users to create vaults managed by a designated leader. The leader controls trading strategy while depositors provide capital under the vault rules.
A depositor should assess:
- strategy concentration and leverage
- leader incentives and track record
- liquidity and withdrawal rules
- drawdowns, not only headline returns
- whether returns came from repeatable execution or one market regime
Why vaults matter
Vaults turn trading expertise into a pooled onchain product. They also show how HyperCore can support more than direct self-trading: the same native market infrastructure can be used by strategies that aggregate third-party capital.
For professional users, the right comparison is closer to managed trading capital than a savings product. Returns come with execution, market, leverage, and strategy risk.