Trade Records and Performance
The Performance page shows exchange fills and published daily reports. Check the network badge, the source, and the covered period before comparing numbers. Testnet balances and results are simulated and do not represent withdrawable mainnet income.
Verify an Execution
A fill is one execution of an order. One order can have several fills, and one transaction can contain several executions. A transaction hash alone does not uniquely identify a fill.
For an individual fill, compare the account, network, market, fill ID, order ID, time, price, quantity, direction, realized P&L, and fee with Hyperliquid. The app preserves individual fill identity when opening execution details.
Valid transaction links open the explorer for the record's network. When a hash is missing, malformed, or all zeros, or the network is unknown, the app does not create a transaction link. A missing link is not proof that a trade failed: some execution types do not have an individual transaction hash.
The explorer may display the submitted order's limit price. That price can differ from the execution price in the fill record. Compare fill price with fill price when checking execution quality.
History Coverage
The exchange-fill view loads up to 2,000 recent fills per request and shows the available date range. Filters and “Load more” apply to that loaded sample. They do not retrieve an unlimited lifetime history.
Hyperliquid's time-based fill endpoint supports pagination, but only the most recent 10,000 fills are available through that endpoint. A complete historical audit therefore needs a durable archive and an explicit coverage check. A current page, cache, or explorer link does not prove that the lifetime ledger is complete.
See the official Hyperliquid Info endpoint reference for current retrieval limits and fill fields.
Understand the P&L Basis
These values answer different questions:
| Value | Meaning |
|---|---|
| Realized P&L before fees | The closed-position P&L reported for an execution |
| Realized P&L less trading fees | Realized P&L minus trading fees in the loaded sample |
| Funding | Separate transfers between perpetual position holders |
| Unrealized P&L | The current marked gain or loss on open positions |
| Change in account equity | Includes trading results, funding, unrealized changes, and external cashflows |
For a consistent period and currency:
Realized net trading P&L = realized execution P&L − trading fees + funding received − funding paid.
To reconcile equity, also include the change in unrealized P&L, deposits, withdrawals, transfers, and any other ledger adjustments. Deposits are capital contributions, not trading earnings. Transfers between the same account's spot and perpetual balances must not be counted twice.
The Performance fill metrics explicitly exclude funding. They are not an all-cost profitability measure. Infrastructure, market data, model usage, and other operating expenses also need to be deducted when assessing the economics of the agent.
Hyperliquid's reported fill fee already includes any builder fee. Do not deduct that component twice. For current fee and funding rules, consult Fees and Funding.
Read the Metrics Correctly
Fill Counts and Closing-Fill Win Rate
“Loaded fills” counts executions, not orders or completed position cycles. A partial close may create several fills.
The closing-fill win rate and profit factor use each closing fill's realized P&L less that fill's own USDC trading fee. They do not allocate earlier entry fees or funding to the close, so they are not the win rate or profit factor of completed positions after all costs. A reversal fill's fee may also cover newly opened quantity. If the fee currency or economics are unavailable, the affected metrics remain unavailable.
A completed-position analysis needs to reconstruct entries, additions, reductions, reversals, and the final close in execution order. Include every entry and exit fee, and allocate funding only when the underlying timestamps support that allocation. An aggregated funding record may not identify an individual intraday position cycle.
Stored Daily Reports
Monthly bars aggregate stored daily reports. Those reports are a separate source and may cover a different period from the available venue-fill sample.
A report's drawdown amount is a decline in its cumulative dollar P&L for that report's period. It must not be interpreted as a percentage decline in total account equity. A valid account drawdown percentage requires a consistent account-equity series with appropriate treatment of cashflows.
Account Return and Sharpe
An account-return Sharpe ratio requires a documented return series, interval, coverage, cashflow treatment, and annualization method. A calculation based on unnormalized daily dollar P&L is not enough to establish that measure. The fill view does not present such a value as a verified account-return Sharpe ratio.
Use metrics with matching periods and definitions. A high win rate, a positive profit factor, or a positive month alone does not demonstrate a durable trading advantage.
Reconciliation and Exceptions
An execution audit should:
- Establish the exact account and network.
- Retrieve and deduplicate all available fills within a fixed cutoff.
- Match individual fill and order identities without substituting another fill that shares a transaction hash.
- Compare quantities, execution prices, realized P&L, and fees.
- Reconcile funding and non-trading cashflows to balances and open positions.
- Record missing history, ambiguous matches, and unexplained residuals explicitly.
Historical application records can contain estimates, partial information, or records from another account period. They should not override exchange execution data. Repairs to immutable audit records or trading reconciliation need the project's review process; an unexplained difference should remain visible until resolved.
Profit Targets
A daily profit objective is a planning target, not a guaranteed outcome. Its feasibility depends on capital, net expectancy after all costs, turnover, market capacity, drawdowns, and whether results survive independent validation.
Testnet results can help validate execution and accounting. They do not establish that the same strategy will earn the same returns with mainnet liquidity, costs, and capital.
Next: Integrate programmatically with the SDK.