Copy trading: research before automation
Copy trading uses a service to reproduce some activity from another account. The service’s rules determine what it copies and how it handles entries, exits, partial fills, and failed orders. A public wallet history does not prove a service can copy those actions or reproduce the same result.
This guide is a due-diligence checklist. It does not identify profitable traders or recommend an allocation.
Separate the record from the claim
First identify where the trader operates. HyperCore and HyperEVM have different interfaces and histories. A claim about activity on one does not verify activity on the other.
For HyperCore, Hyperliquid’s Info API documentation describes account fills and portfolio-history queries. Use the actual account or subaccount address. Hyperliquid explains that an API wallet is a signer, and that account data must be queried with the account address rather than the signer address in its API-wallet documentation . The time-based fills endpoint returns at most 2,000 fills per response and makes only the 10,000 most recent fills available. A partial history cannot establish a complete track record.
For HyperEVM activity, inspect the address and transactions on the correct EVM network. Hyperliquid’s HyperEVM guide describes its JSON-RPC interface. Token transfers and swaps alone do not show a complete account return: transfers between wallets, assets held elsewhere, open positions, gas, and token valuation can all affect the picture.
Evaluate one account at a time
- Confirm the address and account boundary. Record the network, wallet address, and any linked account or subaccount. Ask the trader to explain any addresses included in a published performance claim.
- Choose a fixed observation window. Save the start and end dates and the raw data you used. Do not accept a selected screenshot or a changing leaderboard as a full history.
- Reconcile the balance changes. Account for deposits, withdrawals, transfers, open positions, and assets that have not been sold. Separate realized results from estimated value.
- Include costs. Account for venue fees, provider or builder fees, funding, gas, spread, and the price impact of trades. A gross gain is not a net return.
- Look at losses and concentration. Check whether the result depends on one token, one large trade, leverage, or an unclosed position. Ask how losses, liquidation, and pauses are handled.
- Compare the provider’s rules with the activity. Confirm the supported market, copied actions, sizing method, order limits, exclusions, failure handling, and stop process in current provider documentation.
If the available records do not answer these questions, mark the claim unresolved. Do not substitute social proof, labels, or a high win rate for account-level evidence.
Why a copied result can differ
The follower uses a separate account and receives its own execution. Entry price, trade size, available liquidity, order timing, rejected orders, and fees can differ from the source account. Automation does not remove these differences or make an observed result repeatable.
Before connecting a service, review what it can sign and where key material is stored. Use the bot custody checklist to structure that review. Set your own stop conditions before the first copied order; do not increase exposure to recover a loss.
Further reading
- Find wallets to copy — verify account history and its limits.
- Understanding ROI and win-rate metrics — compare metric definitions and reporting windows.
- Choosing a trading bot — compare documented permissions, fees, and controls.
This material is educational, not financial advice. Automated trading and token trading can lose all funds committed.