Reality
Perfect one-to-one replication does not exist. A mirrored order is a new order, placed after the leader's fill is already public on-chain, into a book that has already moved. Divergence is structural, and the useful question is whether it is bounded and attributable.
In short
Copy trades diverge from the leader for six structural reasons: latency between their fill and your order, order-book depth and slippage, partial fills, proportional sizing with lot rounding, different margin state and leverage caps, and funding accruing to whoever holds the position at the stamp. Divergence can be positive or negative for the follower on any given trade, and it is larger for high-turnover leaders.
| Driver | Effect on your result | What bounds it |
|---|---|---|
| Latency | Worse or better entry than the leader's fill | Nothing structurally — reduced, never removed |
| Book depth / slippage | Different average price | Market choice and order size |
| Partial fills | Smaller position than the target | Retry and reconciliation logic |
| Proportional sizing + rounding | Ratio to the leader is never exact | Larger sleeve size relative to lot size |
| Margin state / caps | Deliberately lower exposure than the leader | Per-leader ceilings and leverage caps |
| Funding timing | Different funding paid or received | Shorter holding periods; nothing else |
The leader's fill becomes visible on-chain, then it has to be observed, evaluated, sized, signed and submitted. Every one of those steps takes time, and the book does not wait. By the time your order arrives, the price that triggered it is history.
This is inherent to mirroring rather than a defect of a particular implementation. Any system that copies a public action acts after it.
The leader consumed the top of the book. Your order fills into what remains, so your average price differs even when the intent is identical. In thin markets or during fast moves that gap widens, and a large order can walk several levels deeper.
Partial fills add a second layer: an order may fill in pieces at different prices, or not complete at all if liquidity disappears, leaving your sleeve with a different position size than the proportional target implied.
Mirrored size is derived from the capital allocated to that leader and their weight in the basket — Capital weighted in proportion to each leader's composite score — not copied notionally from the leader's ticket. Rounding to the market's lot and tick size shifts it again, and the effect is proportionally larger on small sleeves.
Per-leader notional ceilings and independent leverage caps deliberately break the ratio further. They exist to bound exposure, and bounding exposure necessarily means not replicating a leader who is running hotter than your caps allow.
The leader's margin state is unknown to you and unlike yours. What they can hold, add to or survive is a function of their collateral, not yours, so identical intent produces different capacity.
Funding accrues to whoever holds the position at the funding stamp. Entering minutes later can change which payments you receive or pay, and for a position held across many stamps that difference compounds independently of price.
Seconds matter more to a leader who scalps than to one who holds for days. High-turnover strategies therefore contribute more tracking error per unit of capital, and leverage amplifies a small entry-price gap into a large PnL gap.
The corollary is that slower leaders copy better. This is a property of the strategy being mirrored, not of the follower.
Errors across independent leaders are partly uncorrelated, so in a weighted basket they average out rather than compounding into one number. Because each leader runs in one Hyperliquid sub-account per mirrored leader, the error is measurable per leader instead of hidden inside one netted position.
State the honest conclusion plainly: tracking error can be negative for the follower over a period. Direction and structure track closely; entry and exit prices do not. The design goal is bounded and attributable divergence, not zero divergence.
Past performance is not indicative of future results. Perpetual futures are leveraged instruments and carry a substantial risk of loss, including the loss of your entire position.
Methodology
Scoring and replacement are documented in full on How it works and in the Docs (Policy v3). In short: the Elite basket is sticky, emergencies remove a leader immediately, and soft issues accrue at most one strike per UTC day with three strike-days triggering replacement. Read how it works or the documentation for the full table.
Questions
Yes, on any given trade. Divergence is two-sided: entering after the leader can produce a better or worse average price. Over many trades it is not a source of edge in either direction.
Slower, lower-turnover leaders in deep markets. Short-hold, high-frequency strategies diverge most because seconds are material to them.
No. It is execution divergence, separate from the builder fee and from funding. It can be positive or negative for you.
It does not reduce the per-leader error, but it makes it attributable per leader and prevents one leader's fill quality from determining the whole book's outcome.
Score-weighted allocation across up to 10 elite Hyperliquid traders, each isolated in its own sub-account. Your funds never leave your account.
Non-custodial · Agent cannot withdraw · Cancel delegation anytime
