Copying Across Prop Accounts: The Mechanics Nobody Explains

A copier sounds simple: the leader trades, the followers trade the same thing. In practice the followers trade *almost* the same thing, and the gap between "same" and "almost" is where a farm's risk actually lives. We built our own copier for our futures desk, and these are the mechanics that mattered.

What a copier does, literally

It watches the leader's account for order events — new order, fill, modification, cancel, close — and places corresponding orders on each follower, scaled by a multiplier. That sentence hides three separate systems: detection (how fast does it see the leader's event?), placement (how fast does it act on each follower?), and reconciliation (how does it know the follower's position still matches?).

Where fills diverge

The leader fills at one price. Each follower's order arrives a fraction of a second later and fills at whatever the market offers then. In a fast market on a thin contract, that is a different price on every seat. Over many trades the difference averages out in *direction* but not in *risk*: a follower that fills worse has less room to its stop and to its trailing floor than the leader did. The most fragile follower — the one nearest its floor — is also the one most hurt by a worse fill.

Partial closes are where copiers break

Closing half a position on the leader is easy to say and hard to mirror. Followers with different sizes cannot always close "half" in whole contracts; followers that filled late may have a different open size; a follower whose earlier order was rejected has nothing to close. Our own build got partial-close mirroring wrong before it got it right, and every commercial copier we have looked at documents the same class of edge case. The operator's answer is to design the process so that the leader closes in whole units the followers can replicate, and to reconcile positions after every event rather than trusting the last one.

Rejections and the naked follower

Sometimes a follower's order is refused — margin, a rule, a rate limit at the firm. The leader is in a trade; that follower is not. If the copier does not notice, the follower's *exit* will later fire against no position, or worse, open one. Reconciliation — checking every follower's actual position against the leader's after each event — is not optional. A copier that cannot answer "which followers are actually in this trade right now?" is a liability.

Latency is a rule-risk, not just a fill-risk

A news blackout window, a daily close, a payout freeze — these are enforced by the clock on each seat. A copier that places a follower's order two seconds after the leader's can place it *inside* a window the leader's was outside. The copier's latency has to be inside the operator's calendar margins. News blackouts are the sharpest version.

What this means for the two trees

Everything above applies inside a tree. Across the two trees — evaluations and funded accounts — nothing is copied at all, which is the point. The copier's complexity is contained inside groups that share an objective; nothing mechanical ever crosses between objectives.

Kingdom Portfolios is an independent education company. We're not affiliated with, endorsed by, or sponsored by any prop firm, broker, or platform named here, and we don't use affiliate links. Nothing here is investment advice or a recommendation to join any firm or trade any product. Funded-account evaluations cost real money and most participants never pass or get paid — learn first, and trade your own risk. Rules and fees change often; verify current details on each company's own site. Education only.