Holding period and turnover

The arithmetic of delay

If a leader's average position is held for three days, a delay of a few seconds at entry is negligible relative to the move being captured. If the average hold is thirty seconds, that same delay is a significant fraction of the trade's life.

The same logic applies at exit. A late exit on a slow position is a rounding error; a late exit on a scalp can convert the trade's result entirely.

• Delay cost is roughly constant; edge per trade is not.
• Fee drag scales linearly with turnover, including the builder fee on mirrored volume.
• High turnover produces more data points, which helps scoring accuracy.
• Low turnover produces cleaner replication but slower evidence.

The turnover trade-off in scoring

High-frequency records are statistically richer: hundreds of closed trades support a far more confident score than a dozen. That statistical advantage is why high-turnover accounts often top raw scoring — and why score alone is not the allocation decision.

Allocation weighs the confidence of the score against the fidelity of replication. A slightly noisier score on a strategy that copies cleanly is worth more than a precise score on one that does not.

What a follower should expect

Tracking error grows with turnover. Over a month, a low-turnover leader's mirrored result should sit close to theirs; a high-turnover leader's can diverge visibly in either direction without anything malfunctioning.

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.