Slippage in CopyTrading: What 240 Order Book Snapshots Revealed

Subscribers repeat the same trade but may get a different price: the total volume passes through several order book levels, and the average price shifts. This is how slippage occurs in copy trading.

We tested this effect on 240 snapshots of public BTC order books. This calculation is based on visible liquidity, without placing actual orders.

How we measured slippage by order book depth

On August 21, 2026, we collected 60 order book snapshots each from BTC perpetual futures on Binance, Bybit, OKX, and Hyperliquid. The entire series took 10 minutes and 22 seconds.

For each snapshot, we calculated the volume-weighted average price (VWAP) for buying and selling amounts of $1,000, $10,000, $100,000, and $500,000, and compared it with the mid-spread price. In total, there were 1,920 calculations.

How order size affects slippage

For each size, we separately took the highest median and p95 values among the four platforms and two trade directions. The median shows the middle of the distribution. p95 is the boundary below which 95% of observations remain. Numbers in the same row may relate to different platforms or directions.

1 bp = 0.01%.

Up to $100,000 the median changed little. But the upper part of the distribution increased earlier: already at $10,000, the high p95 noticeably distanced from the median.

The highest p95 in the table is 2.110 bp for a model buy order of $500,000. This is about $105.52 difference between the mid-spread price and the calculated average order price. This is not a fee or a forecast of client slippage.

Why the model does not predict partial order execution

All 1,920 calculations found sufficient visible depth, but the actual order may still be incompletely filled. The public order book does not show the execution queue, hidden liquidity, delay between signal and order, fees, funding, rejections, or specific account settings.

The platforms were polled sequentially, and their APIs return varying depth. Therefore, the table is not suitable for ranking exchanges and does not indicate CopyTrader execution quality.

We have saved 240 calculated observations, the summary table, the script, and control hashes. Aggregates can be verified again. The original arrays of price levels in this series are not saved, so the exact historical passage through each order book cannot be reproduced.

Official documentation: Binance, Bybit, OKX and Hyperliquid.

The study was prepared by the EasyTrading team for CopyTrader. Results apply only to the series collected on August 21, 2026, from 13:19:03 to 13:29:24 UTC.

Calculations and the four rows of the final table were cross-checked again with the saved summary on August 22, 2026.

The material is for informational purposes and is not investment advice. Derivative trading and copy trading carry risk of loss. Calculation from order book snapshots does not assess overall risk or strategy profitability nor guarantee actual execution price.

How to compare execution of leader’s and subscriber’s orders

In the analysis of leader strategy copyability, we explained why its trade history does not show how much liquidity remains when the same signal is repeated simultaneously by subscribers.

Before a trade, it is useful to see available depth and permissible deviation limit. After a trade, the actual average price, executed volume, and final position.

The larger the total subscriber order, the more important it is to compare their actual execution with the leader’s execution, rather than judge copyability solely by the leader’s trade history.