Market impact vs order book liquidity

We explore slippage costs for larger notional order sizes focusing on the four most liquid and regulated exchange venues: Coinbase, Kraken, FTX US, and Coinbase Prime.

May 27, 2022

ocean water

Written by

  • Kilian Mie, William Grant & David Duong

Introduction

As institutional activity in digital assets increases, we revisit our previous work (Crypto Liquidity: What trading speed is right for you?) to explore slippage costs for larger notional order sizes (in USD terms). In this study, we focus on the four most liquid and regulated exchange venues: Coinbase, Kraken, FTX US, and Coinbase Prime (the aggregate of Coinbase + LMAX + Bitstamp).

Key observations

  • The flatter the slippage curve, the more liquid the venue
  • Coinbase Prime’s access to smart routing across multiple venues results in significantly lower slippage for all order sizes studied 
    • About 10 bps improvement vs Coinbase at a US$2M notional, and
    • About 30 bps improvement at a $5M notional
    • This analysis does not include Coinbase Prime's access to non-public liquidity providers
  • Kraken also offers a competitive slippage to order size ratio, which is only ~2.5 to 3.5bps above Coinbase Prime’s level
    • Note that Coinbase Prime is in the process of onboarding access to Kraken via the Prime smart order router as well
  • BTC/USD is more liquid than ETH/USD and thus has correspondingly less slippage across all order sizes, which is in line with our expectations
  • Below a notional size of $5M, the level of slippage across exchanges seems relatively balanced regardless of the quote on the bid or ask side.
    • However, the slippage curve for Coinbase on BTC/USD is much steeper on the bid than the ask side (by roughly 18bps) once the notional grows above $5M.
    • Meanwhile, for FTX US, slippage on ETH/USD is actually much lower on the bid than the ask side for notional sizes above $5M
    • These imbalances were affected by the recent market volatility and were not previously visible in the data of earlier periods

Methodology 

  • We source order book snapshots from two sources: Nomics and Cryptocompare
  • For a given period, we measure the slippage of walking up the order book from the mid-quote up to the given USD notional
  • We then compute the average slippage for each notional from April 15 to May 15, 2022, including the corresponding 25% to 75% percentile range

BTC/USD: Slippage (bps) vs Order Size (US$M) for April 15 to May 15, 2022

Filled point represents the mean slippage over time. Shaded region represents the interquartile range.

default image
default image

ETH/USD: Slippage (bps) vs Order Size (US$M) for April 15 to May 15, 2022

Filled point represents the mean slippage over time. Shaded region represents the interquartile range.

default image
default image
newsletter.png

Sign up for our insights

Get the latest market insights, developments and updates, direct to your inbox.