What is the proportion of aggressive orders vs passive orders executed by different types of traders?

Question

It's clear that each aggressive order (or market order or limit crossing BBO) is matched against the same volume of resting limit order(s).

I'm interested in statistics per different types of traders, large or small. My hypothesis is that large traders such as market makers or execution algos execute more than 50% of their volume as resting orders, probably around 70-80%, but I could not find relevant studies. Will appreciate references to such studies. Alternatively, it can be statistics per account size of traders or anything relevant. I'm most interested in futures markets such as CME and Eurex, US stocks markets, and cryptocurrencies.

My main motivation to to ask this question is to evaluate the 'importance' of market depth (order book) data compared to trades data (or times and sales). It seems that trades data is just 3-10% of all market data updates, but it's not a reliable way to answer my (probably vague) question because resting orders can be cancelled or moved at any time.

Answer 1

There is a partial answer to this question in Megarbane, Nicolas, Pamela Saliba, C-A L., and Mathieu Rosenbaum. "The behavior of high-frequency traders under different market stress scenarios" Market Microstructure and Liquidity 3, no. 03n04 (2017).

In the paper, the dataset is made of the 36 most liquid French stocks, from November 2015 to July 2016.

Let me focus on Table 2, that makes the difference between HFT and non-HFT market participants:

Aggressive	Passive	pct
HFT	HFT	33.6%
non-HFT	HFT	22.4%
HFT	non-HFT	31.2%
non-HFT	non-HFT	12.8 %

Another table in the paper shows HFT are polarized: they are not all of them operating the same way, some are mainly removing liquidity and overs providing liquidity.

Another partial answer is that there is a correspondence between liquidity consumption and liquidity provision. Simply because one can only consume the liquidity that is there, and also because liquidity often replenishes the orderbooks. Of course it is not replenished systematically since in around 10% of the case, the price move consecutive to liquidity consumption is permanent, probably because there was information driving the trades.