# Quasi-Centralized Limit Order Books

**Authors:** Martin D. Gould, Mason A. Porter, Sam D. Howison  
**arXiv:** [1502.00680](https://arxiv.org/abs/1502.00680)  
**Date:** 2015-02-02  
**Summary page:** https://www.summarizepaper.com/en/paper/1502.00680v2/

## Key points

- Quasi-centralized limit order books (QCLOBs) allow financial institutions to access trading opportunities only with counterparties they have sufficient bilateral credit with.
- The authors conducted an empirical analysis using a high-quality data set from a large electronic trading platform that utilizes QCLOBs.
- Significant differences were found between the results of this study and those widely reported for other types of limit order books (LOBs).
- A simple linear rescaling can cause distributions describing order flow and market state to collapse onto a single curve, despite varying greatly across different trading days.
- The authors propose a semi-parametric model to describe order flow and market state in a QCLOB on a single trading day, which demonstrates similar performance to parametric curve-fitting techniques but is simpler to compute and faster to implement.
- This research sheds light on the unique characteristics of QCLOBs compared to other LOBs, contributing to our understanding of market dynamics and offering potential improvements in modeling approaches for QCLOBs in electronic trading platforms.

## Summary

In their paper titled "Quasi-Centralized Limit Order Books," authors Martin D. Gould, Mason A. Porter, and Sam D. Howison explore the concept of a quasi-centralized limit order book (QCLOB). This type of limit order book (LOB) allows financial institutions to access trading opportunities only with counterparties they have sufficient bilateral credit with. The authors conduct an empirical analysis using a high-quality data set from a large electronic trading platform that utilizes QCLOBs. The study reveals significant differences between their results and those widely reported for other LOBs. Additionally, the authors make an interesting discovery regarding empirical universality by finding that a simple linear rescaling can cause distributions describing order flow and market state to collapse onto a single curve despite varying greatly across different trading days. Motivated by this finding, the authors propose a semi-parametric model to describe order flow and market state in a QCLOB on a single trading day. Their model demonstrates similar performance to parametric curve-fitting techniques but is simpler to compute and faster to implement. Overall, this research sheds light on the unique characteristics of QCLOBs and provides insights into their behavior compared to other types of LOBs, contributing to our understanding of market dynamics and offering potential improvements in modeling approaches for QCLOBs in electronic trading platforms.