Measurement Error in Equity Market Microstructure

2027 · 2027 Competition

School: School of Computer and Information Sciences
Category: ResearchPrimary

Project Overview

One Liner: Measuring how data preprocessing choices bias standard liquidity estimates in US equity markets, benchmarked against ground truth from Nasdaq order-level data.

Abstract

Standard measures of market liquidity depend on knowing which side initiated each trade, but public equity data does not record this, so researchers infer it using classification rules and a set of preprocessing choices that are rarely documented. This project uses Nasdaq ITCH message data, which identifies the true initiator, to measure how much error those choices introduce and how that error propagates into published liquidity estimates, and tests whether the error changed after the 2025-2026 market structure reforms. The deliverable is an open-source pipeline with characterized error alongside the empirical results.

No video available.

Screenshots

0 image(s)

No screenshots uploaded yet.

Team Members

Emirhan Gencer
Lead

Advisors

Sean Grimes
Sean Grimes