Paper: arXiv 2408.10016
Authors: Sid Bhatia, Sidharth Peri, Sam Friedman, Michelle Malen
Abstract
This research presents a comprehensive framework for analyzing liquidity in financial markets, particularly in the context of high-frequency trading. By leveraging advanced machine learning classification techniques, including Logistic Regression, Support Vector Machine, and Random Forest, the study aims to predict minute-level price movements using an extensive set of liquidity metrics derived from the Trade and Quote (TAQ) data. The findings reveal that employing a broad spectrum of liquidity measures yields higher predictive accuracy compared to models utilizing a reduced subset of features. Key liquidity metrics, such as Liquidity Ratio, Flow Ratio, and Turnover, consistently emerged as significant predictors across all models, with the Random Forest algorithm demonstrating superior accuracy. This study not only underscores the critical role of liquidity in market stability and transaction costs but also highlights the complexities involved in short-interval market predictions. The research suggests that a comprehensive set of liquidity measures is essential for accurate prediction, and proposes future work to validate these findings across different stock datasets to assess their generalizability.
Complexity vs Empirical Score
- Math Complexity: 3.0/10
- Empirical Rigor: 7.5/10
- Quadrant: Street Traders — empirical and applied, lower math complexity
Why this score: The paper relies on standard machine learning classification and statistical feature analysis with minimal advanced mathematical derivations, but it is heavily empirical, utilizing extensive TAQ data, minute-level sampling, and cross-validation to backtest predictive models for price movement.
Research Flowchart
flowchart TD
A["Research Goal<br>Predict minute-level price movements<br>using liquidity metrics"] --> B["Data Input<br>Trade and Quote TAQ Data"]
B --> C["Methodology Feature Engineering<br>Derive liquidity metrics<br>Liquidity Ratio, Flow Ratio, Turnover"]
C --> D["Computational Process<br>ML Classification Models<br>Logistic Regression, SVM, Random Forest"]
D --> E{"Model Training &<br>Evaluation"}
E --> F["Key Finding 1<br>Full feature set yields<br>higher predictive accuracy"]
E --> G["Key Finding 2<br>Random Forest demonstrates<br>superior accuracy"]
E --> H["Key Finding 3<br>Liquidity metrics are<br>essential for prediction"]
Evidence Card
What the authors claim, on what data, and who has checked it. Read from the paper's own text by a language model: every quote is verbatim, nothing was re-run. How to read this card
Proposes
The paper proposes a comprehensive framework for analyzing liquidity in high‑frequency trading markets using machine‑learning classification models.
Source sentence
This research presents a comprehensive framework for analyzing liquidity in financial markets, particularly in the context of high-frequency trading.
Assumes 6
- Liquidity metrics are derived from TAQ data and are suitable predictors for price movements. data
Source sentence
The core aim of this project is to utilize TAQ data to compute liquidity measures that serve as predictors in a model designed to forecast price movements.
- Price movement direction can be modeled as a binary classification (Up or Down). model
Source sentence
The predictive model is approached as a classification problem where the main dependent variable is the price movement direction, categorized as Up or Down.
- Data are sampled only between 11:00 AM and 4:00 PM and aggregated to 1‑minute intervals. data
Source sentence
Data sampling includes daily subsets from 11:00 AM to 4:00 PM. This subset is further refined to 1‑minute frequency data
- The dataset is split 70% training, 15% validation, 15% testing. model
Source sentence
Data will be split in a 70%-15%-15% manner across training, validation, and testing sets respectively.
Show 2 more
- All available liquidity variables are initially used for model fitting before subset selection. model
Source sentence
Initially, all models are fitted with all available independent variables.
- High‑frequency data are taken from the Refinitiv Tick History Dataset. data
Source sentence
This research agenda involves using limit order book datasets to develop predictive models for price movements, utilizing high-frequency data from the Refinitiv Tick History Dataset.
Data
US equities (TAQ data) · equities · intraday
- Trade and Quote (TAQ) data market-prices unknown
Source sentence
The core aim of this project is to utilize TAQ data to compute liquidity measures that serve as predictors in a model designed to forecast price movements.
- Refinitiv Tick History Dataset order-book commercial
Source sentence
This research agenda involves using limit order book datasets to develop predictive models for price movements, utilizing high-frequency data from the Refinitiv Tick History Dataset.
Costs
Not applicable
Finds 3
- Random Forest achieved the highest predictive accuracy among the three classifiers. performance
Source sentence
Notably, the Random Forest algorithm outperformed others, delivering the highest accuracy in our predictions.
- Using all liquidity metrics yields higher accuracy than models with a reduced subset of features. predictive-accuracy
Source sentence
Key insights from these results include the notably high accuracy of the baseline models compared to the tuned models that utilized a subset of liquidity metrics.
- Liquidity Ratio, Flow Ratio, and Turnover consistently rank as the most important features across all models. statistical
Source sentence
Flow Ratio, and Turnover consistently emerge as significant across all models.
Checked by others
No other paper in the archive has been found to support or contradict these findings (compared with 24 related papers). Coverage is partial, so this is neither replication nor refutation.
Code
No repository by the authors found in the paper.
Extracted 2026-11-14 · same card as JSON · A language model chose these items: check the source sentence before relying on one.
Cited References (6)
- A Study on Brexit: Correlations and Tail Events Distribution of Liquidity Measures (2018)
- Detection of Rare Events in Multidimensional Financial Datasets with Zonoid Depth Functions (2017)
- Liquidity Risk and Asset Movement Evidence from Brexit (2017)
- Cluster Analysis of Liquidity Measures in A Stock Market Using High-Frequency Data (2017)
- Rare Events Analysis Using Multidimensional Liquidity Measures in Financial Markets (2021)
- An Event Study of Brexit on Distribution Characteristics of Liquidity Measures (2017)
0 of 6 cited works are in the corpus; resolved via arXiv source bundles, Semantic Scholar and OpenAlex. Corpus-wide: 2596 linked papers.