Comparing Statistical and Machine Learning Outlier Detectors for Bitcoin Order Books
Summary
This study compares thirteen statistical and machine learning models for identifying unusual observations in Bitcoin limit order book data. It evaluates the models in a shared framework intended for real-time anomaly detection, with potential use in examining manipulative trading behavior. The strongest reported model is Empirical Covariance, which the study says outperformed a buy-and-hold benchmark in its backtest.
The evidence comes from 26,204 records from a major exchange. The authors discuss trade-offs among model complexity, trading frequency, and performance, and frame outlier detection as a possible input to trading and risk management. The excerpt gives little detail on data period, transaction costs, benchmark construction, model settings, or validation procedures, so the reported gain should not be assumed to generalize to other venues or market conditions.
Key ideas
- The study evaluates thirteen methods for detecting outliers in Bitcoin limit order books.
- Empirical Covariance is reported as the top-performing model in the backtest.
- The comparison uses 26,204 records from a major exchange.
- Model complexity and trading frequency are presented as relevant performance trade-offs.
- The excerpt does not describe costs or validation details needed to assess live applicability.
Tags
Full text
# A Comparative Analysis of Statistical and Machine Learning Models for Outlier Detection in Bitcoin Limit Order Books # A Comparative Analysis of Statistical and Machine Learning Models for Outlier Detection in Bitcoin Limit Order Books The detection of outliers within cryptocurrency limit order books (LOBs) is of paramount importance for comprehending market dynamics, particularly in highly volatile and nascent regulatory environments. This study conducts a comprehensive comparative analysis of robust statistical methods and advanced machine learning techniques for real-time anomaly identification in cryptocurrency LOBs. Within a unified testing environment, named AITA Order Book Signal (AITA-OBS), we evaluate the efficacy of thirteen diverse models to identify which approaches are most suitable for detecting potentially manipulative trading behaviours. An empirical evaluation, conducted via backtesting on a dataset of 26,204 records from a major exchange, demonstrates that the top-performing model, Empirical Covariance (EC), achieves a 6.70% gain, significantly outperforming a standard Buy-and-Hold benchmark. These findings underscore the effectiveness of outlier-driven strategies and provide insights into the trade-offs between model complexity, trade frequency, and performance. This study contributes to the growing corpus of research on cryptocurrency market microstructure by furnishing a rigorous benchmark of anomaly detection models and highlighting their potential for augmenting algorithmic trading and risk management.
Shown in full with attribution under the source's licence. Licence: abstract CC0
This summary was written by Stratmill's research agent from the original; it is not a copy of the source.