Cryptocurrency High-Frequency Liquidity Strategy based on Orderbook Behavior

Author

Lynn Chu

Keywords:

Cryptocurrency, High-Frequency Trading, Order Book Analysis, Machine Learning

Abstract

This paper presents a high-frequency trading strategy for cryptocurrency markets based on order book dynamics. We develop and evaluate multiple machine learning models to predict price movements using order book data from the Binance cryptocurrency exchange. Our approach incorporates both traditional statistical features and deep learning methods to capture complex patterns in market microstructure. The results demonstrate that our LSTM-based model achieves superior performance in predicting short-term price movements compared to traditional machine learning approaches, with potential applications in algorithmic trading strategies.

Data Source

Data is collected using the Binance Exchange Public OrderBook API. For more details, visit Binance Order Book API.

Shared Overleaf

Access the Overleaf main_07.tex project here