Reproduce “Protecting Retail Investors from Order Book Spoofing using a GRU-based Detection Model” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Jean-Noël Tuccella, Philip Nadler, Ovidiu Şerban
Researchers from Imperial College London developed a Gated Recurrent Unit (GRU) based model to detect order book spoofing in cryptocurrency markets, providing an early warning up to two seconds before a manipulative event concludes. The system, trained on granular Level 2 order book data, aims to protect retail investors in unregulated environments.
Key insights
- Early detection of market manipulation, specifically spoofing, is achievable in cryptocurrency markets using granular Level 2 order book data and advanced deep learning techniques.
- GRUs are effective in capturing the complex temporal dependencies required to identify dynamic order book manipulations in real-time.
- The 2-second lead time provided by the model offers a practical window for retail investors to adjust trading strategies and mitigate potential losses from manipulative actions.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.