Paper: arXiv 2309.00088

Authors: Timothy DeLise

Abstract

Modern financial electronic exchanges are an exciting and fast-paced marketplace where billions of dollars change hands every day. They are also rife with manipulation and fraud. Detecting such activity is a major undertaking, which has historically been a job reserved exclusively for humans. Recently, more research and resources have been focused on automating these processes via machine learning and artificial intelligence. Fraud detection is overwhelmingly associated with the greater field of anomaly detection, which is usually performed via unsupervised learning techniques because of the lack of labeled data needed for supervised learning. However, a small quantity of labeled data does often exist. This research article aims to evaluate the efficacy of a deep semi-supervised anomaly detection technique, called Deep SAD, for detecting fraud in high-frequency financial data. We use exclusive proprietary limit order book data from the TMX exchange in Montréal, with a small set of true labeled instances of fraud, to evaluate Deep SAD against its unsupervised predecessor. We show that incorporating a small amount of labeled data into an unsupervised anomaly detection framework can greatly improve its accuracy.

Complexity vs Empirical Score

  • Math Complexity: 5.5/10
  • Empirical Rigor: 7.5/10
  • Quadrant: Holy Grail — high math complexity, high empirical rigor

Why this score: The paper presents a solid application of Deep SAD to high-frequency limit order book data for market manipulation detection. Empirical rigor is high due to the use of real-world exchange data, though reproducibility is hindered by the proprietary nature of the TMX dataset.

Research Flowchart

  flowchart TD
    A[Research Goal: Evaluate Deep SAD for Fraud Detection in High-Frequency Futures] --> B[Input Data: TMX Limit Order Book with Small Set of Labeled Fraud]
    B --> C[Methodology: Train Deep SAD Framework]
    C --> D[Process: Combine Unsupervised Deep Learning with Supervised Signal from Labeled Data]
    D --> E[Benchmark: Compare Deep SAD Against Unsupervised Predecessor Deep SVDD]
    E --> F[Outcome: Adding Small Labeled Data Significantly Boosts Fraud Detection Accuracy]