Financial Anomaly Detection for the Canadian Market

TL;DR

This paper evaluates topological data analysis, PCA, and neural network methods for detecting financial anomalies in the Canadian market, highlighting the effectiveness of neural networks and TDA.

Contribution

It compares three classes of methods for financial anomaly detection and demonstrates the strong performance of neural networks and TDA in identifying stress events.

Findings

Neural network-based methods outperform PCA in anomaly detection.

TDA methods effectively identify financial stress events.

Neural networks and TDA achieve the strongest detection performance.

Abstract

In this work we evaluate the performance of three classes of methods for detecting financial anomalies: topological data analysis (TDA), principal component analyis (PCA), and Neural Network-based approaches. We apply these methods to the TSX-60 data to identify major financial stress events in the Canadian stock market. We show how neural network-based methods (such as GlocalKD and One-Shot GIN(E)) and TDA methods achieve the strongest performance. The effectiveness of TDA in detecting financial anomalies suggests that global topological properties are meaningful in distinguishing financial stress events.