Forecasting Financial Crisis using Topological Data Analysis Approach

Abstract

Traditional financial forecasting methods often struggle to capture the complex interactions and emerging patterns that precede financial crises. By leveraging on TDA, this research aims to uncover potential topological features that might serve as early warning signals for impending financial crises. The study adopts the utilization of Topological Data Analysis, an initiative mathematical framework to explore and analyze the inherent topological structures within financial data set, using secondary data from ”Yahoo Finance API”. The results of the analysis conducted using Python indicate that persistence homology in TDA successfully identifies key topology features associated with financial crises, implying its potential for developing early warning systems in the financial sector. The insights gained from this analysis could significantly enhance the early detection and proactive management of system risks in financial market, thereby contributing to more robust risk assessment and policy formulation strategies.