Mathematicians from the University of the Philippines have created a novel method for stock traders to examine stock market movements and potential crashes.
By using Topological Data Analysis, which applies concepts from topology to reveal patterns in large datasets, Ela Mae Riñon and Dr. Rachelle Sambayan from the UP Diliman College of Science Institute of Mathematics were able to expect impending market downturns. They’ve published a study on this new data analysis method in the Philippine Journal of Science. Stock data of three Philippine companies—Cebu Air, Century Pacific Food, and PAL Holdings—were included in the analysis from 2019 to 2021.
The authors likened how TDA works to observing patterns in constellations in the night sky, which seems to represent a random arrangement of stars spread across the universe but actually reveals patterns upon extended observation. Similarly, TDA can recognize geometric structures within datasets.
- 6 beginner-friendly online stockbrokers to help you start investing in the Philippines
- How to invest and earn in the Philippine stock market online
One of its key components is persistent homology, which the authors used to observe data points forming specific shapes and how these shapes change as the data is analyzed. They found that the persistence of certain geometric forms decreased as stock market prices and other conditions worsened, thereby indicating heightened risk.
Using TDA, the authors of the study examined Cebu Air’s stock performance during significant downturns brought upon by the COVID-19 pandemic. They noticed a clustering of data points prior to market declines. Meanwhile, PAL and CNPF remained stable as the model predicted during the same period, with only minor dips but no significant crashes.