MANILA, Philippines – Researchers from Ateneo de Manila University have unveiled groundbreaking artificial intelligence (AI) models capable of forecasting market interest rates, a critical factor in financial and policy decision-making. This innovation could offer strategic advantages for governments and businesses in a data-driven economy.
The team tested two deep learning methods—Multi-layer Perceptrons (MLP) and Vanilla Generative Adversarial Networks (VGAN)—which successfully predicted shifts in the Philippine Benchmark Valuation (BVAL) rates. These models processed extensive datasets incorporating 16 economic indicators such as inflation, exchange rates, and credit default swaps to deliver accurate short- and long-term forecasts.
“Interest rates are among the most important macroeconomic factors considered by both government and private entities when making investment and policy decisions,” the researchers explained.
How AI Tools Work
MLP, a type of neural network, identifies patterns in data using a structured sequence of data-processing cells. It performed well in straightforward analyses with smaller datasets. In contrast, VGAN uses dual networks that refine each other’s output to produce highly accurate results, particularly for complex datasets.
Real-World Applications
Financial institutions could leverage these tools for risk management, while governments may optimize debt strategies and minimize borrowing costs. “These AI models present a robust capability to foresee market disruptions,” the team added.
Published on November 15 in AIP Conference Proceedings, the study underscores the growing role of AI in financial strategy and emphasizes the need for advanced neural network designs for enhanced forecasting accuracy.
With such innovative tools, Ateneo’s researchers hope to propel financial systems toward more precise and efficient economic planning in the face of ever-changing market conditions.
Source: AIP Conference Proceedings via Ateneo de Manila University
FAQs
What is the purpose of Ateneo’s AI models?
The AI tools are designed to forecast market interest rates to help governments and businesses make informed financial decisions.
Which AI models were tested in the study?
The researchers tested Multi-layer Perceptrons (MLP) and Vanilla Generative Adversarial Networks (VGAN).
What are the benefits of these tools?
They allow better management of risks and help optimize debt issuance strategies, potentially reducing borrowing costs.
What indicators are used for forecasting?
The models analyze up to 16 domestic and global economic indicators, including inflation, exchange rates, and credit default swaps.
Where can I access the full study?
The research paper is available in AIP Conference Proceedings, published on November 15, 2024.
