Predicting stock returns of banking sector using machine learning models: Evidence from South Asian Economies

Abstract

Banks are very important financial intermediaries in the financial system of a country. The objective of the study is to predict the stock market returns of the banking sector in South Asia. Banks stock return movements may have significant practical ramifications for a variety of decision makers, including current shareholders, potential investors, peer bank managers, and credit rating agencies. This study integrates textual data with financial information from leading news articles to predict stock returns in the banking sector of South Asian Economies with daily observations of 231,458 covering period from January 05, 2010 to January 12, 2024 by applying neural network technique. The study's findings showed that the news articles are important predictor of stock returns of South Asian banking sector. The findings also revealed that the information in news articles in news papers provide important information that may be used for stock returns. Therefore, investors, policymakers and researchers can leverage textual information in business news articles with financial information for better and well-informed decision-making. Investors can concentrate on potential investment opportunities, avoiding market risks and establishing portfolios in South Asia as bank stock is generally considered as an index of future economic expansion.