Mapping the Research Landscape of Exchange Traded Funds using Machine Learning: A Topic Modelling Approach

This study conducts a comprehensive survey of 964 journal articles indexed in Scopus database between 2003 and 2023, solely focused on Exchange Traded Funds (ETFs), employing bibliometric and structural topic modelling approaches. The application of structural topic modelling uncovers five prominent thematic clusters within the dataset: ETF as an investment vehicle, ETF as a trading opportunity, Volatility spillover with other assets, Liquidity, and ETF returns forecasting. The research provides a brief bibliometric assessment, offering insights into publication trends, leading countries, and prominent journals. This information is invaluable for researchers aiming to contribute to esteemed journals in this domain. The USA contributes 249 articles. The "Journal of Portfolio Management" stands out with 31 articles, reflecting its prominence. The study suggests future research avenues, including global influences, market volatility, behavioural insights, sustainable ETFs, and techno- logical integration. The findings have significant implications for academics, industry practitioners, regulators, and policymakers.