Mapping the Intellectual Structure of REIT Research Using Structural Topic Modelling: A Machine Learning Approach

Real Estate Investment Trusts (REITs) are crucial for democratising real estate investment, enabling individuals to participate in property markets. REITs also offer diversification, liquidity, and income potential to investors. Scholarly and practitioner interest in REITs has surged over the last two decades, resulting in a proliferation of research. Nevertheless, a comprehensive literature review delineating thematic clusters in this domain remains lacking. Hence, this study employs the unsupervised machine learning technique known as structural topic modelling which does text mining to unveil latent themes within the REIT literature. This study utilises structural topic modelling along with bibliometric analysis to investigate 1161 articles spanning 2002-2022 from the Scopus database, focusing on REITs. The leading "Journal of Real Estate Finance and Economics" contributes 183 articles (15.8% of the total). Structural topic modelling identifies five key thematic clusters: i) Economic impact of REITs, ii) REIT market pricing strategies, iii) Investment opportunities via REITs, iv) Analysing REITs’ capital market performance, and v) Portfolio diversification with REIT investments. These results shed light on the evolving research trends in the REIT domain. Future studies may explore how economic, environmental, and technological trends impact REIT performance and investor perspectives. These findings hold substantial importance for a diverse audience, encompassing researchers, financial institutions, and professionals.