Bibliometric Analysis of Studies on Shrinkage Techniques and Covariance Matrix Estimation in Optimal Portfolio Selection

Abstract

This study employs bibliometric analysis to explore heavily researched areas (covariance matrix estimation) and emerging fields (shrinkage techniques) within the domain of portfolio optimization. As the volume of financial data grows, traditional sample covariance matrix estimation becomes unstable, often lacking the high precision required to support effective portfolio selection and management. Shrinkage estimators offer an alternative by reducing data dimensionality, thereby improving the accuracy of covariance matrix estimation. Through a comprehensive review of scholarly publications across Web Of Science (WOS) database, this analysis highlights key trends, thematic developments, and core research clusters in this field. Our findings reveal substantial growth in research activity, particularly in methodologies that integrate shrinkage techniques with modern statistical learning theories. The study not only maps the field's structure but also identifies influential authors, institutions, and countries at the forefront of this research. This bibliometric analysis underscores current research trends in covariance matrix estimation and shrinkage techniques within the context of portfolio optimization.