Covariance Matrix Estimation for Portfolio Selection: Linear and Non-Linear Shrinkage Methods on the Frontier Market

The development of a stable, secure, and efficient stock market in Vietnam plays a crucial role in enhancing risk resilience, ensuring a well-structured market, and serving as a platform to attract substantial medium and longterm capital for the economy.Therefore, the study aims to analyze the effectiveness of linear and non-linear shrinkage estimators for portfolio selection in the Vietnamese stock market.Comparing various estimators, including custom-tailored linear Shrinkage, non-linear Shrinkage, and generic linear Shrinkage.The results reveal that the non-linear shrinkage estimator does not always give superior results compared to the generic linear shrinkage estimator, especially when dealing with low numbers of stocks.The custom-tailored linear shrinkage estimator remains a challenging benchmark that the non-linear shrinkage estimator needs to overcome.These findings contribute to further empirical evidence on the effectiveness of shrinkage estimators for portfolio selection in a frontier market.Finally, the author proposes several policy implications for the stock market to promote and create necessary guidance for Vietnamese businesses and economic organizations to apply international financial standards, helping to improve modern management and corporate governance capacity in line with the world's development and progress trends.