EWMA Control Charts for Monitoring Optimal Portfolio Weights

Abstract

The distribution of the asset returns is subject to structural breaks. These breaks may lead to changes in the optimal portfolio weights. A fast detection of any change in the optimal portfolio weights is of great relevance for a portfolio investor.

In this paper we introduce several control charts for the weights of the global minimum variance portfolio. All charts are based on exponential smoothing. The charts are compared with each other using the average and median run length performance criteria. The proposed methodology is applied to empirical data.

Keywords:

• Changes in the covariance matrix
• Global minimum variance portfolio
• Multivariate normal distribution
• Statistical process control

Subject Classifications:

• 62L10
• 91B28
• 62P05

ACKNOWLEDGMENTS

The authors would like to thank the referees for their thoughtful and constructive suggestions, which led to a considerable improvement of the paper. Vasyl Golosnoy expresses his gratitude for financial support from the Stiftungsfond Deutsche Bank.

Notes

Recommended by N. Mukhopadhyay