Search Search

Archive of Press Releases Archive of Press Releases


University of East Anglia research reveals new ways to form efficient portfolios

Mon, 11 Jun 2012

Investors can finally trust financial theory in making portfolio decisions - according to new research from the University of East Anglia.

Six decades have passed since the Nobel laureate, Prof Harry Markowitz, introduced the notion of efficient portfolios - portfolios that offer the highest return for an acceptable level of investment risk.

Despite being the cornerstone of modern portfolio theory, investors avoid the use of efficient portfolios since they are hard to be implemented in practice.

But research published today in the Journal of Banking and Finance proposes a new method that makes portfolio theory more practically valuable and user-friendly.

The implementation of efficient portfolios requires an accurate forecast of the covariance matrix of the returns on the available investments - a statistic that summarizes the risk of the investments and how their returns are associated.

But due to limitations in the information available, investors cannot estimate the covariance matrix with accuracy. As a result, the use of efficient portfolios can lead to significant losses.

The new study develops a novel but simple framework to estimate this important unknown parameter.

Lead author Dr Apostolos Kourtis from UEA's Norwich Business School said: "Households as well as professional investors tend to neglect theoretical portfolio choice methods in favour of naive techniques.

"However, our research provides a way to re-establish the practical value of portfolio theory. This is achieved through a simple but very effective method to estimate the covariance matrix of investment returns.

"Our method's success is based on modifying a powerful statistical tool, known as shrinkage, to fit the investor's need for portfolios that perform well in practice."

Co-author Prof Raphael Markellos, also from UEA's Norwich Business School, said: "In fact, we show that our method generally outperforms most existing portfolio choice techniques in literature and investment practice by offering lower risk and higher risk-adjusted returns.

"Surprisingly, our approach performs very well even when information about prospective investments is quite limited.

"Our research, along with a couple of recent articles, shows that investors can finally trust financial theory in making portfolio decisions."

'Parameter Uncertainty in Portfolio Selection: Shrinking the Inverse Covariance Matrix' is published by the Journal of Banking and Finance.

Co-authors are Dr Apostolos Kourtis and Professor Raphael Markellos from UEA's Norwich Business School, and Dr George Dotsis from the Department of Economics of the University of Athens.