Modern portfolio theory : a quantitative approach to traditional theory and a proposed variant

Abstract

Modern Portfolio Theory has been the most widely accepted theory of portfolio optimisation, yet the shortcomings in the actual application of the theory to financial markets are considerable. This study analyses the tradtional method of portfolio optimisation and the asset allocation it produces. By applying this method to the US stock market, this reseach evaluates whether the traditional method delivers the expected results. In addition, this dissertation proposes an alternative method of portfolio optimisation, one which uses forward-looking inputs in an attempt to produce an asset allocation that is more suitable for the market structure expected in the next period. Finally, this study compares the ability of the two methods in producing a portfolio beta that better predicts the actual portfolio beta during a five-year forecast period. The asset allocation produced by both methods is also put to the test by analysing the performance of the optimal portfolios suggested by the two methods.