Optimal portfolio allocations for global bank stocks in local surrencies and US dollars 1992 - 2001

Abstract

In this paper we examine the optimal composition of global portfolios of bank stocks, expressed in both local currencies and in US dollar terms, over the period January 1992 to June 2001. We estimate optimal global bank stock portfolios using two covariance optimisation algorithms – the Markowitz expected return/variance algorithm (MPT), and the Elton, Gruber and Padberg average correlation algorithm (EGP) – and compare the composition and performance of these portfolios with a portfolio comprising equally-weighted bank stocks. Our study also includes measures of skewness and kurtosis, and risk adjusted return measures based on variance, semivariance and portfolio betas. The purpose of our study is twofold. First, we wish to examine whether the covariance optimisation approaches produce significantly different portfolio allocations over the period of the study. Second, we wish to determine if the two significant events for the global banking sector in this period– the implementation of global risk-based capital adequacy standards in 1992 and the Asian banking crisis of 1997 – may have had any influence on the optimal allocation of global bank stocks in an investment portfolio. To achieve this we construct the optimal bank portfolios, using both optimisation algorithms, for the period 19921996 and 1997-2001. We include bank stock returns for 26 countries in the study. We find that the MPT and EGP optimisation algorithms do produce different portfolio allocations during both periods of the study. If return is measured against variance, the MPT algorithm produces the best performing portfolio. However if return is measured against semi-variance, the results are mixed. We also find that bank portfolios performed better on a risk-adjusted basis in the period leading up to the Asian crisis of 1997. Our most interesting finding is that if the highest risk bank stocks are removed from the portfolio, the terminal wealth of the portfolio falls by around half in each period. This suggests that higher-risk bank stocks are needed to achieve sufficient diversification to ‘protect’ the return for a global portfolio of bank stocks.