Читать книгу Asset Allocation - William Kinlaw, Mark P. Kritzman - Страница 70
Determining Relative Importance by Simulation
ОглавлениеThe associations between standard deviation, correlation, and relative volatility are easy to illustrate when we consider only two asset classes each divided equally between only two securities. These associations become less clear when we consider several asset classes weighted differently among hundreds of securities with a wide range of volatilities and correlations. Under these real-world conditions, it is easier to resolve the question of relative importance by a simulation procedure known as bootstrapping.
Unlike Monte Carlo simulation, which draws random observations from a prespecified distribution, bootstrapping simulation draws random observations with replacement from empirical samples. Specifically, we generate thousands of random portfolios from a large universe of securities that vary only because of asset allocation or security selection. This allows us to observe the distribution of available returns associated with each investment decision as opposed to studying the actual performance of managed funds, which reflects the biases of their investors. Kritzman and Page (2002) conducted such an analysis for investment markets in five countries: Australia, Germany, Japan, the United Kingdom, and the United States based on returns from 1988 to 2001. They showed that the dispersion around average performance arising from security selection was substantially greater than the dispersion arising from asset allocation in every country, and it was particularly large in the United States because the United States has a larger number of individual securities.5
L'Her and Plante (2006) refined the Kritzman and Page methodology to account for the relative capitalization of securities, and they also included a broader set of asset classes. Their analysis showed asset allocation and security selection to be approximately equally important – still a far different result from the conclusion of Brinson, Hood, and Beebower.