EMPIRICAL RESULTS

In Exhibit 16.2, we quantify how much each of the three components—country selection, country noise, and pure stock selection—explains the stock portfolio's holdings. To do this, we rewrite the above decomposition as a weighted average of “equal” parts, given by the components scaled to a constant level of expected risk, using BARRA's estimate of the covariance matrix. The weights in this weighted average indicate the relative importance of the components in the composition of the stock selection portfolio's cross-section of holdings. Any portfolio can be decomposed in this way, and the weights in the decomposition indicate the extent to which each of the components explains the portfolio's positions. The numbers in this exhibit show median values of the component weights over time.

Before we discuss the implications of the results shown in Exhibit 16.2, let us give an intuitive interpretation of these numbers. Consider, for example, the case of the naive stock portfolio based on value in emerging markets. This portfolio's stock weights are 41% explained by a country selection component, 25% explained by a country noise component, and only 34% explained by a pure stock selection component. This breakdown means that investing $100 in this naive stock selection portfolio is equivalent to allocating $41 to a country selection manager, $25 to a country noise “manager,” and only $34 to a pure stock selection manager.8 Note that, by construction, all country-neutral ...

Get Equity Valuation and Portfolio Management now with the O’Reilly learning platform.

O’Reilly members experience books, live events, courses curated by job role, and more from O’Reilly and nearly 200 top publishers.