Mutual fund managers often form portfolios using one of two distinct methods. The first is a 'top-down' macro-economic approach, while the competing method is a 'bottom-up' company specific approach. This study looks at the 'top-down' versus 'bottom-up' approaches to equity portfolio formation, specifically, the effect of focusing on industry sectors versus securities. Two approaches are utilised to assess this. The first is a bootstrapping method, whereby two sets of hypothetical portfolios are formed from US stock data over the period January 1970 to December 2005 through varying either sector or security allocations in the portfolio. This technique allows an assessment of the effect that these two decision making processes can have on portfolio performance. After this, portfolio holdings of equity managers are looked at to assess the level of portfolio performance being generated in US mutual funds from 1980 to 2005 that is attributable to either the manager's sector selection or security selection.
The results indicate that, for a manager with equal skill at either selecting securities or sectors, the larger part of their outperformance would stern from their choice of sectors rather than their selection of stocks within the sectors. From assessing the holdings data of mutual fund managers, it was found that for the average manager, the returns are derived marginally more through their choice of security rather than sector. These two results contradict somewhat, therefore potential explanations are put forth to mediate the divergence that was obtained from these two methodologies.
The primary objective of this thesis is to further the understanding of the optimal method to construct equity portfolios and to assess how this interrelates with how managers typically generate returns for their portfolios.