Article ID: | iaor1991564 |
Country: | United Kingdom |
Volume: | 41 |
Issue: | 7 |
Start Page Number: | 599 |
End Page Number: | 607 |
Publication Date: | Jul 1990 |
Journal: | Journal of the Operational Research Society |
Authors: | Meade N., Salkin G.R. |
Keywords: | programming: quadratic |
An index fund is a portfolio of shares designed to replicate the investment performance of a market index. The index represents the behaviour of the market as a whole. This paper describes the selection of an index fund which minimizes expected tracking error. Using a multivariate model of returns on shares, a development of a univariate model by Taylor, the selection problem is formulated as a quadratic programme. The effects of various constraints on tracking error are demonstrated. Several policies for the readjustment of a fund are examined in the context of the differing objectives of fund managers. As a general rule, regular readjustment is shown to be a more expensive policy than irregular updating.