In Defence of Capitalisation Weights: Evidence from the FTSE 100 and S&P 500 Indices


  • Special thanks go to an anonymous reviewer for her/his very useful comments and to John Doukas (the editor). The author is responsible for any errors that remain. The author would also like to thank the following: the University of Stirling for funding this research; Kevin Campbell, Alan Goodacre, Rezaul Kabir, Chris Veld, and Chris Brooks for providing valuable suggestions; participants at the Scottish BAA conference in September 2006; participants at the Fund Management Conference at the Xfi Centre in Exeter in December 2006; Ser-Huang Poon and Clive Granger for their comments at the 2003 Volatility Workshop held at the University of Strathclyde in Glasgow; Li, Miffre, Brooks, and Sullivan (2008) for allowing the author to use their momentum data for the UK market; and Dimson, Nagel, and Quigley (2003) and Kenneth R. French for their data library.


A simple method for decomposing the variance covariance matrix of portfolio returns at the level of individual stocks is applied to the FTSE 100 Index. During extreme negative shocks, the largest index constituents exhibit lower than average covariance, thereby reducing the volatility of the capitalisation-weighted index. The risk-adjusted returns of the capitalisation-weighted FTSE 100 Index exceed those of an equally-weighted version of the same index and the outperformance is robust to the method of risk adjustment applied. The equally-weighted index also exhibits greater systematic (market) risk than the capitalisation-weighted version.