Comovement and FTSE 100 index changes

We employ the Barberis et al. (2005) methodology to investigate the impact of changes to the FTSE 100 index on return comovement 1992-2002. For FTSE entries, the average weekly increase in the beta coefficient is 0.38 in univariate regressions and 0.60 in bivariate regressions that control for the return on non-FTSE stocks. Stocks deleted from the index display the opposite pattern post exit. The results are robust to a number of factors including size, industry and non-trading effects. They are difficult to explain within a classical framework but complement those found for the USA, Japan and Canada in supporting behavioural finance views of comovement.