Is There Really Excess Co-movement? Causal Evidence from FTSE 100 Index Turnover

Empirical research shows co-movement in excess of covariation of fundamental factors, like discount rates and future cash flows. Apparently, events that have nothing to do with fundamentals influence co-movement: for example, co-movement increases when stocks enter a major stock index and decreases it when they are being removed from it. The excess co-movement is generated by typical patterns applied by investor groups and based on it the empirical literature denies the fundamentals-based hypothesis.

In the paper “Is There Really Excess Co-movement? Causal Evidence from FTSE 100 Index Turnover”, available at: SSRN , author Christian von Drathen analyses further this hypothesis and notices that previous studies rely on variation in index membership, which is not random, and makes it difficult to state if the index turnover is a factor that generates excess co-movement.

No effects found

The paper analyzes if stock index turnover generates a change in co-movement of stock returns and, specifically for the FTSE 100 index, no significant causal effect is found. Index turnover does not generate any change in co-movement, but, on the other side, a change in unobserved stock characteristics correlated with co-movement it is likely to generate index turnover. A causal test of fundamental-based hypothesis is performed, focusing on random membership changes in the FTSE 100 index which were not linked to changes in the covariance structure of fundamental factors.

FTSE 100 balancing index turnover offers no evidence of excess co-movement, so that the author cannot reject the fundamentals-based hypothesis on stock markets. That goes against previous literature conclusions, which explain excess co-movement through common liquidity shocks generated by the price impact of correlated investor demand. Other literature sources find that S&P 500 index turnover changes co-movement and relate this to investors trading index stocks together, but it cannot be put in similarity relationship with the FTSE 100, as the S&P 500 index turnover is correlated with unobserved stock characteristics, while the FTSE membership rules are transparent and mechanical.

Finally, balancing index turnover proves to be a method that offers good perspectives for analyzing asset markets phenomena where selections issues are involved.