The accuracy of spread decomposition models in capturing informed trades Evidence from the London Stock Exchange and behavioural implications

Abstract Purpose – The purpose of this paper is to examine the empirical relationship between trades undertaken by informed agents (managers) and the proxies for informed trades computed by bid-ask spread decomposition models. Design/methodology/approach – An econometric application of spread decomposition models to data from the London Stock Exchange, with an examination of whether the model predictions are co-integrated with actual outcomes. Findings – The authors find overwhelming evidence of non-stationary behaviour between the actual and predicted informed trade prices. The findings suggest that there is a clear need for an alternative to extant spread decomposition models perhaps incorporating findings from behavioural finance. Originality/value – Given the importance of stock market liquidity and the extensive use of spread decomposition models in predicting informed trades, the authors believe that the research conducted in the paper is an important contribution to the market microstructure literature. Keywords Information asymmetry, Behavioural finance, Bid-ask spread, Spread decomposition models, Time series modelling Paper type Research paper