Title: Dynamic estimation of information asymmetry risk in trading
Authors: Jim Griffin - University of Kent (United Kingdom)
Jaideep Oberoi - University of Kent (United Kingdom) [presenting]
Samuel Oduro - easyJet (United Kingdom)
Abstract: A new dynamic indicator is proposed for the risk of informed trading inferred from traded prices, volume and bid-ask spread. Several widely used measures of information asymmetry rely on decomposing traded volume into expected and unexpected components. The approach exploits the non-linear relationship between bid-ask spreads and volume. Both volume and spreads have been shown to be (partly) driven by an underlying liquidity process. We model a stochastic latent information process in a state-space model that allows us to identify different types of volume changes, and their implications for informed trading. We use a Bayesian procedure with a Gibbs sampling algorithm to estimate the model on a sample of 10 stocks over a 10 year period. The results are consistent with the opposing effects of informed and uninformed volume on the spread. We also find that both informed and uninformed trading are significantly persistent over the sample period.