(en) This paper proposes an asymmetric autoregressive conditional duration (ACD) model, which extends the ACD model of Engle and Russell (1998). The asymmetry consists of letting the duration process depend on the state of the price process. If the price has increased, the parameters of the ACD model can differ from what they are if the price has decreased. The model is applied to the bid-ask quotes of two stocks traded on the NYSE and the evidence in favour of asymmetry is strong. Information effects (Easley and O'Hara 1992) are also empirically relevant. As the model is a transition model for the price process, it delivers `market forecasts' of where prices are heading. A trading strategy based on the model is implemented using tick-by-tick data.
Affiliations
Louvain School of ManagementAccounting & Finance
Citations
APA
Chicago
FWB
Bauwens, L., & Giot, P. (2003). Asymmetric ACD models : Introducing price information in ACD model with a two state transition model. Empirical Economics, 28(4), 709-731. https://doi.org/10.1007/s00181-003-0155-7 (Original work published 2003)