Bayesian Inference of Multiscale Stochastic Conditional Duration Models
Posted: 3 Sep 2014
Date Written: September 1, 2014
Abstract
In this paper we revisit the notion that a single factor of duration running on single time scale is adequate to capture the dynamics of the duration process of financial transaction data. The documented poor fit of the left tail of the marginal distribution of the observed durations in some existing one-factor stochastic duration models may be indicative of the possible existence of multiple stochastic duration factors running on different time scales. This paper proposes multiscale stochastic conditional duration (MSCD) models to describe the dynamics of duration of financial transaction data. Suitable algorithms of MCMC are developed to fit the resulting MSCD models under three distributional assumptions about the innovation of the measurement equation. Simulation studies suggest that our proposed models and methods result in improved in-sample fits as well as improved duration forecasts. Applications of our models and methods to two duration data sets of FIAT and IBM indicate the existence of at least two factors governing the dynamics of the duration of the stock transactions.
Keywords: Markov Chain Monte Carlo; Multiscale; Auxiliary particle filter; Probability integral transform; Deviance information criterion.
JEL Classification: C10; C41; G10
Suggested Citation: Suggested Citation