Abstract
In recent years, analysis of financial time series has focused largely on data related to market trading activity. Apart from modelling the conditional variance of returns within the GARCH family of models, presently attention has also been devoted to other market variables, especially volumes, number of trades and durations. The financial econometrics literature has focused on Multiplicative Error Models (MEMs), which are considered particularly suited for modelling certain financial variables. The paper establishes an econometric specification approach for MEMs. In the literature, several procedures are available to perform specification testing for MEMs, but the proposed specification testing method is particularly useful within the context of the MEMs of financial duration. The paper makes a number of important theoretical contributions. Both the proposed specification testing method and the associated theory are established and evaluated through simulations and real data examples.
Highlights
The Multiplicative Error Models (MEMs) is first introduced by Engle (2002) as a general class of time series models for positive-valued random variables, which are decomposed into the product of their conditional mean and a positive-valued error term
The idea of the MEM is wellknown in financial econometrics since it originates from the structure of the autoregressive conditional heteroskedasticity (ARCH) model introduced by Engle (1982) and the stochastic volatility model proposed by Taylor (1982), where the conditional variance is dynamically parameterized and multiplicatively interacts with an innovation term
The flexibility of the autoregressive conditional duration (ACD) class of models depends partially on the design of the conditional duration, which is determined in this paper by the functional form of θ
Summary
The MEM is first introduced by Engle (2002) as a general class of time series models for positive-valued random variables, which are decomposed into the product of their conditional mean and a positive-valued error term. The idea of the MEM is wellknown in financial econometrics since it originates from the structure of the autoregressive conditional heteroskedasticity (ARCH) model introduced by Engle (1982) and the stochastic volatility model proposed by Taylor (1982), where the conditional variance is dynamically parameterized and multiplicatively interacts with an innovation term. Chou (2005) proposes an alternative specification known as the conditional autoregressive range model to study the dynamics of the high/low range of asset prices within fixed time intervals.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have