Filtering techniques are often applied to the estimation of dynamic latent variable models. However, these techniques are often based on a set assumptions which restrict models to be specified in a linear state-space form. Numerical filtering techniques have been propsed that avoid invoking such...

In this paper we propose a general component-driven model to analyze economic data with different characteristics (or regimes) in different time periods. Motivated by empirical data characteristics, our discussion focuses on a simple model driven by a random walk component and a stationary ARMA...

This paper investigates the effect of exchange rate volatility on US-UK bilateral trade flows. As part of econometric problems arising from a generated variable, we consider a special case when an ARCH type auxiliary model is used to measure uncertainty in the exchange rate and discuss a...

The aim of this paper is to identify permanent and transitory shocks. This identification is done according to the size of the shocks or the size of some other important economic variable. In order to be able to carry this identification scheme on, we introduce a new class of threshold models:...

This paper investigates the effect of exchange rate volatility on US-UK bilateral trade flows. As part of econometric problems arising from a generated variable, we consider a special case when an ARCH type auxiliary model is used to measure uncertainty in the exchange rate and discuss a...

In this paper we take a different modeling approach based on the component driven (CD) model developed in Kuan, Huang, and Tsay~(2003) to test the permanent income hypothesis (PIH), an example of intertemporal choice models. A key feature of this approach is that it explicitly allows for state...

The use of the Beveridge Nelson decomposition in macroeconomic analysis involves the truncation and estimation of infinite weighted sums of random variables, whereas the single source of error (SSE) state space approach provides a simple and effective framework that leads to exactly the same...

In real time forecasting, the sample is usually split into an estimation period of R observations and a prediction period of P observations, where T=R+P. Parameters are often estimated in a recursive manner, initially using R observations, then R+1 observations and so on until T-1 observations...