This paper employes a parametric model of structural breaks in the mean of stock returns which allows them to be endogenously driven by large positive or negative stock market return shocks. These shocks can be taken to reflect important market announcements, monetary policy regime shifts and/or...

In this paper we suggest a number of statistical tests based on neural network models, that are designed to be powerful against structural breaks in otherwise stationary time series processes while allowing for a variety of nonlinear specifications for the dynamic model underlying them. It is...

The paper provides a proof of consistency of the ridge estimator for regressions where the number of regressors tends to infinity. Such result is obtained without assuming a factor structure. A Monte Carlo study suggests that shrinkage autoregressive models can lead to very substantial...

The paper provides a proof of consistency of the ridge estimator for regressions where the number of regressors tends to infinity. Such result is obtained without assuming a factor structure. A Monte Carlo study suggests that shrinkage autoregressive models can lead to very substantial...

The persistence properties of economic time series has been a primary object of investigation in a variety of guises since the early days of econometrics. This paper suggests investigating the persistence of processes conditioning on their history. In particular we suggest that examining the...

This paper presents an invariance principle for highly nonstationary long memory processes, defined as processes with long memory parameter lying in (1, 1.5). This principle provides the tools for showing asymptotic validity of the bootstrap in the context of such processes.

This paper constructs tests for the presence of nonlinearity of unknown form in addition to a fractionally integrated, long memory component in a time series process. The tests are based on artificial neural network structures and do not restrict the parametric form of the nonlinearity. The...

A prominent class of nonlinear time series models are threshold autoregressive models. Recently work by Kapetanios (2000) has shown in a Monte Carlo setting that the superconsistency property of the threshold parameter estimates does not translate to superior performance in small samples....

This paper revisits a number of data-rich prediction methods, like factor models, Bayesian ridge regression and forecast combinations, which are widely used in macroeconomic forecasting, and compares these with a lesser known alternative method: partial least squares regression. Under the...