In this paper, different approaches to dealing with nuisance parameters in the likelihood based inference are presented and illustrated by reference to the linear regression model with nonspherical errors. The estimator of the error variance using each of the approaches is also derived for the...

The global linear trend with autocorrelated disturbances is a surprising omission from the M1 competition. This approach to forecasting is therefore evaluated using the 51 non-seasonal series from the competition. It is contrasted with a fully optimized version of Holts trend corrected...

In this paper Kuznets' U-Curve hypothesis is tested on two unbalanced panel data sets of 47 and 62 countries, for the period 1970-93, using two-way fixed and random effects models. Several competing model specifications are estimated and the one best fitting the data is selected by appropriate...

The problem of constructing prediction intervals for linear time series (ARIMA) models is examined. The aim is to find prediction intervals which incorporate an allowance for sampling error associated with parameter estimates. The effect of constraints on parameters arising from stationary and...

A Bayesian estimation procedure is developed for estimating multiple regime vector autoregressive models appropriate for deviations from financial arbitrage relationships. This approach has clear advantages over classical stepwise threshold autoregressive analysis.

The degree of substitution between private and public per capita consumption for the G7 countries is estimated over the period 1960 to 1996. Special attention is given to isolating both long-run and short-run substitution effects.

This article derives analystic finite sample approximations to the bias and standard error of a class of statistics which test the hypothesis of no serial correlation in market returns. They offer an alternative to both the widely used Monte Carlo approach for calculating the bias, as well as...

In this paper we propose a test statistic to compare two or more stationary time series that are not necessarily independent. The test is based on the difference between estimated parameters of the autoregressive models that are fitted to the series.

The segmentation problem arises in many applications in data mining, A.I. and statistics. In this paper, we consider segmenting simple time series. We develop two Bayesian approaches for segmenting a time series, namely the Bayes Factor approach, and the Minimum Message Length (MML) approach. We...

Time series often have patterns that form a basis for comparing them or classifying them into groups. In this paper we present procedures to compare and classify stationary multivariate time series. Simulations studies show that the procedures perform fairly well for reasonably long series.