This paper considers the analysis of cointegrated time series using principal components methods. These methods have the advantage of neither requiring the normalisation imposed by the triangular eror correction model, nor the specification of a finite order vector autoregression.

In this paper we present a test statistic, which will be used to test for significant differences between generating processes of two time series that may be logically connected. The test statistic is based on the differences between estimated parameters of the autoregressive models which are...

This paper studies the All Ordinaries Index in Australia, and its futures contract known as the Share Price Index. We use a new form of smooth transition model to account for a variety of nonlinearities caused by transaction costs and other market/data imperfections, and given the recent...

This paper studies linear and nonlinear autoregressive leading indicator models of business cycles in G7 countries. The models use the spread between short-term and long-term interest rates as leading indicators for GDP, and their success in capturing business cycles is gauged by non-parametric...

This study proposes a new approach to the estimation of daily volatility. This approach is different ( in the sense of using all available intraday price data) and unbiased ( in the sense of accounting for the high levels of autocorrelation found in intraday price data).

We develop nonlinear leading indicator models for GDP growth, with the interest rate spread and growth in M2 as leading indicators. Since policy makers are typically interested in whether or not a recession is imminent, we evaluate these models according to their ability to predict the...

This paper studies linear and linear autoregressive leading indicator models of business cycles in OECD countries. The models use the spread between short-term and long-term interest rates as leading indicators for GDP, and their success in capturing business cycles gauged by the non-parametric...

In many applications, there are multiple time series that are hierarchically organized and can be aggregated at several different levels in groups based on products, geography or some other features. We call these "hierarchical time series". They are commonly forecast using either a "bottom-up"...

We review the past 25 years of time series research that has been published in journals managed by the International Institute of Forecasters (Journal of Forecasting 1982-1985; International Journal of Forecasting 1985-2005). During this period, over one third of all papers published in these...

The basic ideals underlying the Kalman filter are outlined in this paper without direct recourse to the complex formulae normally associated with this method. The novel feature of the paper is its reliance on a new algebraic system based on the first two moments of the multivariate normal...