Since the true nature of a time series process is often unknown it is important to understand the effects of model choice. This paper examines how the choice between modelling stationary time series as ARMA or ARFIMA processes affects the accuracy of forecasts. This is done, for first-order...

This paper demonstrates that long memory leads to spurious rejection of the linearity hypothesis, when a STAR specification constitutes the alternative.

This paper considers nine long Swedish macroeconomic time series whose business cycle properties were discussed by Englund, Persson, and Svensson (1992) using frequency domain techniques. It is found by testing that all but two of the logarithmed and difference series are non-linear. The...

Asymptotic tests for fractional integration are usually badly sized in small samples, even for normally distributed processes. Furthermore, tests that are well-sized under normality may be severely distorted by non-normalities and ARCH errors. This paper demonstrates how the bootstrap can be...

The LM type linearity test for STAR nonlinearities is severely distorted when the process is governed by conditional heteroskedasticity. In order to correct the test we propose a parametric bootstrap. It is shown, by means of Monte Carlo methods, that the bootstrap test is almost exact.

The paper discusses a simple univariate nonlinear parametric time-series model for unemployment rates, focusing on the asymmetry observed in many OECD unemployment rate series. The model is based on a standard logistic smooth transition autoregressive (LSTAR) model for the first difference of...

This paper examines the predictability memory of fractionally integrated ARMA processes. Very long memory is found for positively fractionally integrated processes with large positive AR parameters. However, negative AR parameters absorb, to a great extent, the memory generated by a positive...

Asymptotic tests for fractional integration, such as the Geweke-Porter-Hudak test, the modified rescaled range test and Lagrange multiplier type tests, exhibit size-distortions in small-samples. This paper investigates a parametric bootstrap testing procedure, for size-correction, by means of a...

When testing the null hypothesis of linearity of a univariate time series against smooth transition autoregression (STAR), standard asymptotic distribution results do not apply since nuisance parameters in the model are unidentified under the null hypothesis. The prevailing test of Luukkonen,...

A smooth transition autoregressive model is estimated for the Southern Oscillation Index, an index commonly used as a measure of El Niño events. Using standard measures there is no indication of nonstationarity in the index. A logistic smooth transition autoregressive model describes the most...