A Monte Carlo study of the forecasting performance of empirical SETAR models

Abstract

In this paper we investigate the multi‐period forecast performance of a number of empirical self‐exciting threshold autoregressive (SETAR) models that have been proposed in the literature for modelling exchange rates and GNP, among other variables. We take each of the empirical SETAR models in turn as the DGP to ensure that the ‘non‐linearity’ characterizes the future, and compare the forecast performance of SETAR and linear autoregressive models on a number of quantitative and qualitative criteria. Our results indicate that non‐linear models have an edge in certain states of nature but not in others, and that this can be highlighted by evaluating forecasts conditional upon the regime. Copyright © 1999 John Wiley & Sons, Ltd.

Non-Linear Predictability in Stock and Bond Returns: When and Where is it Exploitable?

“…In fact, out of 36 cells signalling top performance, 16 go to simple benchmarks (AR(1) and random walk), while in 9 additional cases, augmenting the random walk and simple linear prediction models with ARCH-in mean effects, gives accurate predictions. 20 The evidence for French stock returns is mixed, although it is remarkable that in 11 cases we find evidence of accurate forecasting performance from linear predictive models augmented by ARCH-in mean effects.…”

Forecasting Performance of Exponential Smooth Transition Autoregressive Exchange Rate Models

“…For instance, the article by Diebold and Nason (1990) point out that there is no guarantee that the SETAR model will perform better than the linear AR model. A similar view is expressed in Clements and Smith (1997), where they note that the rejection of a null of linearity in favour of nonlinearity does not guarantee that the prediction based on the SETAR model will outperform the AR models.…”

Efficient Market Hypothesis and the RMB-Dollar Rates: A Nonlinear Modeling of the Exchange Rate

“…The results of the Root Mean Square Forecasting Error (RMSE) also give similar outcome. We therefore find it difficult to substantiate the claim of Boero andMarrocu (2002), andClement andSmith (1997) that the RMSE masks the superiority of a model.…”