Papers by Timo Teräsvirta
Macroeconomic Dynamics, 2002
The paper discusses a simple univariate nonlinear parametric time-series model for unemployment r... more The paper discusses a simple univariate nonlinear parametric time-series model for unemployment rates, focusing on the asymmetry observed in many OECD unemployment series. The model is based on a standard logistic smooth transition autoregressive (LSTAR) model for the first difference of unemployment, but it also includes a lagged level term. This model allows for asymmetric behaviour by permitting 'local' nonstationarity in a globally stable model. Linearity tests are performed for a number of quarterly, seasonally unadjusted, unemployment series from OECD countries, and linearity is rejected for a number of them. For a number of series, nonlinearity found by testing can be modelled satisfactorily by use of our smooth transition autoregressive model. The properties of the estimated models, including persistence of the shocks according to them, are illustrated in various ways and discussed. Possible existence of moving equilibria in series not showing asymmetry is investigated and modelled with another STAR model.
Journal of Econometrics, 2002
ABSTRACT Introduction to special Annals issue of papers presented at a conference in Cardiff, UK ... more ABSTRACT Introduction to special Annals issue of papers presented at a conference in Cardiff, UK on July 9–11th 2000
Journal of Econometrics, 2002
Journal of Business & Economic Statistics, 2006
This article contains two novelties. First, a unified framework for testing and evaluating the ad... more This article contains two novelties. First, a unified framework for testing and evaluating the adequacy of an estimated autoregressive conditional duration (ACD) model is presented. Second, two new classes of ACD models, the smooth transition ACD model and the time-varying ACD model, are introduced and their properties discussed. New misspecification tests for the ACD class of models are introduced. These are Lagrange multiplier and Lagrange multiplier-type tests against general forms of additive and multiplicative misspecification of the conditional mean function. These forms include tests against higher-order models, tests of no remaining ACD in the standardized durations, and tests of linearity and parameter constancy. The finite-sample properties of the tests are investigated by simulation. Main statistical properties of the two new classes of ACD models that serve as alternatives in tests of linearity and parameter constancy are investigated. Finally, the tests are applied to ACD models of IBM stock traded at the New York Stock Exchange.
Journal of Applied Econometrics, 1999
This paper considers nine long Swedish macroeconomic time series whose business cycle properties ... more This paper considers nine long Swedish macroeconomic time series whose business cycle properties were discussed by using frequency domain techniques. It is found by testing that all but two of the logarithmed and differenced series are nonlinear. The observed nonlinearity is characterized by STAR models. The statistical and dynamic properties of the estimated STAR models are investigated using, among other things, parametrically estimated 'local' or 'sliced' spectra. Cyclical variation at business cycle frequencies does not seem to be constant over time for all series, and it is difficult to find a 'Swedish business cycle'. Only two series may be regarded as having genuinely asymmetric cyclical variation. Standard Granger noncausality tests are adapted to the nonlinear (STAR) case, and the null hypothesis of noncausality is tested for pairs of series. The results point at strong temporal interactions between series. They also indicate that the assumption of functional form (linear or STAR) strongly affects the outcome of these pairwise tests.
Journal of Applied Econometrics, 2001
In this paper we reconsider an error-correction model of UK broad money demand by Ericsson, Hendr... more In this paper we reconsider an error-correction model of UK broad money demand by Ericsson, Hendry and Prestwich. Their model is non-linear in both variables and parameters, and it can be viewed as an approximation to a smooth transition regression (STR) type specification. The corresponding STR model, when specified and estimated, fits the data better than the original model. Adopting a somewhat more general modelling approach leads to another STR model. This model variance dominates the other two, and the encompassing tests performed in this paper indicate that it is an improvement over the other two specifications.
International Journal of Forecasting, 1994
ELS EVIER International Journal of Forecasting 10 (1994) 4757 The combination of forecasts using ... more ELS EVIER International Journal of Forecasting 10 (1994) 4757 The combination of forecasts using changing weights Melinda Deutsch*'3, Clive WJ Granger3, Timo Terasvirta6 'Department of Economics, University of California, San Diego, La Jolla, CA 92093, USA 'Research ...
Empirical Economics, 1995
This paper considers modelling the annual logarithmed per capita gross national product of the Un... more This paper considers modelling the annual logarithmed per capita gross national product of the United States in 1889-1987. Some authors have suggested that the parameters of the process generating the data have changed over time but formal parameter constancy tests do not support this argument. The series turns out to be nonlinear and can be adequately characterized by an exponential smooth transition autoregressive model. For comparison, a detrended series is also considered, found nonlinear and modelled using a logistic smooth transition autoregressive model. The behaviour of the estimated models is discussed, and it is seen that nonlinearity is needed to describe the response of the process to exceptionally large exogenous shocks. The properties of the models are further investigated by forecasting several years ahead, and the forecasts are compared with those from other linear and nonlinear models.
Economics Letters, 1999
There has been considerable interest in recent years in long-memory, specifically fractionally in... more There has been considerable interest in recent years in long-memory, specifically fractionally integrated I(d) models. These processes have properties, in terms of the rate of decline of the correlogram against lag length, that lie between those of stationary autoregressive ...
In this paper we propose a general method for testing the Granger noncausality hypothesis in stat... more In this paper we propose a general method for testing the Granger noncausality hypothesis in stationary nonlinear models of unknown functional form. These tests are based on a Taylor expansion of the nonlinear model around a given point in a sample space. We study the performance of our tests by a Monte Carlo experiment and compare these to the most
SSRN Electronic Journal, 2000
Page 1. Statistical Properties of the Asymmetric Power ARCH Process by Changli He andTimo Teräsvi... more Page 1. Statistical Properties of the Asymmetric Power ARCH Process by Changli He andTimo Teräsvirta Department of Economic Statistics Stockholm School of Economics Box 6501, S-113 83 Stockholm, Sweden First draft: September 1997 ...
An error correction model for the demand for real M3 money is constructed for the period 1976-199... more An error correction model for the demand for real M3 money is constructed for the period 1976-1994 with real GNP, the GNP deflator as well as a short-term and a long-term interest rate as explanatory variables. Quarterly, seasonally unadjusted data are used in estimating the model. It is found that there is a clear structural break due to the German unification in 1990. On the other hand, once this structural break is accounted for, a stable relation is found which passes a series of misspecification tests. In particular, a number of recent tests of parameter constancy and linearity are applied. Our specification is at variance with findings reported by some other researchers, notably the Deutsche Bundesbank.
Annastiina Silvennoinen School of Finance and Economics, University of Technology Sydney, Box 123... more Annastiina Silvennoinen School of Finance and Economics, University of Technology Sydney, Box 123, Broadway NSW 2007, e-mail: annastiina.silvennoinen@uts.edu.au Timo Teräsvirta CREATES, School of Economics and Management, University of Aarhus, DK-8000 ...
Handbook of Financial Time Series, 2009
Realized volatility is a nonparametric ex-post estimate of the return variation. The most obvious... more Realized volatility is a nonparametric ex-post estimate of the return variation. The most obvious realized volatility measure is the sum of finely-sampled squared return realizations over a fixed time interval. In a frictionless market the estimate achieves consistency for the underlying quadratic return variation when returns are sampled at increasingly higher frequency. We begin with an account of how and why the procedure works in a simplified setting and then extend the discussion to a more general framework. Along the way we clarify how the realized volatility and quadratic return variation relate to the more commonly applied concept of conditional return variance. We then review a set of related and useful notions of return variation along with practical measurement issues (e.g., discretization error and microstructure noise) before briefly touching on the existing empirical applications.
Oxford Handbooks Online, 2011
In this paper, nonlinear models are restricted to mean nonlinear parametric models. Several such ... more In this paper, nonlinear models are restricted to mean nonlinear parametric models. Several such models popular in time series econometrics are presented and some of their properties discussed. This includes two models based on universal approximators: the Kolmogorov-Gabor polynomial model and two versions of a simple arti…cial neural network model. Techniques for generating multi-period forecasts from nonlinear models recursively are considered, and the direct (non-recursive) method for this purpose is mentioned as well. Forecasting with complex dynamic systems, albeit less frequently applied to economic forecasting problems, is brie ‡y highlighted. A number of large published studies comparing macroeconomic forecasts obtained using di¤erent time series models are discussed, and the paper also contains a small simulation study comparing recursive and direct forecasts in a particular case where the data-generating process is a simple arti…cial neural network model. Suggestions for further reading conclude the paper.
Review of Economics and Statistics, 1998
AbstractAn error correction model for the demand for real M3 money is constructed for the period... more AbstractAn error correction model for the demand for real M3 money is constructed for the period of 19761994 with real GNP, the GNP deflator, as well as a short-term and a long-term interest rate as explanatory variables. Quarterly, seasonally unadjusted data are used in ...
Studies in Nonlinear Dynamics & Econometrics, 2000
This paper contains a nonlinear, nonstationary autoregressive model whose intercept changes deter... more This paper contains a nonlinear, nonstationary autoregressive model whose intercept changes deterministically over time. The intercept is a flexible function of time, and its construction bears some resemblance to neural network models. A modelling technique, modified from one for single hidden-layer neural network models, is developed for specification and estimation of the model. Its performance is investigated by simulation and further illustrated by two applications to macroeconomic time series.
Oxford Bulletin of Economics and Statistics, 2009
Journal of Time Series Analysis, 1993
Abstract. Recently, a new linearity test for time series was introduced based on concepts from th... more Abstract. Recently, a new linearity test for time series was introduced based on concepts from the theory of neural networks. Lee et al. have already studied the power properties of this test and they are further investigated here. They are compared by simulation with those of a ...
Uploads
Papers by Timo Teräsvirta