nep-ets New Economics Papers
on Econometric Time Series
Issue of 2014‒04‒05
six papers chosen by
Yong Yin
SUNY at Buffalo

  1. Noncausal Bayesian Vector Autoregression By Markku Lanne; Jani Luoto
  2. Two EGARCH models and one fat tail By Michele Caivano; Andrew Harvey
  3. Forecasting Co-Volatilities via Factor Models with Asymmetry and Long Memory in Realized Covariance By Manabu Asai; Michael McAleer
  4. Reduced-rank time-varying vector autoregressions By Joris de Wind; Luca Gambetti
  5. On uniqueness of moving average representations of heavy-tailed stationary processes By Gouriéroux, Christian; Zakoian, Jean-Michel
  6. The seasonal KPSS Test: some extensions and further results By El Montasser, Ghassen

  1. By: Markku Lanne (University of Helsinki and CREATES); Jani Luoto (University of Helsinki)
    Abstract: We propose a Bayesian inferential procedure for the noncausal vector autoregressive (VAR) model that is capable of capturing nonlinearities and incorporating effects of missing variables. In particular, we devise a fast and reliable posterior simulator that yields the predictive distribution as a by-product. We apply the methods to postwar quarterly U.S. inflation and GDP growth series. The noncausal VAR model turns out to be superior in terms of both in-sample fit and out-of-sample forecasting performance over its conventional causal counterpart. In addition, we find GDP growth to have predictive power for the future distribution of inflation over and above the own history of inflation, but not vice versa. This may be interpreted as evidence against the new Keynesian model that implies Granger causality from inflation to GDP growth, provided GDP growth is a reasonable proxy of the marginal cost.
    Keywords: Noncausal time series, non-Gaussian time series, Bayesian analysis, New Keynesian model
    JEL: C11 C32 E31
    Date: 2014–05–24
  2. By: Michele Caivano (Bank of Italy); Andrew Harvey (University of Cambridge)
    Abstract: We compare two EGARCH models, which belong to a new class of models in which the dynamics are driven by the score of the conditional distribution of the observations. Models of this kind are called dynamic conditional score (DCS) models and their form facilitates the development of a comprehensive and relatively straightforward theory for the asymptotic distribution of the maximum likelihood estimator. The EGB2 distribution is light-tailed, but with a higher kurtosis than the normal distribution. Hence it is complementary to the fat-tailed t. The EGB2-EGARCH model gives a good fit to many exchange rate return series, prompting an investigation into the misleading conclusions liable to be drawn from tail index estimates.
    Keywords: exchange rates, heavy tails, Hill’s estimator, score, robustness, EGB2, Student’s t, tail index
    JEL: C22 G17
    Date: 2014–03
  3. By: Manabu Asai; Michael McAleer (University of Canterbury)
    Abstract: Modelling covariance structures is known to suffer from the curse of dimensionality. In order to avoid this problem for forecasting, the authors propose a new factor multivariate stochastic volatility (fMSV) model for realized covariance measures that accommodates asymmetry and long memory. Using the basic structure of the fMSV model, the authors extend the dynamic correlation MSV model, the conditional/stochastic Wishart autoregressive models, the matrix-exponential MSV model, and the Cholesky MSV model. Empirical results for 7 financial asset returns for US stock returns indicate that the new fMSV models outperform existing dynamic conditional correlation models for forecasting future covariances. Among the new fMSV models, the Cholesky MSV model with long memory and asymmetry shows stable and better forecasting performance for one-day, five-day and ten-day horizons in the periods before, during and after the global financial crisis.
    Keywords: Dimension reduction; Factor Model; Multivariate Stochastic Volatility; Leverage Effects; Long Memory; Realized Volatility
    JEL: C32 C53 C58 G17
    Date: 2014–03–17
  4. By: Joris de Wind; Luca Gambetti
    Abstract: The standard time-varying VAR workhorse suffers from overparameterization, which is a serious problem as it limits the number of variables and lags that can be incorporated in the model. Read also: CPB Discussion Paper 271 ' Time variation in the dynamic effects of unanticipated changes in tax policy '. As a solution for the overparameterization problem, we propose a new, more parsimonious time-varying VAR model setup with which we can reliably estimate larger models including more variables and/or more lags than was possible until now. The new model setup implies cross-equation restrictions on the time variation that are empirically supported, theoretically appealing, and make the Bayesian estimation procedure much faster.
    JEL: C52 C53 E37
    Date: 2014–03
  5. By: Gouriéroux, Christian; Zakoian, Jean-Michel
    Abstract: We prove the uniqueness of linear i.i.d. representations of heavy-tailed processes whose distribution belongs to the domain of attraction of an $\alpha$-stable law, with $\alpha
    Keywords: $\alpha$-stable distribution; Domain of attraction; Infinite moving average; Linear process; Mixed causal/noncausal process; Nonparametric identification; Unobserved component model.
    JEL: C14 C22 C32
    Date: 2014–03–31
  6. By: El Montasser, Ghassen
    Abstract: The literature distinguishes finite sample studies of seasonal stationarity quite less intensely than it shows for seasonal unit root tests. Therefore, the use of both types of tests for better exploring time series dynamics is seldom noticed in the relative studies on such a topic. Recently, Lyhagen (2006) introduced for quarterly data the seasonal KPSS test which null hypothesis is no seasonal unit roots. In the same manner, as most unit root limit theory, the asymptotic theory of the seasonal KPSS test depends on whether the data has been filtered by a preliminary regression. More specifically, one may proceed to the extraction of deterministic components – such as the mean and trend – from the data before testing. In this paper, I took account of de-trending on the seasonal KPSS test. A sketch of its limit theory was provided in this case. Also, I studied in finite sample the behaviour of the test for monthly time series. This could enrich our knowledge about it since it was – as I mentioned above – early introduced for quarterly data. Overall, the obtained results showed that the seasonal KPSS test preserved its good size and power properties. Furthermore, like the test of Kwiatkowski et al. [KPSS] (1992), the nonparametric corrections of residual variances may smooth the wide variations of the seasonal KPSS empirical sizes.
    Keywords: KPSS test, deterministic seasonality, Brownian motion, LM test
    JEL: C32
    Date: 2014–03–10

This nep-ets issue is ©2014 by Yong Yin. It is provided as is without any express or implied warranty. It may be freely redistributed in whole or in part for any purpose. If distributed in part, please include this notice.
General information on the NEP project can be found at For comments please write to the director of NEP, Marco Novarese at <>. Put “NEP” in the subject, otherwise your mail may be rejected.
NEP’s infrastructure is sponsored by the School of Economics and Finance of Massey University in New Zealand.