nep-ets New Economics Papers
on Econometric Time Series
Issue of 2007‒06‒02
six papers chosen by
Yong Yin
SUNY at Buffalo

  1. Comparing smooth transition and Markov switching autoregressive models of US Unemployment By Philippe J. Deschamps
  2. Dynamic Panel Data Models with Cross Section Dependence and Heteroscedasticity By Kazuhiko Hayakawa
  3. A Simple Efficient Instrumental Variable Estimator in Panel AR(p) Models By Kazuhiko Hayakawa
  4. Wavelet Analysis and Denoising: New Tools for Economists By Iolanda Lo Cascio
  5. A look into the factor model black box - publication lags and the role of hard and soft data in forecasting GDP By Marta Ba?bura; Gerhard Rünstler
  6. Does implied volatility reflect a wider information set than econometric forecasts? By Ralf Becker; Adam Clements; James Curchin

  1. By: Philippe J. Deschamps (Department of Quantitative Economics)
    Abstract: Logistic smooth transition and Markov switching autoregressive models of a logistic transform of the monthly US unemployment rate are estimated by Markov chain Monte Carlo methods. The Markov switching model is identified by constraining the first autoregression coefficient to differ across regimes. The transition variable in the LSTAR model is the lagged seasonal difference of the unemployment rate. Out of sample forecasts are obtained from Bayesian predictive densities. Although both models provide very similar descriptions, Bayes factors and predictive efficiency tests (both Bayesian and classical) favor the smooth transition model.
    Keywords: Logistic smooth transition autoregressions; Hidden Markov models; Density forecasts; Markov chain Monte Carlo; Bridge sampling; Unemployment rate
    JEL: C11 C22 C53 E24 E27
    Date: 2007–05–24
  2. By: Kazuhiko Hayakawa
    Abstract: In this paper, we show that the bias-corrected first-difference (BCFD) estimator suggested by Chowdhury (1987) can be applied to the case where the error terms are cross-sectionally dependent and heteroscedastic. By deriving the finite sample bias of the BCFD estimator, we find that the BCFD estimator has small bias when T, the dimension of the time series, is not very large and ƒÏ, the autoregressive parameter, is close to one. Simulation results show that the BCFD estimator performs better than existing estimators, especially when T is not very large.
    Date: 2007–05
  3. By: Kazuhiko Hayakawa
    Abstract: In this paper, we show that for panel AR(p) models with iid errors, an instrumental variable (IV) estimator with instruments in the backward orthogonal deviation has the same asymptotic distribution as the infeasible optimal IV estimator when both N and T, the dimensions of the cross section and the time series, are large. If we assume that the errors are normally distributed, the asymptotic variance of the proposed IV estimator is shown to attain the lower bound when both N and T are large. A simulation study is conducted to assess the estimator.
    Keywords: panel AR(p) models, the optimal instruments, the backward orthogonal deviation
    JEL: C12 C23
    Date: 2007–04
  4. By: Iolanda Lo Cascio (Queen Mary, University of London)
    Abstract: This paper surveys the techniques of wavelets analysis and the associated methods of denoising. The Discrete Wavelet Transform and its undecimated version, the Maximum Overlapping Discrete Wavelet Transform, are described. The methods of wavelets analysis can be used to show how the frequency content of the data varies with time. This allows us to pinpoint in time such events as major structural breaks. The sparse nature of the wavelets representation also facilitates the process of noise reduction by nonlinear <i>wavelet shrinkage</i>, which can be used to reveal the underlying trends in economic data. An application of these techniques to the UK real GDP (1873-2001) is described. The purpose of the analysis is to reveal the true structure of the data - including its local irregularities and abrupt changes - and the results are surprising.
    Keywords: Wavelets, Denoising, Structural breaks, Trend estimation
    JEL: C22 C14 C53
    Date: 2007–05
  5. By: Marta Ba?bura (ECARES, Université Libre de Bruxelles, Avenue Franklin D. Roosevelt 50, B-1050 Brussels, Belgium.); Gerhard Rünstler (Directorate General Research, European Central Bank, Kaiserstrasse 29, 60311 Frankfurt am Main, Germany.)
    Abstract: We derive forecast weights and uncertainty measures for assessing the role of individual series in a dynamic factor model (DFM) to forecast euro area GDP from monthly indicators. The use of the Kalman filter allows us to deal with publication lags when calculating the above measures. We find that surveys and financial data contain important information beyond the monthly real activity measures for the GDP forecasts. However, this is discovered only, if their more timely publication is properly taken into account. Differences in publication lags play a very important role and should be considered in forecast evaluation. JEL Classification: E37, C53.
    Keywords: Dynamic factor models, forecasting, filter weights.
    Date: 2007–05
  6. By: Ralf Becker; Adam Clements; James Curchin
    Abstract: Much research has addressed the relative performance of option implied volatilities and econometric model based forecasts in terms of forecasting asset return volatility. The general theme to come from this body of work is that implied volatility is a superior forecast. Some authors attribute this to the fact that option markets use a wider information set when forming their forecasts of volatility. This article considers this issue and determines whether S&P 500 implied volatility reflects a set of economic information beyond its impact on the prevailing level of volatility. It is found, that while the implied volatility subsumes this information, as do model based forecasts, this is only due to its impact on the current or prevailing level of volatility. Therefore, it appears as though implied volatility does not reflect a wider information set than model based forecasts, implying that implied volatility forecasts simply reflect volatility persistence in much the same way of as do econometric models.
    Keywords: Implied volatility, VIX, volatility forecasts, informational efficiency
    JEL: C12 C22 G00 G14
    Date: 2007–05–22

This nep-ets issue is ©2007 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.