|
on Econometric Time Series |
| By: | Yongmiao Hong (Cornell University); Yoon-Jin Lee (Indiana University Bloomington) |
| Abstract: | We propose a new class of specification tests for Autoregressive Conditional Duration (ACD) models. Both linear and nonlinear ACD models are covered, and standardized innovations can have time-varying conditional dispersion and higher order conditional moments of unknown form. No specific estimation method is required, and the tests have a convenient null asymptotic N(0,1) distribution. To reduce the impact of parameter estimation uncertainty in finite samples, we adopt Wooldridge's (1990a) device to our context and justify its validity. Simulation studies show that the finite sample correction gives better sizes in finite samples and are robust to parameter estimation uncertainty. And, it is important to take into account time-varying conditional dispersion and higher order conditional moments in standardized innovations; failure to do so can cause strong overrejection of a correctly specified ACD model. The proposed tests have reasonable power against a variety of popular linear and nonlinear ACD alternatives. |
| Keywords: | Autoregressive Conditional Duration, Dispersion Clustering, Finite Sample Correction, Generalized Spectral Derivative, Nonlinear Time Series, Parameter Estimation Uncertainty, Wooldridge's Device |
| JEL: | C4 C2 |
| Date: | 2007–09 |
| URL: | http://d.repec.org/n?u=RePEc:inu:caeprp:2007019&r=ets |
| By: | João Victor Issler (EPGE/FGV); Luiz Renato Lima |
| Date: | 2007–09 |
| URL: | http://d.repec.org/n?u=RePEc:fgv:epgewp:650&r=ets |
| By: | Todd E. Clark; Michael W. McCracken |
| Abstract: | Motivated by the common finding that linear autoregressive models forecast better than models that incorporate additional information, this paper presents analytical, Monte Carlo, and empirical evidence on the effectiveness of combining forecasts from nested models. In our analytics, the unrestricted model is true, but as the sample size grows, the data generating process converges to the restricted model. This approach captures the practical reality that the predictive content of variables of interest is often low. We derive MSE-minimizing weights for combining the restricted and unrestricted forecasts. Monte Carlo and empirical analyses verify the practical effectiveness of our combination approach. |
| Date: | 2007 |
| URL: | http://d.repec.org/n?u=RePEc:fip:fedgfe:2007-43&r=ets |
| By: | Katrin Assenmacher-Wesche (Swiss National Bank); M. Hashem Pesaran (Cambridge University, CIMF USC and IZA) |
| Abstract: | We investigate the effect of forecast uncertainty in a cointegrating vector error correction model for Switzerland. Forecast uncertainty is evaluated in three different dimensions. First, we investigate the effect on forecasting performance of averaging over forecasts from different models. Second, we look at different estimation windows. We find that averaging over estimation windows is at least as effective as averaging over different models and both complement each other. Third, we explore whether using weighting schemes from the machine learning literature improves the average forecast. Compared to equal weights the effect of the weighting scheme on forecast accuracy is small in our application. |
| Keywords: | Bayesian model averaging, choice of observation window, long-run structural vector autoregression |
| JEL: | C53 C32 |
| Date: | 2007–09 |
| URL: | http://d.repec.org/n?u=RePEc:iza:izadps:dp3071&r=ets |
| By: | Costantini, Mauro; Lupi, Claudio; Popp, Stephan |
| Abstract: | In this paper we propose the extension of the covariate-augmented Dickey Fuller (CADF) test for unit roots developed by Hansen (1995} to the panel case. We show that the extension is viable and gives power gains with respect to the time series approach. Particular attention is paid to cross-unit dependence. |
| Keywords: | Unit root, Panel data, Cross-unit dependence |
| JEL: | C12 C22 C23 |
| Date: | 2007–09–10 |
| URL: | http://d.repec.org/n?u=RePEc:mol:ecsdps:esdp07039&r=ets |
| By: | Luís Francisco Aguiar-Conraria (Universidade do Minho - NIPE); Maria Joana Soares (Universidade do Minho - Departamento de Matemática) |
| Abstract: | A large body of empirical literature has suggested that oil price shocks have an important effect on economic activity. But in most of the literature the analysis is exclusively done in the time domain. However, interesting relations exist at different frequencies. We use (cross) wavelet analysis to uncover some of these relations, estimating the spectral characteristics of the time-series as a function of time. Our analysis suggests that the volatility of both the inflation rate and the output growth rate started to decrease in the decades of 1950 and 1960, suggesting that the great moderation started then,but that it was temporarily interrupted due to the oils crisis of the 1970s, whose effects extend until the mid 1980s. We also show that while at business cycle frequencies oil prices lead industrial production, in the very long run production increases lead oil price increases. The exception to this long-run relation occurred between the mid 1970s and mid 1980s. Our analysis also suggests that monetary policy became much more eficient after 1980 to deal with the inflationary pressures of oil shocks. |
| Keywords: | Business cycles, time-frequency analysis, non-stationary time series, wavelets, cross wavelets, wavelet coherency. |
| JEL: | C10 E32 |
| Date: | 2007 |
| URL: | http://d.repec.org/n?u=RePEc:nip:nipewp:16/2007&r=ets |
| By: | Luís Francisco Aguiar-Conraria (Universidade do Minho - NIPE); Maria Joana Soares (Universidade do Minho - Departamento de Matemática); Nuno Azevedo (Universidade do Porto - Faculdade de Ciências) |
| Abstract: | Economic agents simultaneously operate at different horizons. Many economic processes are the result of the actions of several agents with different term objectives. Therefore, economic time-series is a combination of components operating on different frequencies. Several questions about the data are connected to the understanding of the time-series behavior at different frequencies. While Fourier analysis is not appropriate to study the cyclical nature of economic time-series, because these are rarely stationary, wavelet analysis performs the estimation of the spectral characteristics of a time-series as a function of time. In spite of all its advantages, wavelets are hardly ever used in economics. The purpose of this paper is to show that cross wavelet analysis can be used to directly study the interactions different time-series in the time-frequency domain. We use wavelets to analyze the impact of interest rate price changes on some macroeconomic variables: Industrial Production, Inflation and the monetary aggregates M1 and M2. Specifically, three tools are utilized: the wavelet power spectrum, wavelet coherency and wavelet phase-difference. These instruments illustrate how the use of wavelets may help to unravel economic time-frequency relations that would otherwise remain hidden. |
| Keywords: | Monetary policy, time-frequency analysis, non-stationary time series, wavelets, cross wavelets, wavelet coherency. |
| Date: | 2007 |
| URL: | http://d.repec.org/n?u=RePEc:nip:nipewp:17/2007&r=ets |
| By: | Lucia Alessi; Matteo Barigozzi; Marco Capasso |
| Abstract: | We review, under a historical perspective, the developement of the problem of non- fundamentalness of Moving Average (MA) representations of economic models, starting from the work by Hansen and Sargent [1980]. Nonfundamentalness typically arises when agents' information space is larger than the econometrican's one. Therefore it is impos- sible for the latter to use standard econometric techniques, as Vector AutoRegression (VAR), to estimate economic models. We re-state the conditions under which it is pos- sible to invert an MA representation in order to get an ordinary VAR, and we consider how the latter is used in the literature to assess the validity of Dynamic Stochastic Gen- eral Equilibrium models, providing some interesting examples. We believe that possible nonfundamental representations of considered models are too often neglected in the liter- ature. We consider how factor models can be seen as an alternative to VAR for assessing the validity of an economic model without having to deal with the problem of nonfun- damentalness. We then review the works by Lippi and Reichlin [1993] and Lippi and Reichlin [1994] which are the first attempts to give to nonfundamental representations the economic relevance that they deserve, and to outline a method to obtain such repre- sentations starting from an estimated VAR. |
| Keywords: | Nonfundamentalness, Structural VAR, Dynamic Stochastic General Equilibrium Models, Factor Models |
| Date: | 2007–10–01 |
| URL: | http://d.repec.org/n?u=RePEc:ssa:lemwps:2007/22&r=ets |
| By: | Carlos Santos (Faculdade de Economia e Gestão, Universidade Católica Portuguesa (Porto)) |
| Abstract: | A two-stage procedure based on impulse saturation is suggested to distinguish mean and variance shifts. The resulting zero-mean innovation test statistic has a non standard distribution, with a nuisance parameter. Hence, simulation-based critical values are provided for some cases of interest. Monte Carlo evidence reveals the test has good power properties to discriminate mean and variance shifts identified through the impulse saturation break test. |
| Keywords: | breaks; mean shift; variance shift; impulse saturation; nuisance parameter |
| JEL: | C12 C15 C22 C52 |
| Date: | 2007–08 |
| URL: | http://d.repec.org/n?u=RePEc:cap:wpaper:142007&r=ets |