Search

This paper proposes an exponential class of dynamic binary choice panel data models for the analysis of short T (time dimension) large N (cross section dimension) panel data sets that allow for unobserved heterogeneity (fixed effects) to be arbitrarily correlated with the covariates. The paper derives moment conditions that are invariant to the fixed effects which are then used to identify and estimate the parameters of the model. Accordingly, generalized method of moments (GMM) estimators are proposed that are consistent and asymptotically normally distributed at the root-Nrate. We also study the conditional likelihood approach and show that under exponential specification, it can identify the effect of state dependence but not the effects of other covariates. Monte Carlo experiments show satisfactory finite sample performance for the proposed estimators and investigate their robustness to misspecification. [ABSTRACT FROM AUTHOR]

In this paper, we examine the issue of detecting explosive behavior in economic and financial time series when an explosive episode is both ongoing at the end of the sample and of finite length. We propose a testing strategy based on a subsampling method in which a suitable test statistic is calculated on a finite number of end-of-sample observations, with a critical value obtained using subsample test statistics calculated on the remaining observations. This approach also has the practical advantage that, by virtue of how the critical values are obtained, it can deliver tests which are robust to, among other things, conditional heteroskedasticity and serial correlation in the driving shocks. We also explore modifications of the raw statistics to account for unconditional heteroskedasticity using studentization and a White-type correction. We evaluate the finite sample size and power properties of our proposed procedures and find that they offer promising levels of power, suggesting the possibility for earlier detection of end-of-sample bubble episodes compared to existing procedures. [ABSTRACT FROM AUTHOR]

This paper presents results on the size and power of first generation panel unit root and stationarity tests obtained from a large scale simulation study. The tests developed in the following papers are included: Levin et al. (2002), Harris and Tzavalis (1999), Breitung (2000), Im et al. (19972003), Maddala and Wu (1999), Hadri (2000), and Hadri and Larsson (2005). Our simulation set-up is designed to address inter alia the following issues. First, we assess the performance as a function of the time and the cross-section dimensions. Second, we analyze the impact of serial correlation introduced by positive MA roots, known to have detrimental impact on time series unit root tests, on the performance. Third, we investigate the power of the panel unit root tests (and the size of the stationarity tests) for a variety of first order autoregressive coefficients. Fourth, we consider both of the two usual specifications of deterministic variables in the unit root literature. [ABSTRACT FROM AUTHOR]

In this paper we consider testing that an economic time series follows a martingale difference process. The martingale difference hypothesis has typically been tested using information contained in the second moments of a process, that is, using test statistics based on the sample autocovariances or periodograms. Tests based on these statistics are inconsistent since they cannot detect nonlinear alternatives. In this paper we consider tests that detect linear and nonlinear alternatives. Given that the asymptotic distributions of the considered tests statistics depend on the data generating process, we propose to implement the tests using a modified wild bootstrap procedure. The paper theoretically justifies the proposed tests and examines their finite sample behavior by means of Monte Carlo experiments. [ABSTRACT FROM AUTHOR]

The debate on whether macroeconomic series are trend or difference stationary, initiated by Nelson and Plosser [Nelson, C. R.; Plosser, C. I. (1982). Trends and random walks in macroeconomic time series: some evidence and implications. Journal of Monetary Economics 10:139–162] remains unresolved. The main objective of the paper is to contribute toward a resolution of this issue by bringing into the discussion the problem of statistical adequacy. The paper revisits the empirical results of Nelson and Plosser [Nelson, C. R.; Plosser, C. I. (1982). Trends and random walks in macroeconomic time series: some evidence and implications. Journal of Monetary Economics 10:139–162] and Perron [Perron, P. (1989). The great crash, the oil price shock, and the unit root hypothesis. Econometrica 57:1361–1401] and shows that several of their estimated models are misspecified. Respecification with a view to ensuring statistical adequacy gives rise to heteroskedastic AR(k) models for some of the price series. Based on estimated models which are statistically adequate, the main conclusion of the paper is that the majority of the data series are trend stationary. [ABSTRACT FROM AUTHOR]

In the two decades since the influential paper by researchers C.R. Nelson and C.I. Plosser documented that many macroeconomic time series appear to exhibit unit roots, a large literature has investigated their data, reaching a number of potentially conflicting conclusions about various characteristics such as unit roots, structural change and linearity. For example, researcher P. Perron noted that the conclusion of difference stationarity may have been due in part to failure to account for structural change and illustrated that using a model allowing for structural change, the null of a unit root was rejected in favor of trend stationarity for 10 of the original 13 Nelson-Plosser series. Many more articles related to the difference versus trend stationarity debate have been written, most either applying existing techniques to new datasets or, like those previously mentioned, taking issue with an earlier maintained assumption to develop a modified specification in the hope of obtaining a more clear solution. In fact, what the large body of literature and the continuing debate mainly illustrates is a truth so fundamental that it is perhaps often forgotten, that inference depends crucially on model specification and the underlying assumptions inherent in the chosen model.

The article focuses on statistical adequacy and the testing of trend versus difference stationarity. Researchers E. Andreou and A. Spanos (A & S), hereafter, have written an important and inspiring paper about the role of statistical adequacy and model specification when testing for unit-roots in economic time series. Their thesis is that the statistical adequacy of the univariate model used for testing unit roots should be scrutinized and that the model needs to be respecified if any departures the underlying assumptions are detected. As a blueprint for practice A & S approach shares some important elements with econometric methodologies, which stress the importance of congruency of models, and which advocate general-to-specific modeling. In fact, one way of reading the paper is as a coherent argument for applying the same standards of model evaluation when testing for unit roots, as the prudent researcher is advised to adopt for econometric modelling. In the first section, the author gives an example of the practical relevance of one of A & S main points, namely that unless statistical adequacy is established, standard unit root tests can give rise to wrong conclusions about a variable's temporal properties.

This paper reviews and extends the literature on the finite sample behavior of tests for sample selection bias. Monte Carlo results show that, when the “multicollinearity problem” identified by Nawata (1993) is severe, (i) the t -test based on the Heckman–Greene variance estimator can be unreliable, (ii) the Likelihood Ratio test remains powerful, and (iii) nonnormality can be interpreted as severe sample selection bias by Maximum Likelihood methods, leading to negative Wald statistics. We also confirm previous findings (Leung and Yu, 1996) that the standard regression-based t -test (Heckman, 1979) and the asymptotically efficient Lagrange Multiplier test (Melino, 1982), are robust to nonnormality but have very little power. [ABSTRACT FROM AUTHOR]

In this paper we discuss different aspects of long memory behavior and applicable parametric models. We discuss the confusion that can arise when the empirical autocorrelation function decreases in a hyperbolic way. [ABSTRACT FROM AUTHOR]

It is widely known that significant in-sample evidence of predictability does not guarantee significant out-of-sample predictability. This is often interpreted as an indication that in-sample evidence is likely to be spurious and should be discounted. In this paper, we question this interpretation. Our analysis shows that neither data mining nor dynamic misspecification of the model under the null nor unmodelled structural change under the null are plausible explanations of the observed tendency of in-sample tests to reject the no-predictability null more often than out-of-sample tests. We provide an alternative explanation based on the higher power of in-sample tests of predictability in many situations. We conclude that results of in-sample tests of predictability will typically be more credible than results of out-of-sample tests. [ABSTRACT FROM AUTHOR]

This paper reconsiders the nonlinearity test proposed by Kocenda (Kocenda, E. (2001). An alternative to the BDS test: integration across the correlation integral. Econometric Reviews 20:337–351). When the analyzed series is non-Gaussian, the empirical rejection rates can be much larger than the nominal size. In this context, the necessity of tabulating the empirical distribution of the statistic each time the test is computed is stressed. To that end, simple random permutation works reasonably well. This paper also shows, through Monte Carlo experiments, that Kocenda's test can be more powerful than the Brock et al. (Brock, W., Dechert, D., Scheickman, J., LeBaron, B. (1996). A test for independence based on the correlation dimension. Econometric Reviews 15:197–235) procedure. However, more than one range of values for the proximity parameter should be used. Finally, empirical evidence on exchange rates is reassessed. [ABSTRACT FROM AUTHOR]

Researchers E. Andreou and A. Spanos provide a further analysis of the unit root versus trend stationary debate using the data set of researchers C.R. Nelson and C.I. Plosser. They present an interesting and novel approach to the problem. The main argument of the paper is that any such analysis should be preceded by a statistical adequacy analysis. As they state, "By statistical adequacy we mean the empirical validity of the probabilistic assumptions underlying the estimated statistical model. Establishing statistical adequacy provides a sufficient condition for statistical inference because the latter presupposes the validity of the model specification." The criteria adopted to assess statistical adequacy are: normality of the process; linearity; homoskedaticity; t-invariance of the parameters; and errors are an innovation process. Conclusions are basically the same as those of researcher P. Perron except for 3 series: GNP deflator; velocity; and stock prices. All rejections, except for the unemployment rate of course, are obtained introducing breaks in the trend function as in Perron showing the importance of the feature in the data.

This paper adresses the measurement of technical efficiency of textile, clothing, and leather (TCL) industries in Tunisia through a panel data estimation of a dynamic translog production frontier. It provides a perspective on productivity and efficiency that should be instructive to a developing economy which will face substantial competitive pressure along the gradual economic liberalisation process. The importance of TCL industries in Tunisian manufacturing sector is a reason for obtaining more knowledge of productivity and efficiency for this key industry. Dynamic is introduced to reflect the production consequences of the adjustment costs, which are associated with changes in factor inputs. Estimation of a dynamic error components model is considered using the system generalized method of moments (GMM) estimator suggested by Arellano and Bover (1995). Another look at the instrumental–variable estimation of error–components models, J. Econometrics 68:29–51) and Blundell and Bond (Blundell, R., Bond, S. (1998a), Initial conditions and moment restrictions in dynamic panel data models. J. Econometrics 87:115–143; Blundell, R., Bond, S. (1998b), GMM estimation with persistent panel data: an application to production functions, Paper presented at the Eighth International Conference on Panel Data, Goteborg University). Our study evaluates the sensitivity of the results, particularly of the efficiency measures, to different specifications. Firm-specific time-invariant technical efficiency is obtained using the Schmidt and Sickles (Schmidt, P., Sickles, R. C. (1984). Production frontiers and panel data. J. Bus. Econ. Stat. 2:367–374) approach after estimating the dynamic frontier. We stress the importance of allowing for lags in adjustment of output to inputs and of controlling for time-invariant variables when estimating firm-specific efficiency. The results suggest that the system GMM estimation of the dynamic specification... [ABSTRACT FROM AUTHOR]

This paper adopts a unified approach to the derivation of the asymptotic distributions of various seasonal unit root tests. The procedures considered are those of Dickey et al. [DHF], Kunst, Hylleberg et al. [HEGY], Osborn et al. [OCSB], Ghysels et al. [GHL] and Franses. This unified approach shows that the asymptotic distributions of all these test statistics are functions of the same vector of Brownian motions. The Kunst test and the overall HEGY F-test are, indeed, equivalent both asymptotically and in finite samples, while the Franses and GHL tests are shown to have equivalent parameterizations. The OCSB and DHF test regressions are viewed as restricted forms of the Kunst-HEGY regressions, and these restrictions may have non-trivial asymptotic implications. [ABSTRACT FROM AUTHOR]

This paper provides a means of accurately simulating explosive autoregressive processes and uses this method to analyze the distribution of the likelihood ratio test statistic for an explosive second-order autoregressive process of a unit root. While the standard Dickey-Fuller distribution is known to apply in this case, simulations of statistics in the explosive region are beset by the magnitude of the numbers involved, which cause numerical inaccuracies. This has previously constituted a bar on supporting asymptotic results by means of simulation, and analyzing the finite sample properties of tests in the explosive region. [ABSTRACT FROM AUTHOR]

In this paper we compare Bartlett-corrected, bootstrap, and fast double bootstrap tests on maximum likelihood estimates of cointegration parameters. The key result is that both the bootstrap and the Bartlett-corrected tests must be based on the unrestricted estimates of the cointegrating vectors: procedures based on the restricted estimates have almost no power. The small sample size bias of the asymptotic test appears so severe as to advise strongly against its use with the sample sizes commonly available; the fast double bootstrap test minimizes size bias, while the Bartlett-corrected test is somehow more powerful. [ABSTRACT FROM AUTHOR]

This paper generalizes Nagar's (1959) approximation to the finite sample mean squared error (MSE) of the instrumental variables (IV) estimator to the case in which the errors possess an elliptical distribution whose moments exist up to infinite order. This allows for types of excess kurtosis exhibited by some financial data series. This approximation is compared numerically to Knight's (1985) formulae for the exact moments of the IV estimator under nonnormality. We use the results to explore two questions on instrument selection. First, we complement Buse's (1992) analysis by considering the impact of additional instruments on both bias and MSE. Second, we evaluate the properties of Andrews's (1999) selection method in terms of the bias and MSE of the resulting IV estimator. [ABSTRACT FROM AUTHOR]

This paper studies the behavior of the HEGY statistics for quarterly data, for seasonal autoregressive unit roots, when the analyzed time series is deterministic seasonal stationary but exhibits a change in the seasonal pattern. We analyze also the HEGY test for the nonseasonal unit root. Our results show that when the break magnitudes are finite, the HEGY test statistics are not asymptotically biased toward the nonrejection of the seasonal and nonseasonal unitroot hypotheses. However, the finite sample power properties may be substantially affected, the behavior of the tests depending on the type of the break. [ABSTRACT FROM AUTHOR]

This paper develops optimal tests based on sequential predictive moment conditions. We show that an appropriate weighting version of the predictive test achieves the same power as optimal structural change tests proposed by Sowell (1996a) Optimal tests for parameter instability in the generalized method of moments framework. Econometrica 64:1085–1107 and (1996b) Tests for Violations of MOMENT conditions. Manuscript. Graduate School of Industrial Administration, Carnegie Mellon University. Consequently, we can apply directly Sowell's results. Optimal predictive tests for parameter instability and overidentifying restriction stability are proposed. The finite sample properties of LM, Wald, LR-type and predictive tests for parameter instability are studied via a simulation study. [ABSTRACT FROM AUTHOR]

The aim of this paper is to compare the relative performance of several tests for the null hypothesis of cointegration, in terms of size and power in finite samples. This is carried out using Monte Carlo simulations for a range of plausible data-generating processes. We also analyze the impact on size and power of choosing different procedures to estimate the long run variance of the errors. We found that the parametrically adjusted test of McCabe et al. (1997) is the most well-balanced test, displaying good power and relatively few size distortions. [ABSTRACT FROM AUTHOR]

The present paper focuses attention on the sensitivity of technical inefficiency to most commonly used one-sided distributions of the inefficiency error term, namely the truncated normal, the half-normal, and the exponential distributions. A generalized version of the half-normal, which does not embody the zero-mean restriction, is also explored. For each distribution, the likelihood function and the counterpart of the estimator of technical efficiency are explicitly stated (Jondrow, J., Lovell, C. A. K., Materov, I. S., Schmidt, P. (1982), On estimation of technical inefficiency in the stochastic frontier production function model, J. Econometrics 19:233–238). Based on our panel data set, related to Tunisian manufacturing firms over the period 1983–1993, formal tests lead to a strong rejection of the zero-mean restriction embodied in the half normal distribution. Our main conclusion is that the degree of measured inefficiency is very sensitive to the postulated assumptions about the distribution of the one-sided error term. The estimated inefficiency indices are, however, unaffected by the choice of the functional form for the production function. [ABSTRACT FROM AUTHOR]

This paper constructs a consistent model specification test based on the difference between the nonparametric kernel sum of squares of residuals and the sum of squares of residuals from a parametric null model. We establish the asymptotic normality of the proposed test statistic under the null hypothesis of correct parametric specification and show that the wild bootstrap method can be used to approximate the null distribution of the test statistic. Results from a small simulation study are reported to examine the finite sample performance of the proposed tests. [ABSTRACT FROM AUTHOR]

Using the augmented Dickey-Fuller test to verify the existence of a unit root in an autoregressive process often requires the correctly specified intercept, since the test statistics can be distinctive under different model specifications and lead to contradictory results at times. In this article, we develop a unified inference that not only unifies the specifications of the intercept but also accommodates different degrees of persistence of the underlying process and heteroscedastic errors. A simulation study shows that the resulting unified unit root test exhibits excellent size control and reasonably good power. In an empirical application, we implement the proposed test to re-examine the presence of unit roots within eleven widely used variables in stock return predictability. [ABSTRACT FROM AUTHOR]

The present article discusses alternative regression models and estimation methods for dealing with multivariate fractional response variables. Both conditional mean models, estimable by quasi-maximum likelihood, and fully parametric models (Dirichlet and Dirichlet-multinomial), estimable by maximum likelihood, are considered. A new parameterization is proposed for the parametric models, which accommodates the most common specifications for the conditional mean (e.g., multinomial logit, nested logit, random parameters logit, dogit). The text also discusses at some length the specification analysis of fractional regression models, proposing several tests that can be performed through artificial regressions. Finally, an extensive Monte Carlo study evaluates the finite sample properties of most of the estimators and tests considered. [ABSTRACT FROM PUBLISHER]

Based on Fieller's method for the estimation of a confidence set for a ratio, I construct a polar plot of the test statistics for all angles associated with the ratio. This polar confidence plot clarifies and systematizes the inherent properties of the confidence set for ratios and, in particular, determines how the confidence set may be uninformative or disconnected. [ABSTRACT FROM AUTHOR]

In economics, common factors are often assumed to underlie the co-movements of a set of macroeconomic variables. For this reason, many authors have used estimated factors in the construction of prediction models. In this article, we begin by surveying the extant literature on diffusion indexes. We then outline a number of approaches to the selection of factor proxies (observed variables that proxy unobserved estimated factors) using the statistics developed in Bai and Ng (2006a,b). Our approach to factor proxy selection is examined via a small Monte Carlo experiment, where evidence supporting our proposed methodology is presented, and via a large set of prediction experiments using the panel dataset of Stock and Watson (2005). One of our main empirical findings is that our “smoothed” approaches to factor proxy selection appear to yield predictions that are often superior not only to a benchmark factor model, but also to simple linear time series models which are generally difficult to beat in forecasting competitions. In some sense, by using our approach to predictive factor proxy selection, one is able to open up the “black box” often associated with factor analysis, and to identify actual variables that can serve as primitive building blocks for (prediction) models of a host of macroeconomic variables, and that can also serve as policy instruments, for example. Our findings suggest that important observable variables include various S&P500 variables, including stock price indices and dividend series; a 1-year Treasury bond rate; various housing activity variables; industrial production; and exchange rates. [ABSTRACT FROM AUTHOR]

Røed et al. (1999) demonstrate that the standard result of known negative duration bias does not necessarily hold in a two-state mixed proportional hazard (MPH) model. We show that the duration bias is still ambiguous in a MPH model with a multivariate gamma distribution. A discrete time two-state version of our MPH model is developed to analyze the duration of higher education. The estimation results show that we cannot reject the hypothesis that the two unobserved heterogeneity variables are uncorrelated. Accepting this hypothesis implies that the standard result holds in our analysis. [ABSTRACT FROM AUTHOR]

In this article, a multivariate threshold varying conditional correlation (TVCC) model is proposed. The model extends the idea of Engle (2002) and Tse and Tsui (2002) to a threshold framework. This model retains the interpretation of the univariate threshold GARCH model and allows for dynamic conditional correlations. Techniques of model identification, estimation, and model checking are developed. Some simulation results are reported on the finite sample distribution of the maximum likelihood estimate of the TVCC model. Real examples demonstrate the asymmetric behavior of the mean and the variance in financial time series and the ability of the TVCC model to capture these phenomena. [ABSTRACT FROM AUTHOR]

We describe estimation, learning, and prediction in a treatment-response model with two outcomes. The introduction of potential outcomes in this model introduces four cross-regime correlation parameters that are not contained in the likelihood for the observed data and thus are not identified. Despite this inescapable identification problem, we build upon the results of Koop and Poirier (1997) to describe how learning takes place about the four nonidentified correlations through the imposed positive definiteness of the covariance matrix. We then derive bivariate distributions associated with commonly estimated “treatment parameters” (including the Average Treatment Effect and effect of Treatment on the Treated), and use the learning that takes place about the nonidentified correlations to calculate these densities. We illustrate our points in several generated data experiments and apply our methods to estimate the joint impact of child labor on achievement scores in language and mathematics. [ABSTRACT FROM AUTHOR]

Three classes of structural change tests (or tests for parameter instability) that have been receiving much attention in both the statistics and the econometrics communities but have been developed in rather loosely connected lines of research are unified by embedding them into the framework of generalized M -fluctuation tests (Zeileis and Hornik, 2003). These classes are tests based on maximum likelihood scores (including the Nyblom–Hansen test), on F statistics (sup F , ave F , exp F tests), and on OLS residuals (OLS-based CUSUM and MOSUM tests). We show that (representatives from) these classes are special cases of the generalized M -fluctuation tests, based on the same functional central limit theorem but employing different functionals for capturing excessive fluctuations. After embedding these tests into the same framework and thus understanding the relationship between these procedures for testing in historical samples, it is shown how the tests can also be extended to a monitoring situation. This is achieved by establishing a general M -fluctuation monitoring procedure and then applying the different functionals corresponding to monitoring with ML scores, F statistics, and OLS residuals. In particular, an extension of the sup F test to a monitoring scenario is suggested and illustrated on a real-world data set. [ABSTRACT FROM AUTHOR]

We compare the finite sample performance of a range of tests of linear restrictions for linear panel data models estimated using the generalized method of moments (GMM). These include standard asymptotic Wald tests based on one-step and two-step GMM estimators; two bootstrapped versions of these Wald tests; a version of the two-step Wald test that uses a finite sample corrected estimate of the variance of the two-step GMM estimator; the LM test; and three criterion-based tests that have recently been proposed. We consider both the AR (1 ) panel model and a design with predetermined regressors. The corrected two-step Wald test performs similarly to the standard one-step Wald test, whilst the bootstrapped one-step Wald test, the LM test, and a simple criterion-difference test can provide more reliable finite sample inference in some cases. [ABSTRACT FROM AUTHOR]

We review the different block bootstrap methods for time series, and present them in a unified framework. We then revisit a recent result of Lahiri [Lahiri, S. N. (1999b). Theoretical comparisons of block bootstrap methods, Ann. Statist. 27:386–404] comparing the different methods and give a corrected bound on their asymptotic relative efficiency; we also introduce a new notion of finite-sample “attainable” relative efficiency. Finally, based on the notion of spectral estimation via the flat-top lag-windows of Politis and Romano [Politis, D. N., Romano, J. P. (1995). Bias-corrected nonparametric spectral estimation. J. Time Series Anal. 16:67–103], we propose practically useful estimators of the optimal block size for the aforementioned block bootstrap methods. Our estimators are characterized by the fastest possible rate of convergence which is adaptive on the strength of the correlation of the time series as measured by the correlogram. [ABSTRACT FROM AUTHOR]

Estimation of a cross-sectional spatial model containing both a spatial lag of the dependent variable and spatially autoregressive disturbances are considered. Kelejian and Prucha (1998) described a generalized two-stage least squares procedure for estimating such a spatial model. Their estimator is, however, not asymptotically optimal. We propose best spatial 2SLS estimators that are asymptotically optimal instrumental variable (IV) estimators. An associated goodness-of-fit (or over identification) test is available. We suggest computationally simple and tractable numerical procedures for constructing the optimal instruments. [ABSTRACT FROM AUTHOR]