We consider the problem of hypothesis testing in a modified version of the stochastic integration and cointegration framework of Harris, McCabe, and Leybourne (2002, Journal of Econometrics 111, 363–384). This nonlinear setup allows for volatility in excess of that catered for by the standard integration/cointegration paradigm through the introduction of nonstationary heteroskedasticity. We propose a test for stochastic cointegration against the alternative of no cointegration and a secondary test for stationary cointegration against the heteroskedastic alternative. Asymptotic distributions of these tests under their respective null hypotheses are derived, and consistency under their respective alternatives is established. Monte Carlo evidence suggests that the tests will perform well in practice. An empirical application to the term structure of interest rates is also given.We are most grateful to the Associate Editor and two anonymous referees for providing helpful comments on earlier versions of this paper.

We consider weak convergence of sample averages of nonlinearly transformed stochastic triangular arrays satisfying a functional invariance principle. A fundamental paradigm for such processes is constituted by integrated processes. The results obtained are extensions of recent work in the literature to the multivariate and non-Gaussian case. As admissible nonlinear transformation, a new class of functionals (so-called locally p-integrable functions) is introduced that adapts the concept of locally integrable functions in Pötscher (2004, Econometric Theory 20, 1–22) to the multidimensional setting.

The inverse of a noncentral Wishart matrix occurs in a variety of contexts in multivariate statistical work, including instrumental variable (IV) regression, but there has been very little work on its properties. In this paper, we first provide an expression for the expectation of the inverse of a noncentral Wishart matrix, and then go on to do the same for a number of scalar-valued functions of the inverse. The main result is obtained by exploiting simple but powerful group-equivariance properties of the expectation map involved. Subsequent results exploit the consequences of other invariance properties.

Despite their theoretical advantages, Integrated Conditional Moment (ICM) specification tests are not commonly employed in the econometrics practice. An important reason is that the employed test statistics are nonpivotal, and so critical values are not readily available. This article proposes an omnibus test in the spirit of the ICM tests of Bierens and Ploberger (1997, Econometrica 65, 1129–1151) where the test statistic is based on the minimized value of a quadratic function of the residuals of time series econometric models. The proposed test falls under the category of overidentification restriction tests started by Sargan (1958, Econometrica 26, 393–415). The corresponding projection interpretation leads us to propose a straightforward wild bootstrap procedure that requires only linear regressions to estimate the critical values irrespective of the model functional form. Hence, contrary to other existing ICM tests, the critical values are easily calculated while the test preserves the admissibility property of ICM tests.

A unified framework is proposed for tests of unobserved heterogeneity in parametric statistic models based on Neyman’s C(α) approach. Such tests are irregular in the sense that the first order derivative of the log likelihood with respect to the heterogeneity parameter is identically zero, and consequently the conventional Fisher information about the parameter is zero. Nevertheless, local asymptotic optimality of the C(α) tests can be established via LeCam’s differentiability in quadratic mean and the limit experiment approach. This leads to local alternatives of order n−1/4 . The scalar case result is already familiar from existing literature and we extend it to the multidimensional case. The new framework reveals that certain regularity conditions commonly employed in earlier developments are unnecessary, i.e. the symmetry or third moment condition imposed on the heterogeneity distribution. Additionally, the limit experiment for the multidimensional case suggests modifications on existing tests for slope heterogeneity in cross sectional and panel data models that lead to power improvement. Since the C(α) framework is not restricted to the parametric model and the test statistics do not depend on the particular choice of the heterogeneity distribution, it is useful for a broad range of applications for testing parametric heterogeneity.

Exact expressions are derived for the density function, variance, and kurtosis of a linear combination of the elements of a two-stage estimator for the coefficients in a single equation of a SUR system. The estimator is the first iterate in the iterative generalized least squares procedure described by Telser [14]. Our results generalize all previously known results for this estimator and, in certain special cases, also generalize some earlier exact results for Zellner's unrestricted covariance matrix estimator, to which it reduces in these special cases.

This note considers an asymptotic property of the class of closest moments estimators. Each such estimator is obtained by setting a vector of sample moments close to corresponding population moments. It is shown that each such estimator is asymptotically equivalent to a GMM estimator, which has a quadratic distance function. An implication of this result is that the estimator that is asymptotically efficient in the GMM class is also asymptotically efficient in the wider class of closest moment estimators.

Ridder (1990, Review of Economic Studies 57, 167–182) provides an identification result for the Generalized Accelerated Failure-Time (GAFT) model. We point out that Ridder’s proof of this result is incomplete, and provide an amended proof with an additional necessary and sufficient condition that requires that a function varies regularly at 0 and ∞. We also give more readily interpretable sufficient conditions on the tails of the error distribution or the asymptotic behavior of the transformation of the dependent variable. The sufficient conditions are shown to encompass all previous results on the identification of the Mixed Proportional Hazards (MPH) model. Thus, this paper not only clarifies, but also unifies the literature on the nonparametric identification of the GAFT and MPH models.

The present paper deals with the distributions related to the fractional Brownian motion (fBm). In particular, we try to compute the distributions of (ratios of) its quadratic functionals, not by simulations, but by numerically inverting the associated characteristic functions (c.f.s). Among them is the fractional unit root distribution. It turns out that the derivation of the c.f.s based on the standard approaches used for the ordinary Bm is inapplicable. Here the martingale approximation to the fBm suggested in the literature is used to compute an approximation to the distributions of such functionals. The associated c.f. is obtained via the Fredholm determinant. Comparison of the first two moments of the approximate with true distributions is made, and simulations are conducted to examine the performance of the approximation. We also find an interesting moment property of the approximate fractional unit root distribution, and a conjecture is given that the same property will hold for the true fractional unit root distribution.

This paper considers generalized method of moments (GMM) estimation of the inclusive panel AR(1) model that contains the covariance stationary panel AR(1) model and the panel AR(1) model with a unit root as special cases. The paper presents a two-step optimal linear GMM (OLGMM) estimator for the inclusive model that is asymptotically equivalent to the optimal nonlinear GMM estimator of Ahn and Schmidt (1997, Journal of Econometrics 76, 309–321) when the data are covariance stationary. Next the paper derives the asymptotic distribution of the OLGMM estimator when the model has a unit root under a variety of assumptions about the initial observations and the initial estimator. It is shown that in most cases the OLGMM estimator is superconsistent. In addition it is shown that the iterated OLGMM estimator is superefficient when the variance of the initial observations is finite and fixed, i.e., small compared to the cross-sectional dimension of the panel. The paper also conducts a Monte Carlo study in which the finite-sample properties of various GMM estimators for the inclusive panel AR(1) model are compared.I thank Steve Bond and Frank Windmeijer for kindly making one of their computer programs available to me. I also thank two anonymous referees and a co-editor for very helpful comments. This research was funded by the ESRC under grant R000239139.

Proxies for unobserved skills and technologies are increasingly available in empirical data. For dynamic discrete choice models of forward-looking agents where a continuous state variable is unobserved but its proxy is available, we derive closed-form identification of the structure by explicitly solving integral equations. In the first step, we derive closed-form identification of Markov components, including the conditional choice probabilities and the law of state transition. In the second step, we plug in these first-step identifying formulas to obtain primitive structural parameters of dynamically optimizing agents.

In this paper we develop a nonparametric estimation technique for semiparametric transformation models of the form: H (Y) = φ(Z) + X′β + U where H,φ are unknown functions, β is an unknown finite-dimensional parameter vector and the variables (Y,Z) are endogenous. Identification of the model and asymptotic properties of the estimator are analyzed under the mean independence assumption between the error term and the instruments. We show that the estimators are consistent, and a $\sqrt N$-convergence rate and asymptotic normality for $\hat \beta$ can be attained. The simulations demonstrate that our nonparametric estimates fit the data well.

Estimating a treatment’s effect on an outcome conditional on covariates is a primary goal of many empirical investigations. Accurate estimation of the treatment effect given covariates can enable the optimal treatment to be applied to each unit or guide the deployment of limited treatment resources for maximum program benefit. Applications of conditional treatment effect estimation are found in direct marketing, economic policy, and personalized medicine. When estimating conditional treatment effects, the typical practice is to select a statistical model or procedure based on sample data. However, combining estimates from the candidate procedures often provides a more accurate estimate than the selection of a single procedure. This article proposes a method of model combination that targets accurate estimation of the treatment effect conditional on covariates. We provide a risk bound for the resulting estimator under squared error loss and illustrate the method using data from a labor skills training program.

The fundamental contributions made by Paul Newbold have highlighted how crucial it is to detect when economic time series have unit roots. This paper explores the effects that model specification has on our ability to do that. Asymptotic power, a natural choice to quantify these effects, does not accurately predict finite-sample power. Instead, here the Kullback–Leibler divergence between the unit root null and any alternative is used and its numeric and analytic properties detailed. Numerically it behaves in a similar way to finite-sample power. However, because it is analytically available we are able to prove that it is a minimizable function of the degree of trending in any included deterministic component and of the correlation of the underlying innovations. It is explicitly confirmed, therefore, that it is approximately linear trends and negative unit root moving average innovations that minimize the efficacy of unit root inferential tools. Applied to the Nelson and Plosser macroeconomic series the effect that different types of trends included in the model have on unit root inference is clearly revealed.

Many regression models have two dimensions, say time (t = 1,…,T) and households (i = 1,…,N), as in panel data, error components, or spatial econometrics. In estimating such models we need to specify the structure of the error variance matrix Ω, which is of dimension T N × T N. If T N is large, then direct computation of the determinant and inverse of Ω is not practical. In this note we define structures of Ω that allow the computation of its determinant and inverse, only using matrices of orders T and N, and at the same time allowing for heteroskedasticity, for household- or station-specific autocorrelation, and for time-specific spatial correlation.

Recently, increasing interest in the issue of fractional cointegration has emerged from theoretical and empirical viewpoints. Here, as opposed to the traditional prescription of unit root observables with weak dependent cointegrating errors, the orders of integration of these series are allowed to take real values, but, as in the traditional framework, equality of the orders of at least two observable series is necessary for cointegration. This assumption, in view of the real-valued nature of these orders, could pose some difficulties, and in the present paper we explore some ideas related to this issue in a simple bivariate framework. First, in a situation of “near-cointegration,” where the only difference with respect to the “usual” fractional cointegration is that the orders of the two observable series differ in an asymptotically negligible way, we analyze properties of standard estimates of the cointegrating parameter. Second, we discuss the estimation of the cointegrating parameter in a situation where the orders of integration of the two observables are truly different but their corresponding balanced versions (with same order of integration) are cointegrated in the usual sense. A Monte Carlo study of finite-sample performance and simulated series is included.I thank Adrian Pagan, James Davidson, and seminar participants at the 2004 Econometric Society European Meeting and the 2004 Simposio de Análisis Económico for helpful comments. I also thank two referees and a co-editor whose comments led to improvements of the paper. This research was supported by the Spanish Ministerio de Educación y Ciencia through a contract Juan de la Cierva and ref. SEJ2005-07657/ECON, and also by the Universidad de Navarra, ref. 16037001.

We consider truncated (or conditional) sum of squares estimation of a parametric model composed of a fractional time series and an additive generalized polynomial trend. Both the memory parameter, which characterizes the behavior of the stochastic component of the model, and the exponent parameter, which drives the shape of the deterministic component, are considered not only unknown real numbers but also lying in arbitrarily large (but finite) intervals. Thus, our model captures different forms of nonstationarity and noninvertibility. As in related settings, the proof of consistency (which is a prerequisite for proving asymptotic normality) is challenging due to nonuniform convergence of the objective function over a large admissible parameter space, but, in addition, our framework is substantially more involved due to the competition between stochastic and deterministic components. We establish consistency and asymptotic normality under quite general circumstances, finding that results differ crucially depending on the relative strength of the deterministic and stochastic components. Finite-sample properties are illustrated by means of a Monte Carlo experiment.

In multilevel modeling of clustered survival data, to account for the differences among different clusters, a commonly used approach is to introduce cluster effects, either random or fixed, into the model. Modeling with random effects may lead to difficulties in the implementation of the estimation procedure for the unknown parameters of interest because the numerical computation of multiple integrals may become unavoidable when the cluster effects are not scalars. On the other hand, if fixed effects are used, there is a danger of having estimators with large variances because there are too many nuisance parameters involved in the model. In this article, using the idea of the homogeneity pursuit, we propose a new multilevel modeling approach for clustered survival data. The proposed modeling approach does not have the potential computational problem as modeling with random effects, and it also involves far fewer unknown parameters than modeling with fixed effects. We also establish asymptotic properties to show the advantages of the proposed model and conduct intensive simulation studies to demonstrate the performance of the proposed method. Finally, the proposed method is applied to analyze a dataset on the second-birth interval in Bangladesh. The most interesting finding is the impact of some important factors on the length of the second-birth interval variation over clusters and its homogeneous structure.