We provide new asymptotic theory for kernel density estimators, when these are applied to autoregressive processes exhibiting moderate deviations from a unit root. This fills a gap in the existing literature, which has to date considered only nearly integrated and stationary autoregressive processes. These results have applications to nonparametric predictive regression models. In particular, we show that the null rejection probability of a nonparametric t test is controlled uniformly in the degree of persistence of the regressor. This provides a rigorous justification for the validity of the usual nonparametric inferential procedures, even in cases where regressors may be highly persistent.

There are many environments in econometrics which require nonseparable modeling of a structural disturbance. In a nonseparable model with endogenous regressors, key conditions are validity of instrumental variables and monotonicity of the model in a scalar unobservable variable. Under these conditions the nonseparable model is equivalent to an instrumental quantile regression model. A failure of the key conditions, however, makes instrumental quantile regression potentially inconsistent. This article develops a methodology for testing the hypothesis whether the instrumental quantile regression model is correctly specified. Our test statistic is asymptotically normally distributed under correct specification and consistent against any alternative model. In addition, test statistics to justify the model simplification are established. Finite sample properties are examined in a Monte Carlo study and an empirical illustration is provided.

It is well known that when investors evaluate risk or opportunity, they often depart from predictions of expected utility. In addition, for both academic and financial communities it is a familiar stylized fact that stock return distributions are not normal. Both empirical evidence and experimental evidence indicate that distributional information of asset returns has an important impact on investors. In this paper, we argue that the right-tail distributional information of returns can provide very valuable information to investors and portfolio managers, and right-tail information should be used together with other (say, left-tail) information in analyzing financial markets. Here, we introduce measures for the right-tail distribution. Quantile regression estimators for the right-tail measures are proposed, and their asymptotic properties are developed. Statistical inference on testing for changes of right-tail distribution is also discussed. A Monte Carlo experiment is conducted to evaluate the performance of the proposed estimator. The proposed estimation method may also be applied to estimation of other measures in finance.

A size correction to adjust the critical values of asymptotic χ2 tests in a SUR context has been developed by Rothenberg and verified by Phillips. The adjustment centers the distribution of the observed test statistic so that it coincides more closely with its hypothetical distribution. The correction itself is an approximation to O(T−1), and while the adjustment should be superior to the asymptotic test (no adjustment) in small-sample situations, its performance can be expected to deteriorate as the sample size decreases. The Edgeworth corrections, using formulas provided by Phillips, are calculated for the Wald, likelihood ratio, and Lagrange multiplier tests in the context of symmetry testing in demand analysis using the Laitinen–Meisner simulated data set. The sample consists of 31 observations and the corrections were “adequate to useful” in the context of 5, 8, and 11 equations. However, in the extreme case of 14 equations the corrections typically only took up 60% of the adjustment required. The computational cost of calculating the Edgeworth correction for large SUR systems with a large number of restrictions also turns out to be quite heavy. The conclusion reached is that while Edgeworth corrections are not totally satisfactory, they are easy to include in a program and provide a useful critical value correction.

We develop the limit theory of the quantilogram and cross-quantilogram under long memory. We establish the sub-root-n central limit theorems for quantilograms that depend on nuisance parameters. We propose a moving block bootstrap (MBB) procedure for inference and establish its consistency, thereby enabling a consistent confidence interval construction for the quantilograms. The newly developed reduction principles for the quantilograms serve as the main technical devices used to derive the asymptotics and establish the validity of MBB. We report some simulation evidence that our methods work satisfactorily. We apply our method to quantile predictive relations between financial returns and long-memory predictors.

This article proposes the use of spectral methods to pool cross-sectional replications (N) of time series data (T) for time series analysis. Spectral representations readily suggest a weighting scheme to pool the data. The asymptotically desirable properties of the resulting estimators seem to translate satisfactorily into samples as small as T = 25 with N = 5. Simulation results, Monte Carlo results, and an empirical example help confirm this finding. The article concludes that there are many empirical situations where spectral methods canbe used where they were previously eschewed.

This paper extends the Andrews (2002, Econometrica 71, 1661–1694) and Andrews and Kim (2006, Journal of Business & Economic Statistics 24, 379–394) ordinary least squares–based end-of-sample instability tests for linear regression models. The author proposes to quasi-difference the data first using a consistent estimate of the sum of the autoregressive coefficients of the error process and then test for the end-of-sample instability. For the cointegration model, the feasible quasi-generalized least squares (FQGLS) version of the Andrews and Kim (2006) P test is considered and is shown, by simulations, to be more robust to serial correlation in the error process and to have power no less than Andrews and Kim’s original test. For the linear time trend model, the FQGLS version of the Andrews (2002) S test is considered with the error process allowed to be nonstationary up to one unit root, and the new test is shown to be robust to potentially nonstationary serial correlation. A simulation study also shows that the finite-sample properties of the proposed test can be further improved when the Andrews (1993, Econometrica 61,139–165) or Andrews and Chen (1994, Journal of Business & Economic Statistics 12, 187–204) median unbiased estimate of the sum of the autoregressive coefficients is used.

A common test in econometrics is the Dickey–Fuller test, which is based on the test statistic . We investigate the behavior of the test statistic if the data yt are given by an exponential random walk exp(Zt) where Zt = Zt−1 + σεt and the εt are independent and identically distributed random variables. The test statistic DF(T) is a nonlinear transformation of the partial sums of εt process. Under certain moment conditions on the εt we show that tends to one as λ → 0. For the particular case that the εt define a simple random walk it is shown that plimT→∞DF(T)/T exists and the limit is evaluated. The theoretical results are illustrated by some simulation experiments.We gratefully acknowledge the help of an anonymous referee whose comments on the first two versions of this paper enabled us to reduce the number of mistakes and to increase the clarity of presentation. The authors' research was supported in part by Sonderforschungsbereich 475, University of Dortmund.

Long-run restrictions are a very popular method for identifying structural vector autoregressions, but they suffer from weak identification when the data is very persistent, i.e., when the highest autoregressive roots are near unity. Near unit roots introduce additional nuisance parameters and make standard weak-instrument-robust methods of inference inapplicable. We develop a method of inference that is robust to both weak identification and strong persistence. The method is based on a combination of the Anderson-Rubin test with instruments derived by filtering potentially nonstationary variables to make them near stationary using the IVX instrumentation method of Magdalinos and Phillips (2009). We apply our method to obtain robust confidence bands on impulse responses in two leading applications in the literature.

We derive conditions for the identification of the structural parameters in Markov decision model under the assumptions of Rust (1987, Econometrica 55, 999–1033) when the payoff function is parametrically specified. Identification in this class of dynamic problems is difficult to establish since the parameters of interest enter the value function nonlinearly, and the value function is only defined implicitly as a fixed point of some functional equation. We show it is sufficient to verify identification in the pseudomodel, which is more tractable as it is originally designed to reduce the computational burden in the estimation problem, for the identification of the data generating parameter of the underling model. Our results extend naturally to a class of dynamic discrete action games commonly used in empirical industrial organizations.

We introduce a new approach for comparing the predictive accuracy of two nested models that bypasses the difficulties caused by the degeneracy of the asymptotic variance of forecast error loss differentials used in the construction of commonly used predictive comparison statistics. Our approach continues to rely on the out of sample mean squared error loss differentials between the two competing models, leads to nuisance parameter-free Gaussian asymptotics, and is shown to remain valid under flexible assumptions that can accommodate heteroskedasticity and the presence of mixed predictors (e.g., stationary and local to unit root). A local power analysis also establishes their ability to detect departures from the null in both stationary and persistent settings. Simulations calibrated to common economic and financial applications indicate that our methods have strong power with good size control across commonly encountered sample sizes.

In this paper, we complement joint time-series and cross-section convergence results derived in a companion paper Hahn, Kuersteiner, and Mazzocco (2016, Central Limit Theory for Combined Cross-Section and Time Series) by allowing for serial correlation in the time-series sample. The implications of our analysis are limiting distributions that have a well-known form of long-run variances for the time-series limit. We obtain these results at the cost of imposing strict stationarity for the time-series model and conditional independence between the time-series and cross-section samples. Our results can be applied to estimators that combine time-series and cross-section data in the presence of aggregate uncertainty in models with rationally forward-looking agents.

This paper studies asymptotic properties of a posterior probability density and Bayesian estimators of spatial econometric models in the classical statistical framework. We focus on the high-order spatial autoregressive model with spatial autoregressive disturbance terms, due to a computational advantage of Bayesian estimation. We also study the asymptotic properties of Bayesian estimation of the spatial autoregressive Tobit model, as an example of nonlinear spatial models. Simulation studies show that even when the sample size is small or moderate, the posterior distribution of parameters is well approximated by a normal distribution, and Bayesian estimators have satisfactory performance, as classical large sample theory predicts.

On June 4, 1999 G.S. Maddala (popularly and affectionately known as GS) passed away in Columbus, Ohio at the age of 66. A leading figure in the econometrics profession for more than three decades, he held the University Eminent Scholar Professorship in the Department of Economics at Ohio State University at the time of his death. GS is survived by his wife Kameswari, “Kay,” and several members of his immediate family: his daughter, Tara, of Houston; his son, Vivek, of San Francisco; and two sisters who live in India.

We consider the method of moments estimation of a structural equation in a panel dynamic simultaneous equations model under different sample size combinations of cross-sectional dimension, N, and time series dimension, T. Two types of linear transformation to remove the individual-specific effects from the model, first difference and forward orthogonal demeaning, are considered. We show that the Alvarez and Arellano (2003) type GMM estimator under both transformations is consistent only if ${T \over N} \to 0$ as $\left( {N,T} \right) \to \infty $. However, it is asymptotically biased if ${{{T^3}} \over N} \to \kappa \ne 0 < \infty$ as $\left( {N,T} \right) \to \infty $. Since the validity of statistical inference depends critically on whether an estimator is asymptotically unbiased, we suggest a jackknife bias reduction method and derive its limiting distribution. Monte Carlo studies are conducted to demonstrate the importance of using an asymptotically unbiased estimator to obtain valid statistical inference.

We consider a multivariate continuous-time process, generated by a system of linear stochastic differential equations, driven by white noise, and involving coefficients that possibly vary over time. The process is observable only at discrete, but not necessarily equally-spaced, time points (though equal spacing significantly simplifies matters). Such settings represent partial extensions of ones studied extensively by A.R. Bergstrom. A model for the observed time series is deduced. Initially we focus on a first-order model, but higher-order models are discussed in the case of equally-spaced observations. Some discussion of issues of statistical inference is included.

In this study, we explore the partial identification of nonseparable models with continuous endogenous and binary instrumental variables. We show that the structural function is partially identified when it is monotone or concave in the explanatory variable. D’Haultfœuille and Février (2015, Econometrica 83(3), 1199–1210) and Torgovitsky (2015, Econometrica 83(3), 1185–1197) prove the point identification of the structural function under a key assumption that the conditional distribution functions of the endogenous variable for different values of the instrumental variables have intersections. We demonstrate that, even if this assumption does not hold, monotonicity and concavity provide identification power. Point identification is achieved when the structural function is flat or linear with respect to the explanatory variable over a given interval. We compute the bounds using real data and show that our bounds are informative.

It is common for an applied researcher to use filtered data, like seasonally adjusted series, for instance, to estimate the parameters of a dynamic regression model. In this paper, we study the effect of (linear) filters on the distribution of parameters of a dynamic regression model with a lagged dependent variable and a set of exogenous regressors. So far, only asymptotic results are available. Our main interest is to investigate the effect of filtering on the small sample bias and mean squared error. In general, these results entail a numerical integration of derivatives of the joint moment generating function of two quadratic forms in normal variables. The computation of these integrals is quite involved. However, we take advantage of the Laplace approximations to the bias and mean squared error, which substantially reduce the computational burden, as they yield relatively simple analytic expressions. We obtain analytic formulae for approximating the effect of filtering on the finite sample bias and mean squared error. We evaluate the adequacy of the approximations by comparison with Monte Carlo simulations, using the Census X-11 filter as a specific example