协整向量误设对相关检验的影响

常用于检验既定协整关系的统计量有tDF和tECM两种，但由于真实数据生成过程未知，估计模型中可能存在一定程度的协整向量误设，从而使统计量的分布特征受到影响。本文首先探讨tDF检验的隐含系数约束α=γ，即短期弹性等于先验长期弹性；其次分析零假设下两种统计量的分布特征，以及先验设定γ对信号噪声比q进而对tECM分布特征的影响；最后在局部备择假设下，给出两种统计量的渐近分布，并表明向量误设会降低协整检验的势，其程度与设定误差d正相关。     

MIS‐SPECIFICATION TESTING IN RETROSPECT

The primary objective of this paper is threefold. First, to undertake a retrospective view of Mis‐Specification (M‐S) testing, going back to the early 20th century, with a view to (i) place it in the broader context of modeling and inference and (ii) bring out some of its special features. Second, to call into question several widely used arguments undermining the importance of M‐S testing in favor of relying on weak probabilistic assumptions in conjunction with generic robustness claims and asymptotic inference. Third, to bring out the crucial role of M‐S testing in securing trustworthy inference results. This is achieved by extending/modifying Fisher's statistical framework with a view to draw a clear line between the modeling and the inference facets of statistical induction. The proposed framework untangles the statistical from the substantive (structural) model and focuses on how to secure the adequacy of the statistical model before probing for substantive adequacy. A case is made for using joint M‐S tests based on custom‐built auxiliary regressions with a view to enhance the effectiveness and reliability of probing for potential statistical misspecifications.     

Tests of independence in a bivariate exponential distribution

The problem of testing independence in a two component series system is considered. The joint distribution of component lifetimes is modeled by the Pickands bivariate exponential distribution, which includes the widely used Marshall and Olkins distribution and the Gumbels type II distribution. The case of identical components is first addressed. Uniformly most powerful unbiased test (UMPU) and likelihood ratio test are obtained. It is shown that inspite of a nuisance parameter, the UMPU test is unconditional and this test turns out to be the same as the likelihood ratio test. The case of nonidentical components is also addressed and both UMPU and likelihood ratio tests are obtained. A UMPU test is obtained to test the identical nature of the components and extensions to the type II censoring scheme and multiple component systems are also discussed. Some modifications to account for the difference in parameters under test and use conditions are also discussed.     

Specification via model selection in vector error correction models

This paper proposes a model selection approach for the specification of the cointegrating rank in the VECM representation of VAR models. Asymptotic properties of estimates are derived and their features compared with the traditional likelihood ratio based approach.     

面板协整检验有限样本性质的模拟比较

面板协整检验是基于渐近分布的检验,有限样本下统计量的检验水平和检验功效的表现涉及检验的可靠性。本文针对目前实证研究中应用最广的一类基于残差的统计量及文献中最新提出的基于准残差的统计量进行蒙特卡罗模拟,比较10个检验统计量在不同DGP设定下的检验水平和检验功效,尤其是在DGP误设时的表现。模拟结果表明:基于准残差的面板协整检验大多数情况下有着更好的检验水平和检验功效表现。这一研究为解决实证中面临的统计量可靠性甄别与选择问题提供了依据。     

Export-led growth: a survey of the empirical literature and some non-causality results. Part 2

This paper continues the investigation of Giles and Williams (2000) on export-led growth (ELG). In the first part, we surveyed the empirical export-led growth literature; it was evident that Granger non-causality tests are commonly applied as a test for ELG. In this paper, we explore the sensitivity of the test for exclusions restrictions often used as the Granger non-causality test for ELG by reconsidering two applications: Oxley's (1993) study for Portugal and Henriques and Sadorsky's (1996) analysis for Canada. We focus on the robustness of the method adopted to deal with non-stationarity, including the choice of deterministic trend degree. We show that different noncausality outcomes are easy to obtain, and consequently we recommend that readers interpret the empirical ELG literature with care. Our analysis also highlights the importance of examining the robustness of Granger non-causality test results to avoid spurious outcomes in applications.     

Maximum-likelihood estimation of misspecified models

Misspecified models occur frequently in econometric practice. It is therefore important to study the sampling distribution of maximum-likelihood estimators of parameters of misspecified models. This note exhibits the asymptotic covariance matrix of the ML estimator of a misspecified model. It points out that the expression for this matrix given by White is incorrect except for the very special case, rarely occuring in econometrics, that each observation is independent and identically distributed. An illustration using the standard linear regression model is provided.     

Statistical adequacy and the trustworthiness of empirical evidence: Statistical vs. substantive information

The paper focuses on how the traditional textbook approach to econometrics, by conflating statistical and substantive information, has contributed significantly to the mountains of untrustworthy evidence accumulated over the last century. In a nutshell, the problem is that when one's favorite theory is foisted on the data, the end result is invariably an empirical model which is both statistically and substantively misspecified, but one has no way to disentangle the two sources of error in order to draw reliable inferences. It is argued that ignoring statistical misspecification, and focusing exclusively on the evaluation of the statistical results – taken at face value – on substantive grounds, has proved a disastrous strategy for learning from data. Moreover, the traditional textbook stratagems of error-fixing designed to alleviate statistical misspecification often make matters worse. Instead, the paper proposes a number of strategies that separate the statistical and substantive sources of information, ab initio, and address the problem by replacing goodness-of-fit with statistical adequacy to secure the statistical reliability of inference, and then proceed to pose questions of substantive adequacy.     

Weighted least squares estimators in possibly misspecified nonlinear regression

The behavior of estimators for misspecified parametric models has been well studied. We consider estimators for misspecified nonlinear regression models, with error and covariates possibly dependent. These models are described by specifying a parametric model for the conditional expectation of the response given the covariates. This is a parametric family of conditional constraints, which makes the model itself close to nonparametric. We study the behavior of weighted least squares estimators both when the regression function is correctly specified, and when it is misspecified and also involves possible additional covariates.