Correcting Estimation Bias in Dynamic Term Structure Models

The affine dynamic term structure model (DTSM) is the canonical empirical finance representation of the yield curve. However, the possibility that DTSM estimates may be distorted by small-sample bias has been largely ignored. We show that conventional estimates of DTSM coefficients are indeed severely biased, and this bias results in misleading estimates of expected future short-term interest rates and of long-maturity term premia. We provide a variety of bias-corrected estimates of affine DTSMs, for both maximally flexible and overidentified specifications. Our estimates imply interest rate expectations and term premia that are more plausible from a macrofinance perspective. This article has supplementary material online.     

Asymptotic Equivalence of Seat Bias Models

Apportionment methods round vote proportions to integer numbers of seats in a parliament. An important issue is whether a given apportionment method treats larger and smaller parties equally or gives rise to seat biases. In this paper two models of quantifying seat biases of popular apportionment methods are compared to each other. The models are found to result in the same asymptotic behaviour when the number of seats available for apportionment grows large.     

Bootstrapping regression models with BLUS residuals

To bootstrap a regression problem, pairs of response and explanatory variables or residuals can be resam‐pled, according to whether we believe that the explanatory variables are random or fixed. In the latter case, different residuals have been proposed in the literature, including the ordinary residuals (Efron 1979), standardized residuals (Bickel & Freedman 1983) and Studentized residuals (Weber 1984). Freedman (1981) has shown that the bootstrap from ordinary residuals is asymptotically valid when the number of cases increases and the number of variables is fixed. Bickel & Freedman (1983) have shown the asymptotic validity for ordinary residuals when the number of variables and the number of cases both increase, provided that the ratio of the two converges to zero at an appropriate rate. In this paper, the authors introduce the use of BLUS (Best Linear Unbiased with Scalar covariance matrix) residuals in bootstrapping regression models. The main advantage of the BLUS residuals, introduced in Theil (1965), is that they are uncorrelated. The main disadvantage is that only n —p residuals can be computed for a regression problem with n cases and p variables. The asymptotic results of Freedman (1981) and Bickel & Freedman (1983) for the ordinary (and standardized) residuals are generalized to the BLUS residuals. A small simulation study shows that even though only n — p residuals are available, in small samples bootstrapping BLUS residuals can be as good as, and sometimes better than, bootstrapping from standardized or Studentized residuals.     

On Bias Reduction in Robust Inference for Generalized Linear Models

Abstract.  It is well known that one or more outlying points in the data may adversely affect the consistency of the quasi-likelihood or the likelihood estimators for the regression effects. Similar to the quasi-likelihood approach, the existing outliers-resistant Mallow's type quasi-likelihood (MQL) estimation approach may also produce biased regression estimators. As a remedy, by using a fully standardized score function in the MQL estimating equation, in this paper, we demonstrate that the fully standardized MQL estimators are almost unbiased ensuring its higher consistency performance. Both count and binary responses subject to one or more outliers are used in the study. The small sample as well as asymptotic results for the competitive estimators are discussed.     

Hierarchical Models for Cross-Classified Overdispersed Multinomial Data

When a vector of sample proportions is not obtained through a simple random sampling, the covariance matrix for the sample vector can differ substantially from the one corresponding to the multinomial model (Wilson 1989). For example, clustering effects of subject effects in repeated-measure experiments can cause the variance of the observed proportions to be much larger than variances under the multinomial model. The phenomenon is generally referred to as overdispersion. Tallis (1962) proposed a model for identically distributed multinomials with a common measure of correlation and referred to it as the generalized multinomial model. This generalized multinomial model is extended in this article to account for overdispersion by allowing the vectors of proportions to vary according to a Dirichlet distribution. The generalized Dirichlet-multinomial model (as it is referred to here) allows for a second order of pairwise correlation among units, a type of assumption found reasonable in some biological data (Kupper and Haseman 1978) and introduced here to business data. An alternative derivation allowing for two kinds of variation is also considered. Asymptotic normal properties of parameter estimators are used to construct Wald statistics for testing hypotheses. The methods are illustrated with applications to performance evaluation monthly data and an integrated circuit yield analysis.     

Identification of Multiple Outliers in Logistic Regression

The use of logistic regression modeling has seen a great deal of attention in the literature in recent years. This includes all aspects of the logistic regression model including the identification of outliers. A variety of methods for the identification of outliers, such as the standardized Pearson residuals, are now available in the literature. These methods, however, are successful only if the data contain a single outlier. In the presence of multiple outliers in the data, which is often the case in practice, these methods fail to detect the outliers. This is due to the well-known problems of masking (false negative) and swamping (false positive) effects. In this article, we propose a new method for the identification of multiple outliers in logistic regression. We develop a generalized version of standardized Pearson residuals based on group deletion and then propose a technique for identifying multiple outliers. The performance of the proposed method is then investigated through several examples.     

Using Multinomial Mixture Models to Cluster Internet Traffic

The paper considers the clustering of two large sets of Internet traffic data consisting of information measured from headers of transmission control protocol packets collected on a busy arc of a university network connecting with the Internet. Packets are grouped into 'flows' thought to correspond to particular movements of information between one computer and another. The clustering is based on representing the flows as each sampled from one of a finite number of multinomial distributions and seeks to identify clusters of flows containing similar packet‐length distributions. The clustering uses the EM algorithm, and the data‐analytic and computational details are given.     

Recursive computation for repeated measures models with patterned variance matrices

A convenient recursive computational method for repeated measures analysis, provided by McGilchrist and Cullis (1990), has been extended by the authors to heterogeneous error structures and also to the repeated measures model with random coefficients. The approach is outlined briefly in this paper. A computing program for the approach has been written and used to obtain results for simulated data having various error structures. A summary of the results is given. The computing program together with some subroutines is available from the authors.     

On Variance Estimation in Semiparametric Regression Models

ABSTRACT

This article considers estimation of the error variance in a semiparametric regression model. The estimator, based on the semiparametric residuals, is shown to be consistent (with certain rate) for the error variance.     

Parsimonious Estimation of the Covariance Matrix in Multinomial Probit Models

This article presents a Bayesian analysis of a multinomial probit model by building on previous work that specified priors on identified parameters. The main contribution of our article is to propose a prior on the covariance matrix of the latent utilities that permits elements of the inverse of the covariance matrix to be identically zero. This allows a parsimonious representation of the covariance matrix when such parsimony exists. The methodology is applied to both simulated and real data, and its ability to obtain more efficient estimators of the covariance matrix and regression coefficients is assessed using simulated data.     

Bias correction through filtering omitted variables and instruments

This paper proposes a combination of the particle-filter-based method and the expectation-maximization algorithm (PFEM), in order to filter unobservable variables and hence, to reduce the omitted variables bias. Furthermore, I consider as an unobservable variable, an exogenous one that can be used as an instrument in the instrumental variable (IV) methodology. The aim is to show that the PFEM is able to eliminate or reduce both the omitted variable bias and the simultaneous equation bias by filtering the omitted variable and the unobserved instrument, respectively. In other words, the procedure provides (at least approximately) consistent estimates, without using additional information embedded in the omitted variable or in the instruments, since they are filtered by the observable variables. The validity of the procedure is shown both through simulations and through a comparison to an IV analysis which appeared in an important previous publication. As regards the latter point, I demonstrate that the procedure developed in this article yields similar results to those of the original IV analysis.     

Determination of the functional form of the relationship of covariates to the log hazard ratio in a Cox model

In this paper, we review available methods for determination of the functional form of the relation between a covariate and the log hazard ratio for a Cox model. We pay special attention to the detection of influential observations to the extent that they influence the estimated functional form of the relation between a covariate and the log hazard ratio. Our paper is motivated by a data set from a cohort study of lung cancer and silica exposure, where the nonlinear shape of the estimated log hazard ratio for silica exposure plotted against cumulative exposure and hereafter referred to as the exposure–response curve was greatly affected by whether or not two individuals with the highest exposures were included in the analysis. Formal influence diagnostics did not identify these two individuals but did identify the three highest exposed cases. Removal of these three cases resulted in a biologically plausible exposure–response curve.     

国外CPI偏差及其测度研究综述

在国内关于CPI是否存在偏差及如何测度偏差研究很少的学术背景下,主要围绕CPI价格指数是否存在偏差、偏差的比较基准、偏差的来源及如何测度偏差、降低偏差的主要方法等问题,对近年来国外的相关研究进行了综述,以期能对中国关于CPI的研究提供帮助。     

Estimation and comparison of lognormal parameters in the presence of censored data

It is assumed that k(k?>?2) independent samples of sizes n _i (i?=?1, …, k) are available from k lognormal distributions. Four hypothesis cases (H ₁ – H ₄) are defined. Under H ₁, all k median parameters as well as all k skewness parameters are equal; under H ₂, all k skewness parameters are equal but not all k median parameters are equal; under H ₃, all k median parameters are equal but not all k skewness parameters are equal; under H ₄, neither the k median parameters nor the k skewness parameters are equal. The Expectation Maximization (EM) algorithm is used to obtain the maximum likelihood (ML) estimates of the lognormal parameters in each of these four hypothesis cases. A (2k???1) degree polynomial is solved at each step of the EM algorithm for the H ₃ case. A two-stage procedure for testing the equality of the medians either under skewness homogeneity or under skewness heterogeneity is also proposed and discussed. A simulation study was performed for the case k?=?3.     

Bias in Small-Sample Inference With Count-Data Models

Abstract

Both Poisson and negative binomial regression can provide quasi-likelihood estimates for coefficients in exponential-mean models that are consistent in the presence of distributional misspecification. It has generally been recommended, however, that inference be carried out using asymptotically robust estimators for the parameter covariance matrix. As with linear models, such robust inference tends to lead to over-rejection of null hypotheses in small samples. Alternative methods for estimating coefficient estimator variances are considered. No one approach seems to remove all test bias, but the results do suggest that the use of the jackknife with Poisson regression tends to be least biased for inference.     

Bias reduction for the ratio of means

Confidence intervals of third-order accuracy are given for the ratio of the means of two normal distributions. A simulation study is conducted to compare these intervals with ones known. A comparison is also made with non-parametric alternatives.     

Multiplicative Bias Correction in Kernel Hazard Estimation

A multiplicative bias reducing technique is introduced for kernel hazard estimation. Similar methods were introduced by Linton & Nielsen (1994) in non-parametric regression and Jones et al . (1995) in non-parametric density estimation. A simulation study indicates good performance of the method. An application is provided on the development of Danish mortality.     

Monitoring Structural Changes in Generalized Linear Models

In this article, we introduce two monitoring schemes to (sequentially) detect structural changes in generalized linear models and develop asymptotic theories for them. The first method is based on cumulative sums (CUSUM) of weighted residuals, in which the unknown in-control parameters have been replaced by its maximum likelihood (ML) estimate from the training sample, whereas the second scheme makes use of moving sums (MOSUM) of weighted residuals. We characterize the limit distribution of the test statistic and show that these tests are consistent. Moreover, we also obtain and tabulate the asymptotic critical values of the tests. Finally, we study the speed of detection under different conditions. The methods are illustrated and compared in several simulations.     

Influence diagnostics in gamma ridge regression model

In this article, we proposed some influence diagnostics for the gamma regression model (GRM) and the gamma ridge regression model (GRRM). We assess the impact of influential observations on the GRM and GRRM estimates by extending the work of Pregibon [Logistic regression diagnostics. Ann Stat. 1981;9:705–724] and Walker and Birch [Influence measures in ridge regression. Technometrics. 1988;30:221–227]. Comparison of both models is made and demonstrated with the help of a simulation study and a real data set. We report some momentous results in detecting the influential observations and their effects on the GRM and GRRM estimates.     

Bias reduction in kernel density estimation

In this paper, we propose two kernel density estimators based on a bias reduction technique. We study the properties of these estimators and compare them with Parzen–Rosenblatt's density estimator and Mokkadem, A., Pelletier, M., and Slaoui, Y. (2009, ‘The stochastic approximation method for the estimation of a multivariate probability density’, J. Statist. Plann. Inference, 139, 2459–2478) is density estimators. It turns out that, with an adequate choice of the parameters of the two proposed estimators, the rate of convergence of two estimators will be faster than the two classical estimators and the asymptotic MISE (Mean Integrated Squared Error) will be smaller than the two classical estimators. We corroborate these theoretical results through simulations.