Journal of business & economic statistics.

Assessing Sensitivity to Unconfoundedness: Estimation and Inference.-- Identification of a Triangular Two Equation System Without Instruments.--Homogeneity and Sparsity Analysis for High-Dimensional Panel Data Models.-- Covariance Model with General Linear Structure and Divergent Parameters.-- Graphical Assistant Grouped Network Autoregression Model: A Bayesian Nonparametric Recourse.-- Likelihood Ratio Tests for Lorenz Dominance.-- Identification of Time-Varying Factor Models.-- Estimation of a Structural Break Point in Linear Regression Models.-- Getting the ROC into Sync.-- Forecasting a Nonstationary Time Series Using a Mixture of Stationary and Nonstationary Factors as Predictors.-- Bayesian Nonparametric Panel Markov-Switching GARCH Models.-- Two-Sample Testing for Tail Copulas with an Application to Equity Indices.-- Low Frequency Cointegrating Regression with Local to Unity Regressors and Unknown Form of Serial Dependence.-- Optimal Subsampling Bootstrap for Massive Data.-- Prediction Using Many Samples with Models Possibly Containing Partially Shared Parameters.-- Specification Tests for GARCH Processes with Nuisance Parameters on the Boundary.-- On the Least Squares Estimation of Multiple-Threshold-Variable Autoregressive Models.-- On Bivariate Time-Varying Price Staleness.-- Estimations and Tests for Generalized Mediation Models with High-Dimensional Potential Mediators.-- A One-Sided Refined Symmetrized Data Aggregation Approach to Robust Mutual Fund Selection.-- Probabilistic Forecast Reconciliation under the Gaussian Framework.-- High-Dimensional Censored Regression via the Penalized Tobit Likelihood.-- Two-Directional Simultaneous Inference for High-Dimensional Models.-- Estimation, Inference, and Empirical Analysis for Time-Varying VAR Models.-- Matrix Factor Analysis: From Least Squares to Iterative Projection.-- A Dynamic Binary Probit Model with Time-Varying Parameters and Shrinkage Prior.

This article provides a set of methods for quantifying the robustness of treatment effects estimated using the unconfoundedness assumption. Specifically, we estimate and do inference on bounds for various treatment effect parameters, like the Average Treatment Effect (ATE) and the average effect of treatment on the treated (ATT), under nonparametric relaxations of the unconfoundedness assumption indexed by a scalar sensitivity parameter c. These relaxations allow for limited selection on unobservables, depending on the value of c. For large enough c, these bounds equal the no assumptions bounds. Using a nonstandard bootstrap method, we show how to construct confidence bands for these bound functions which are uniform over all values of c. We illustrate these methods with an empirical application to the National Supported Work Demonstration program. We implement these methods in the companion Stata module tesensitivity for easy use in practice.

We show that a standard linear triangular two equation system can be point identified, without the use of instruments or any other side information. We find that the only case where the model is not point identified is when a latent variable that causes endogeneity is normally distributed. In this nonidentified case, we derive the sharp identified set. We apply our results to Acemoglu and Johnson’s model of life expectancy and GDP, obtaining point identification and comparable estimates to theirs, without using their (or any other) instrument.

In this article, we are interested in detecting latent group structures and significant covariates in a high-dimensional panel data model with both individual and time fixed effects. The slope coefficients of the model are assumed to be subject dependent, and there exist group structures where the slope coefficients are homogeneous within groups and heterogeneous between groups. We develop a penalized estimator for recovering the group structures and the sparsity patterns simultaneously. We propose a new algorithm to optimize the objective function. Furthermore, we propose a strategy to reduce the computational complexity by pruning the penalty terms in the objective function, which also improves the accuracy of group structure detection. The proposed estimator can recover the latent group structures and the sparsity patterns consistently in large samples. The finite sample performance of the proposed estimator is evaluated through Monte Carlo studies and illustrated with a real dataset.

For estimating the large covariance matrix with a limited sample size, we propose the covariance model with general linear structure (CMGL) by employing the general link function to connect the covariance of the continuous response vector to a linear combination of weight matrices. Without assuming the distribution of responses, and allowing the number of parameters associated with weight matrices to diverge, we obtain the quasi-maximum likelihood estimators (QMLE) of parameters and show their asymptotic properties. In addition, an extended Bayesian information criteria (EBIC) is proposed to select relevant weight matrices, and the consistency of EBIC is demonstrated. Under the identity link function, we introduce the ordinary least squares estimator (OLS) that has the closed form. Hence, its computational burden is reduced compared to QMLE, and the theoretical properties of OLS are also investigated. To assess the adequacy of the link function, we further propose the quasi-likelihood ratio test and obtain its limiting distribution. Simulation studies are presented to assess the performance of the proposed methods, and the usefulness of generalized covariance models is illustrated by an analysis of the U.S. stock market.

Vector autoregression model is ubiquitous in classical time series data analysis. With the rapid advance of social network sites, time series data over latent graph is becoming increasingly popular. In this article, we develop a novel Bayesian grouped network autoregression model, which can simultaneously estimate group information (number of groups and group configurations) and group-wise parameters. Specifically, a graphically assisted Chinese restaurant process is incorporated under the framework of the network autoregression model to improve the statistical inference performance. An efficient Markov chain Monte Carlo sampling algorithm is used to sample from the posterior distribution. Extensive studies are conducted to evaluate the finite sample performance of our proposed methodology. Additionally, we analyze two real datasets as illustrations of the effectiveness of our approach.

In testing hypotheses pertaining to Lorenz dominance (LD), researchers have examined second- and third-order stochastic dominance using empirical Lorenz processes and integrated stochastic processes with the aid of bootstrap analysis. Among these topics, analysis of third-order stochastic dominance (TSD) based on the notion of risk aversion has been examined using crossing (generalized) Lorenz curves. These facts motivated the present study to characterize distribution pairs displaying the TSD without second-order (generalized Lorenz) dominance. It further motivated the development of likelihood ratio (LR) goodness-of-fit tests for examining the respective hypotheses of the LD, crossing (generalized) Lorenz curves, and TSD through approximate Chi-squared distributions. The proposed LR tests were assessed using simulated distributions, and applied to examine the COVID-19 regional death counts of bivariate samples collected by the World Health Organization between March 2020 and February 2021.

This study proposes a point estimator of the break location for a one-time structural break in linear regression models. If the break magnitude is small, the least-squares estimator of the break date has two modes at the ends of the finite sample period, regardless of the true break location. To solve this problem, I suggest an alternative estimator based on a modification of the least-squares objective function. The modified objective function incorporates estimation uncertainty that varies across potential break dates. The new break point estimator is consistent and has a unimodal finite sample distribution under small break magnitudes. A limit distribution is provided under an in-fill asymptotic framework. Monte Carlo simulation results suggest that the new estimator outperforms the least-squares estimator. I apply the method to estimate the break date in U.S. and U.K. stock return prediction models.

Judging the conformity of binary events in macroeconomics and finance has often been done with indices that measure synchronization. In recent years, the use of Receiver Operating Characteristic (ROC) curve has become popular for this task. This article shows that the ROC and synchronization approaches are closely related, and each can be derived from a decision-making framework. Furthermore, the resulting global measures of the degree of conformity can be identified and estimated using the standard method of moments estimators. The impact of serial dependence in the underlying series upon inferences can therefore be allowed for. Such serial correlation is common in macroeconomic and financial data.

We develop a method for constructing prediction intervals for a nonstationary variable, such as GDP. The method uses a Factor Augmented Regression (FAR) model. The predictors in the model include a small number of factors generated to extract most of the information in a set of panel data on a large number of macroeconomic variables that are considered to be potential predictors. The novelty of this article is that it provides a method and justification for a mixture of stationary and nonstationary factors as predictors in the FAR model; we refer to this as mixture-FAR method. This method is important because typically such a large set of panel data, for example the FRED-QD, is likely to contain a mixture of stationary and nonstationary variables. In our simulation study, we observed that the proposed mixture-FAR method performed better than its competitor that requires all the predictors to be nonstationary; the MSE of prediction was at least 33% lower for mixture-FAR. Using the data in FRED-QD for the United States, we evaluated the aforementioned methods for forecasting the nonstationary variables, GDP and Industrial Production. We observed that the mixture-FAR method performed better than its competitors.

This article proposes Bayesian nonparametric inference for panel Markov-switching GARCH models. The model incorporates series-specific hidden Markov chain processes that drive the GARCH parameters. To cope with the high-dimensionality of the parameter space, the article assumes soft parameter pooling through a hierarchical prior distribution and introduces cross sectional clustering through a Bayesian nonparametric prior distribution. An MCMC posterior approximation algorithm is developed and its efficiency is studied in simulations under alternative settings. An empirical application to financial returns data in the United States is offered with a portfolio performance exercise based on forecasts. A comparison shows that the Bayesian nonparametric panel Markov-switching GARCH model provides good forecasting performances and economic gains in optimal asset allocation.

A novel, general two-sample hypothesis testing procedure is established for testing the equality of tail copulas associated with bivariate data. More precisely, using a martingale transformation of a natural two-sample tail copula process, a test process is constructed, which is shown to converge in distribution to a standard Wiener process. Hence, from this test process a myriad of asymptotically distribution-free two-sample tests can be obtained. The good finite-sample behavior of our procedure is demonstrated through Monte Carlo simulations. Using the new testing procedure, no evidence of a difference in the respective tail copulas is found for pairs of negative daily log-returns of equity indices during and after the global financial crisis.

This article develops new t and F tests in a low-frequency transformed triangular cointegrating regression when one may not be certain that the economic variables are exact unit root processes. We first show that the low-frequency transformed and augmented OLS (TA-OLS) method exhibits an asymptotic bias term in its limiting distribution. As a result, the test for the cointegration vector can have substantially large size distortion, even with minor deviations from the unit root regressors. To correct the asymptotic bias of the TA-OLS statistics for the cointegration vector, we develop modified TA-OLS statistics that adjust the bias and take account of the estimation uncertainty of the long-run endogeneity arising from the bias correction. Based on the modified test statistics, we provide Bonferroni-based tests of the cointegration vector using standard t and F critical values. Monte Carlo results show that our approach has the correct size and reasonable power for a wide range of local-to-unity parameters. Additionally, our method has advantages over the IVX approach when the serial dependence and the long-run endogeneity in the cointegration system are important.

The bootstrap is a widely used procedure for statistical inference because of its simplicity and attractive statistical properties. However, the vanilla version of bootstrap is no longer feasible computationally for many modern massive datasets due to the need to repeatedly resample the entire data. Therefore, several improvements to the bootstrap method have been made in recent years, which assess the quality of estimators by subsampling the full dataset before resampling the subsamples. Naturally, the performance of these modern subsampling methods is influenced by tuning parameters such as the size of subsamples, the number of subsamples, and the number of resamples per subsample. In this article, we develop a novel hyperparameter selection methodology for selecting these tuning parameters. Formulated as an optimization problem to find the optimal value of some measure of accuracy of an estimator subject to computational cost, our framework provides closed-form solutions for the optimal hyperparameter values for subsampled bootstrap, subsampled double bootstrap and bag of little bootstraps, at no or little extra time cost. Using the mean square errors as a proxy of the accuracy measure, we apply our methodology to study, compare and improve the performance of these modern versions of bootstrap developed for massive data through numerical study. The results are promising.

We consider prediction based on a main model. When the main model shares partial parameters with several other helper models, we make use of the additional information. Specifically, we propose a Model Averaging Prediction (MAP) procedure that takes into account data related to the main model as well as data related to the helper models. We allow the data related to different models to follow different structures, as long as they share some common covariate effect. We show that when the main model is misspecified, MAP yields the optimal weights in terms of prediction. Further, if the main model is correctly specified, then MAP will automatically exclude all incorrect helper models asymptotically. Simulation studies are conducted to demonstrate the superior performance of MAP. We further implement MAP to analyze a dataset related to the probability of credit card default.

This article develops tests for the correct specification of the conditional variance function in GARCH models when the true parameter may lie on the boundary of the parameter space. The test statistics considered are of Kolmogorov-Smirnov and Cramér-von Mises type, and are based on empirical processes marked by centered squared residuals. The limiting distributions of the test statistics depend on unknown nuisance parameters in a nontrivial way, making the tests difficult to implement. We therefore introduce a novel bootstrap procedure which is shown to be asymptotically valid under general conditions, irrespective of the presence of nuisance parameters on the boundary. The proposed bootstrap approach is based on shrinking of the parameter estimates used to generate the bootstrap sample toward the boundary of the parameter space at a proper rate. It is simple to implement and fast in applications, as the associated test statistics have simple closed form expressions. Although the bootstrap test is designed for a data generating process with fixed parameters (i.e., independent of the sample size n), we also discuss how to obtain valid inference for sequences of DGPs with parameters approaching the boundary at the
n
−
1
/
2
rate. A simulation study demonstrates that the new tests: (i) have excellent finite sample behavior in terms of empirical rejection probabilities under the null as well as under the alternative; (ii) provide a useful complement to existing procedures based on Ljung-Box type approaches. Two data examples illustrate the implementation of the proposed tests in applications.

Most threshold models to-date contain a single threshold variable. However, in many empirical applications, models with multiple threshold variables may be needed and are the focus of this article. For the sake of readability, we start with the Two-Threshold-Variable Autoregressive (2-TAR) model and study its Least Squares Estimation (LSE). Among others, we show that the respective estimated thresholds are asymptotically independent. We propose a new method, namely the weighted Nadaraya-Watson method, to construct confidence intervals for the threshold parameters, that turns out to be, as far as we know, the only method to-date that enjoys good probability coverage, regardless of whether the threshold variables are endogenous or exogenous. Finally, we describe in some detail how our results can be extended to the K-Threshold-Variable Autoregressive (K-TAR) model, K > 2. We assess the finite-sample performance of the LSE by simulation and present two real examples to illustrate the efficacy of our modeling.

Price staleness refers to the extent of zero returns in price dynamics. Bandi, Pirino, and Reno introduce two types of staleness: systematic and idiosyncratic staleness. In this study, we allow price staleness to be time-varying and study the statistical inference for idiosyncratic and common price staleness between two assets. We propose consistent estimators for both time-varying idiosyncratic and systematic price staleness and derive their asymptotic theory. Moreover, we develop a feasible nonparametric test for the simultaneous constancy of idiosyncratic and common price staleness. Our inference is based on infill asymptotics. Finally, we conduct simulation studies under various scenarios to assess the finite sample performance of the proposed approaches and provide an empirical application of the proposed theory.

Motivated by an empirical analysis of stock reaction to COVID-19 pandemic, we propose a generalized mediation model with high-dimensional potential mediators to study the mediation effects of financial metrics that bridge company’s sector and stock value. We propose an estimation procedure for the direct effect via a partial penalized maximum likelihood method and establish its theoretical properties. We develop a Wald test for the indirect effect and show that the proposed test has a
χ
2
limiting null distribution. We also develop a partial penalized likelihood ratio test for the direct effect and show that the proposed test asymptotically follows a
χ
2
-distribution under null hypothesis. A more efficient estimator of indirect effect under complete mediation model is also developed. Simulation studies are conducted to examine the finite sample performance of the proposed procedures and compare with some existing methods. We further illustrate the proposed methodology with an empirical analysis of stock reaction to COVID-19 pandemic via exploring the underlying mechanism of the relationship between companies’ sectors and their stock values.

We consider the problem of identifying skilled funds among a large number of candidates under the linear factor pricing models containing both observable and latent market factors. Motivated by the existence of non-strong potential factors and diversity of error distribution types of the linear factor pricing models, we develop a distribution-free multiple testing procedure to solve this problem. The proposed procedure is established based on the statistical tool of symmetrized data aggregation, which makes it robust to the strength of potential factors and distribution type of the error terms. We then establish the asymptotic validity of the proposed procedure in terms of both the false discovery rate and true discovery proportion under some mild regularity conditions. Furthermore, we demonstrate the advantages of the proposed procedure over some existing methods through extensive Monte Carlo experiments. In an empirical application, we illustrate the practical utility of the proposed procedure in the context of selecting skilled funds, which clearly has much more satisfactory performance than its main competitors.

Forecast reconciliation of multivariate time series maps a set of incoherent forecasts into coherent forecasts to satisfy a given set of linear constraints. Available methods in the literature either follow a projection matrix-based approach or an empirical copula-based reordering approach to revise the incoherent future sample paths to obtain reconciled probabilistic forecasts. The projection matrices are estimated either by optimizing a scoring rule such as energy or variogram score or simply using a projection matrix derived for point forecast reconciliation.

This article proves that (a) if the incoherent predictive distribution is jointly Gaussian, then MinT (minimum trace) minimizes the logarithmic scoring rule for the hierarchy; and (b) the logarithmic score of MinT for each marginal predictive density is smaller than that of OLS (ordinary least squares). We illustrate these theoretical results using a set of simulation studies and the Australian domestic tourism dataset. The estimation of MinT needs to estimate the covariance matrix of the base forecast errors. We have evaluated the performance using the sample covariance matrix and shrinkage estimator. It was observed that the theoretical properties noted above are greatly impacted by the covariance matrix used and highlighted the importance of estimating it reliably, especially with high dimensional data.

High-dimensional regression and regression with a left-censored response are each well-studied topics. In spite of this, few methods have been proposed which deal with both of these complications simultaneously. The Tobit model—long the standard method for censored regression in economics—has not been adapted for high-dimensional regression at all. To fill this gap and bring up-to-date techniques from high-dimensional statistics to the field of high-dimensional left-censored regression, we propose several penalized Tobit models. We develop a fast algorithm which combines quadratic majorization with coordinate descent to compute the penalized Tobit solution path. Theoretically, we analyze the Tobit lasso and Tobit with a folded concave penalty, bounding the
l
2
estimation loss for the former and proving that a local linear approximation estimator for the latter possesses the strong oracle property. Through an extensive simulation study, we find that our penalized Tobit models provide more accurate predictions and parameter estimates than other methods on high-dimensional left-censored data. We use a penalized Tobit model to analyze high-dimensional left-censored HIV viral load data from the AIDS Clinical Trials Group and identify potential drug resistance mutations in the HIV genome. A supplementary file contains intermediate theoretical results and technical proofs.

This article proposes a general two-directional simultaneous inference (TOSI) framework for high-dimensional models with a manifest variable or latent variable structure, for example, high-dimensional mean models, high-dimensional sparse regression models, and high-dimensional latent factors models. TOSI performs simultaneous inference on a set of parameters from two directions, one to test whether the assumed zero parameters indeed are zeros and one to test whether exist zeros in the parameter set of nonzeros. As a result, we can better identify whether the parameters are zeros, thereby keeping the data structure fully and parsimoniously expressed. We theoretically prove that the single-split TOSI is asymptotically unbiased and the multi-split version of TOSI can control the Type I error below the prespecified significance level. Simulations are conducted to examine the performance of the proposed method in finite sample situations and two real datasets are analyzed. The results show that the TOSI method can provide more predictive and more interpretable estimators than existing methods.

Vector autoregressive (VAR) models are widely used in practical studies, for example, forecasting, modeling policy transmission mechanism, and measuring connection of economic agents. To better capture the dynamics, this article introduces a new class of time-varying VAR models in which the coefficients and covariance matrix of the error innovations are allowed to change smoothly over time. Accordingly, we establish a set of asymptotic properties including the impulse response analyses subject to structural VAR identification conditions, an information criterion to select the optimal lag, and a Wald-type test to determine the constant coefficients. Simulation studies are conducted to evaluate the theoretical findings. Finally, we demonstrate the empirical relevance and usefulness of the proposed methods through an application on U.S. government spending multipliers.

In this article, we study large-dimensional matrix factor models and estimate the factor loading matrices and factor score matrix by minimizing square loss function. Interestingly, the resultant estimators coincide with the Projected Estimators (PE) in Yu et al. which was proposed from the perspective of simultaneous reduction of the dimensionality and the magnitudes of the idiosyncratic error matrix. In other word, we provide a least-square interpretation of the PE for the matrix factor model, which parallels to the least-square interpretation of the PCA for the vector factor model. We derive the convergence rates of the theoretical minimizers under sub-Gaussian tails. Considering the robustness to the heavy tails of the idiosyncratic errors, we extend the least squares to minimizing the Huber loss function, which leads to a weighted iterative projection approach to compute and learn the parameters. We also derive the convergence rates of the theoretical minimizers of the Huber loss function under bounded fourth or even
(
2
+
ϵ
)
th moment of the idiosyncratic errors. We conduct extensive numerical studies to investigate the empirical performance of the proposed Huber estimators relative to the state-of-the-art ones. The Huber estimators perform robustly and much better than existing ones when the data are heavy-tailed, and as a result can be used as a safe replacement in practice. An application to a Fama-French financial portfolio dataset demonstrates the empirical advantage of the Huber estimator.

This article studies a time series binary probit model in which the underlying latent variable depends on its lag and exogenous regressors. The regression coefficients for the latent variable are allowed to vary over time to capture possible model instability. Bayesian shrinkage priors are applied to automatically differentiate fixed and truly time-varying coefficients and thus avoid unnecessary model complexity. I develop an MCMC algorithm for model estimation that exploits parameter blocking to boost sampling efficiency. An efficient Monte Carlo approximation based on the Kalman filter is developed to improve the numerical stability for computing the predictive likelihood of the binary outcome. Benefits of the proposed model are illustrated in a simulation study and an application to forecast economic recessions.