arXiv subjects

Clifford M. Hurvich

Publications and source records attributed to Clifford M. Hurvich.

12 recordsLinked to original sources

arXiv2026

Long Memory in Intrinsically Dynamic Factor Models

We study the generalized dynamic factor model in a long-memory setting. Unlike most recent work, which assumes a finite-dimensional factor space and short memory, our framework allows the factor space to be infinite-dimensional and the common components to exhibit long memory. We employ the two-sided estimation method of Forni, Hallin, Lippi and Reichlin (2000, Review of Economics and Statistics) to recover the common component. The long memory structure of the common component poses a challenge, as it introduces unboundedness/discontinuity in the spectral density. We address this issue by leveraging two key facts: First, the estimated operator is a projection onto the leading eigenspace and thus the eigengap provides an intrinsic scaling that partially mitigates the blow-up. Second, we perform most of our estimation in $L^p$-norm, rather than pointwise. Experimental results are presented to provide evidence supporting the theory, as well as potential improvements to it.

math.ST↗

arXiv2025

Automatic Order, Bandwidth Selection and Flaws of Eigen Adjustment in HAC Estimation

In this paper, we propose a new heteroskedasticity and autocorrelation consistent covariance matrix estimator based on the prewhitened kernel estimator and a localized leave-one-out frequency domain cross-validation (FDCV). We adapt the cross-validated log likelihood (CVLL) function to simultaneously select the order of the prewhitening vector autoregression (VAR) and the bandwidth. The prewhitening VAR is estimated by the Burg method without eigen adjustment as we find the eigen adjustment rule of Andrews and Monahan (1992) can be triggered unnecessarily and harmfully when regressors have nonzero mean. Through Monte Carlo simulations and three empirical examples, we illustrate the flaws of eigen adjustment and the reliability of our method.

econ.EM↗

arXiv2023

A Unified Frequency Domain Cross-Validatory Approach to HAC Standard Error Estimation

A unified frequency domain cross-validation (FDCV) method is proposed to obtain a heteroskedasticity and autocorrelation consistent (HAC) standard error. This method enables model/tuning parameter selection across both parametric and nonparametric spectral estimators simultaneously. The candidate class for this approach consists of restricted maximum likelihood-based (REML) autoregressive spectral estimators and lag-weights estimators with the Parzen kernel. Additionally, an efficient technique for computing the REML estimators of autoregressive models is provided. Through simulations, the reliability of the FDCV method is demonstrated, comparing favorably with popular HAC estimators such as Andrews-Monahan and Newey-West.

econ.EM↗

arXiv2021

On the Use of Information Criteria for Subset Selection in Least Squares Regression

Least squares (LS)-based subset selection methods are popular in linear regression modeling. Best subset selection (BS) is known to be NP-hard and has a computational cost that grows exponentially with the number of predictors. Recently, Bertsimas (2016) formulated BS as a mixed integer optimization (MIO) problem and largely reduced the computation overhead by using a well-developed optimization solver, but the current methodology is not scalable to very large datasets. In this paper, we propose a novel LS-based method, the best orthogonalized subset selection (BOSS) method, which performs BS upon an orthogonalized basis of ordered predictors and scales easily to large problem sizes. Another challenge in applying LS-based methods in practice is the selection rule to choose the optimal subset size k. Cross-validation (CV) requires fitting a procedure multiple times, and results in a selected k that is random across repeated application to the same dataset. Compared to CV, information criteria only require fitting a procedure once, but they require knowledge of the effective degrees of freedom for the fitting procedure, which is generally not available analytically for complex methods. Since BOSS uses orthogonalized predictors, we first explore a connection for orthogonal non-random predictors between BS and its Lagrangian formulation (i.e., minimization of the residual sum of squares plus the product of a regularization parameter and k), and based on this connection propose a heuristic degrees of freedom (hdf) for BOSS that can be estimated via an analytically-based expression. We show in both simulations and real data analysis that BOSS using a proposed Kullback-Leibler based information criterion AICc-hdf has the strongest performance of all of the LS-based methods considered and is competitive with regularization methods, with the computational effort of a single ordinary LS fit.

stat.ME↗

arXiv2020

Selection of Regression Models under Linear Restrictions for Fixed and Random Designs

Many important modeling tasks in linear regression, including variable selection (in which slopes of some predictors are set equal to zero) and simplified models based on sums or differences of predictors (in which slopes of those predictors are set equal to each other, or the negative of each other, respectively), can be viewed as being based on imposing linear restrictions on regression parameters. In this paper, we discuss how such models can be compared using information criteria designed to estimate predictive measures like squared error and Kullback-Leibler (KL) discrepancy, in the presence of either deterministic predictors (fixed-X) or random predictors (random-X). We extend the justifications for existing fixed-X criteria Cp, FPE and AICc, and random-X criteria Sp and RCp, to general linear restrictions. We further propose and justify a KL-based criterion, RAICc, under random-X for variable selection and general linear restrictions. We show in simulations that the use of the KL-based criteria AICc and RAICc results in better predictive performance and sparser solutions than the use of squared error-based criteria, including cross-validation.

stat.ME↗

arXiv2016

On the Sensitivity of the Lasso to the Number of Predictor Variables

The Lasso is a computationally efficient regression regularization procedure that can produce sparse estimators when the number of predictors (p) is large. Oracle inequalities provide probability loss bounds for the Lasso estimator at a deterministic choice of the regularization parameter. These bounds tend to zero if p is appropriately controlled, and are thus commonly cited as theoretical justification for the Lasso and its ability to handle high-dimensional settings. Unfortunately, in practice the regularization parameter is not selected to be a deterministic quantity, but is instead chosen using a random, data-dependent procedure. To address this shortcoming of previous theoretical work, we study the loss of the Lasso estimator when tuned optimally for prediction. Assuming orthonormal predictors and a sparse true model, we prove that the probability that the best possible predictive performance of the Lasso deteriorates as p increases is positive and can be arbitrarily close to one given a sufficiently high signal to noise ratio and sufficiently large p. We further demonstrate empirically that the amount of deterioration in performance can be far worse than the oracle inequalities suggest and provide a real data example where deterioration is observed.

stat.ML↗

arXiv2013

Limit Laws in Transaction-Level Asset Price Models

We consider pure-jump transaction-level models for asset prices in continuous time, driven by point processes. In a bivariate model that admits cointegration, we allow for time deformations to account for such effects as intraday seasonal patterns in volatility, and non-trading periods that may be different for the two assets. We also allow for asymmetries (leverage effects). We obtain the asymptotic distribution of the log-price process. We also obtain the asymptotic distribution of the ordinary least-squares estimator of the cointegrating parameter based on data sampled from an equally-spaced discretization of calendar time, in the case of weak fractional cointegration. For this same case, we obtain the asymptotic distribution for a tapered estimator under more

math.ST↗

arXiv2013

Efficiency for Regularization Parameter Selection in Penalized Likelihood Estimation of Misspecified Models

It has been shown that AIC-type criteria are asymptotically efficient selectors of the tuning parameter in non-concave penalized regression methods under the assumption that the population variance is known or that a consistent estimator is available. We relax this assumption to prove that AIC itself is asymptotically efficient and we study its performance in finite samples. In classical regression, it is known that AIC tends to select overly complex models when the dimension of the maximum candidate model is large relative to the sample size. Simulation studies suggest that AIC suffers from the same shortcomings when used in penalized regression. We therefore propose the use of the classical corrected AIC (AICc) as an alternative and prove that it maintains the desired asymptotic properties. To broaden our results, we further prove the efficiency of AIC for penalized likelihood methods in the context of generalized linear models with no dispersion parameter. Similar results exist in the literature but only for a restricted set of candidate models. By employing results from the classical literature on maximum-likelihood estimation in misspecified models, we are able to establish this result for a general set of candidate models. We use simulations to assess the performance of AIC and AICc, as well as that of other selectors, in finite samples for both SCAD-penalized and Lasso regressions and a real data example is considered.

stat.ML↗

arXiv2007

Semiparametric estimation of fractional cointegrating subspaces

We consider a common-components model for multivariate fractional cointegration, in which the $s\geq1$ components have different memory parameters. The cointegrating rank may exceed 1. We decompose the true cointegrating vectors into orthogonal fractional cointegrating subspaces such that vectors from distinct subspaces yield cointegrating errors with distinct memory parameters. We estimate each cointegrating subspace separately, using appropriate sets of eigenvectors of an averaged periodogram matrix of tapered, differenced observations, based on the first $m$ Fourier frequencies, with $m$ fixed. The angle between the true and estimated cointegrating subspaces is $o_p(1)$. We use the cointegrating residuals corresponding to an estimated cointegrating vector to obtain a consistent and asymptotically normal estimate of the memory parameter for the given cointegrating subspace, using a univariate Gaussian semiparametric estimator with a bandwidth that tends to $\infty$ more slowly than $n$. We use these estimates to test for fractional cointegration and to consistently identify the cointegrating subspaces.

math.ST↗

arXiv2007

Long Memory in Nonlinear Processes

It is generally accepted that many time series of practical interest exhibit strong dependence, i.e., long memory. For such series, the sample autocorrelations decay slowly and log-log periodogram plots indicate a straight-line relationship. This necessitates a class of models for describing such behavior. A popular class of such models is the autoregressive fractionally integrated moving average (ARFIMA) which is a linear process. However, there is also a need for nonlinear long memory models. For example, series of returns on financial assets typically tend to show zero correlation, whereas their squares or absolute values exhibit long memory. Furthermore, the search for a realistic mechanism for generating long memory has led to the development of other nonlinear long memory models. In this chapter, we will present several nonlinear long memory models, and discuss the properties of the models, as well as associated parametric andsemiparametric estimators.

math.ST↗

arXiv2007

Asymptotics for Duration-Driven Long Range Dependent Processes

We consider processes with second order long range dependence resulting from heavy tailed durations. We refer to this phenomenon as duration-driven long range dependence (DDLRD), as opposed to the more widely studied linear long range dependence based on fractional differencing of an $iid$ process. We consider in detail two specific processes having DDLRD, originally presented in Taqqu and Levy (1986), and Parke (1999). For these processes, we obtain the limiting distribution of suitably standardized discrete Fourier transforms (DFTs) and sample autocovariances. At low frequencies, the standardized DFTs converge to a stable law, as do the standardized sample autocovariances at fixed lags. Finite collections of standardized sample autocovariances at a fixed set of lags converge to a degenerate distribution. The standardized DFTs at high frequencies converge to a Gaussian law. Our asymptotic results are strikingly similar for the two DDLRD processes studied. We calibrate our asymptotic results with a simulation study which also investigates the properties of the semiparametric log periodogram regression estimator of the memory parameter.

math.ST↗

arXiv2006

Propagation of Memory Parameter from Durations to Counts

We establish sufficient conditions on durations that are stationary with finite variance and memory parameter $d \in [0,1/2)$ to ensure that the corresponding counting process $N(t)$ satisfies $\textmd{Var} N(t) \sim C t^{2d+1}$ ($C>0$) as $t \to \infty$, with the same memory parameter $d \in [0,1/2)$ that was assumed for the durations. Thus, these conditions ensure that the memory in durations propagates to the same memory parameter in counts and therefore in realized volatility. We then show that any utoregressive Conditional Duration ACD(1,1) model with a sufficient number of finite moments yields short memory in counts, while any Long Memory Stochastic Duration model with $d>0$ and all finite moments yields long memory in counts, with the same $d$.

math.ST↗