pith. sign in

arxiv: 2605.15596 · v1 · pith:RWKSDC7Snew · submitted 2026-05-15 · 📊 stat.ME

Tail postcoloring in long-run variance estimation of time series

Xu Liu , Kin Wai Chan This is my paper

Pith reviewed 2026-05-20 17:14 UTC · model grok-4.3

classification 📊 stat.ME
keywords long-run variance estimationtime seriestail postcoloringprewhiteningparametric robustnessscaling factorautocovariancesMCMC output analysis
0
0 comments X

The pith

Tail postcoloring projects neglected tail autocovariances onto parametric models through a scaling factor to estimate long-run variance.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper develops tail postcoloring to improve long-run variance estimation when time series exhibit strong autocorrelation. A scaling factor projects the tail autocovariances missed by nonparametric methods onto a parametric coloring model. This lets the estimator switch arms via bandwidth without full prewhitening of the data. The result is a parametric convergence rate under correct specification and greater robustness to misspecification in small samples. It also sidesteps variance inflation from recoloring and extends to multiple models or multivariate series.

Core claim

The central claim is that tail postcoloring, implemented through a scaling factor, successfully incorporates parametric information about tail autocovariances into nonparametric long-run variance estimators. This construction yields the parametric convergence rate whenever the coloring model is correctly specified. In finite samples the estimator tolerates misspecification of that model better than the whitening step in prewhitening methods and avoids the variance or power penalties that recoloring can impose. The same scaling construction supports multiply robust versions using several parametric models at once and applies unchanged to multivariate time series.

What carries the argument

The scaling factor that projects the tail autocovariances neglected by the nonparametric estimator onto the chosen parametric coloring model, with bandwidth governing the switch.

If this is right

  • When the coloring model is well-specified, a parametric convergence rate is achieved.
  • The estimator is more robust to misspecification of the coloring model than the standard prewhitening approach in finite samples.
  • It avoids severe potential variance inflation or power reduction caused by the recoloring factor.
  • Multiple parametric models can be used to construct a multiply robust tail postcolored estimator.
  • It naturally works for multivariate time series.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • If applied to Markov chain Monte Carlo output, the method may produce more accurate variance estimates for the simulated chains.
  • Researchers could investigate whether the multiply robust version reduces sensitivity to the choice of any single parametric model in practice.
  • The automatic bandwidth switching may generalize to other hybrid nonparametric-parametric estimators in dependent data settings.

Load-bearing premise

The scaling factor must project the neglected tail autocovariances from the nonparametric estimator onto the parametric model without bias or variance inflation, and the bandwidth must correctly switch between the two arms.

What would settle it

Simulation experiments where data follow the parametric coloring model exactly, yet the tail postcolored estimator fails to attain the parametric rate or has higher error than a fully parametric estimator would indicate the claim is false.

Figures

Figures reproduced from arXiv: 2605.15596 by Kin Wai Chan, Xu Liu.

Figure 1
Figure 1. Figure 1: (a) A schematic illustration of tail autocovariances. If ℓ = 15, only {γ˜k}∣k∣≤15 are used, while {γ˜k}∣k∣>15 are unused in the nonparametric estimator ˜v. Tail postcoloring projects the data to the nearest fitted model, and estimates the tail autocovariances {γk}∣k∣>15 parametrically. (b) The optimal bandwidth ℓ defined in (3.4). Bartlett kernel and ar(1) tail postcoloring model are used. The data are gen… view at source ↗
Figure 2
Figure 2. Figure 2: Heat maps display the asymptotic differences of the MSEs after standardizing by the true value of v 2 . Plot (a), (b), and (c) show the limiting values of {MSE(vˆpw)−MSE(v˜un)}n 2/3 /v 2 , {MSE(vˆtail)−MSE(v˜un)}n 2/3 /v 2 , and {MSE(vˆtail)−MSE(vˆpw)}n 2/3 /v 2 , respectively. See Example 3.2 for the details. A difference less than zero (i.e., blueish regions) indicates an improvement of the first estimat… view at source ↗
Figure 3
Figure 3. Figure 3: Plots (a) and (b) show the root mean-squared error and bias, i.e., MSE(⋅) 1/2 and Bias(⋅), of ˆv [1] tail (green dashed), ˆv [2] tail (blue dot-dashed), ˆv [3] tail (black dotted), and the proposed multi-model tail postcolored estimator ˆv [1∶2] tail (red solid) against different c, respectively. The value of MSE(vˆ [3] tail) 1/2 , which ranges from 0.36 to 0.72, is obviously larger the root mean squared e… view at source ↗
Figure 4
Figure 4. Figure 4: Upper row: Power curves for testing β0 = 0, where the black dashed horizontal lines designate the nominal size 5% and 0. Lower row: Empirical distribution functions of the p-values under H0, where the black dashed line is the distribution function of Unif(0, 1). Tests (a)–(d) use V˜ un (purple dotted curves), Vˆ pw (green dashed curves), V˜ lugs (blue dash-dotted curves), and Vˆ tail (red solid curves), re… view at source ↗
read the original abstract

Prewhitening is a common approach to deal with strong autocorrelation. In this article, we propose a new approach called tail postcoloring, motivated by it. It uses parametric models to project, or color back, the neglected tail autocovariances in nonparametric estimators onto the final estimator. This approach bridges the non-parametric variance estimator and the parametric coloring model through a scaling factor. It automatically switches between these two arms using a bandwidth parameter, without the need to transform the entire dataset into residuals, as in the standard prewhitening approach. When the coloring model is well-specified, a parametric rate can be achieved. In finite samples, it is also more robust to misspecification of the coloring model compared to the whitening model in the standard approach. Besides, it avoids severe potential variance inflation or power reduction caused by the recoloring factor in the standard approach. We show that multiple parametric models can be used to construct a multiply robust tail postcolored estimator. It also naturally works for multivariate time series. A real-data example in Markov chain Monte Carlo output analysis is provided.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

3 major / 1 minor

Summary. The manuscript proposes tail postcoloring as a new method for long-run variance estimation in time series. It projects neglected tail autocovariances from a nonparametric kernel estimator onto a parametric coloring model via a scaling factor, with an automatic bandwidth switch between the nonparametric and parametric arms; this avoids transforming the full data series as in standard prewhitening. The central claims are that a parametric convergence rate is attained when the coloring model is correctly specified, that the estimator is more robust to misspecification in finite samples than the whitening approach, that it avoids variance inflation from recoloring factors, that multiple parametric models yield a multiply robust version, and that the method extends naturally to multivariate series, illustrated by an MCMC output analysis example.

Significance. If the asymptotic results and finite-sample advantages hold, the hybrid postcoloring construction could provide a practical improvement over existing HAC estimators by combining nonparametric flexibility with parametric efficiency without the data-transformation step or recoloring-induced variance penalty of prewhitening, potentially benefiting inference in strongly dependent time series and MCMC settings.

major comments (3)
  1. [Abstract] Abstract: the claim that a parametric rate is achieved when the coloring model is well-specified rests on the scaling factor correctly projecting the tail autocovariances without retaining nonparametric bias; however, no rate conditions on the bandwidth or moment assumptions are stated to ensure the kernel truncation bias vanishes in the limit.
  2. [Method description] The bandwidth parameter is presented as the device that switches arms and makes the nonparametric contribution negligible, yet the manuscript supplies neither the precise growth rate required for this switch to be asymptotically sharp nor the conditions under which the projection remains bias-free.
  3. [Finite-sample analysis] Assertions of greater finite-sample robustness to misspecification of the coloring model and avoidance of variance inflation relative to standard prewhitening lack supporting simulation evidence or finite-sample bounds, leaving the comparative advantage unverified.
minor comments (1)
  1. Notation for the scaling factor and the precise definition of the bandwidth selector could be introduced earlier to improve readability.

Simulated Author's Rebuttal

3 responses · 0 unresolved

We thank the referee for the constructive and detailed comments, which help clarify the presentation of our asymptotic results and strengthen the empirical support for the finite-sample properties. We address each major comment below and indicate the revisions planned for the next version of the manuscript.

read point-by-point responses

  1. Referee: [Abstract] Abstract: the claim that a parametric rate is achieved when the coloring model is well-specified rests on the scaling factor correctly projecting the tail autocovariances without retaining nonparametric bias; however, no rate conditions on the bandwidth or moment assumptions are stated to ensure the kernel truncation bias vanishes in the limit.

    Authors: We agree that the abstract claim of parametric convergence requires explicit supporting conditions to be rigorous. In the revised manuscript we will add the necessary rate conditions on the bandwidth together with moment assumptions (e.g., finite fourth moments and a kernel with appropriate order) that ensure the truncation bias of the nonparametric estimator is o_p(n^{-1/2}) when the coloring model is correctly specified. These conditions will be stated both in the abstract and in the main asymptotic theorem so that the projection via the scaling factor is bias-free at the parametric rate. revision: yes

  2. Referee: [Method description] The bandwidth parameter is presented as the device that switches arms and makes the nonparametric contribution negligible, yet the manuscript supplies neither the precise growth rate required for this switch to be asymptotically sharp nor the conditions under which the projection remains bias-free.

    Authors: We accept that the precise growth rate for the bandwidth and the conditions guaranteeing an asymptotically sharp switch were not stated explicitly. In the revision we will specify that the bandwidth must satisfy h → ∞ while h = o(n^{1/4}) (or the analogous rate compatible with the kernel order) so that the nonparametric contribution becomes negligible at the parametric rate. We will also add the regularity conditions on the kernel and on the parametric coloring model that keep the projection asymptotically unbiased under correct specification. revision: yes

  3. Referee: [Finite-sample analysis] Assertions of greater finite-sample robustness to misspecification of the coloring model and avoidance of variance inflation relative to standard prewhitening lack supporting simulation evidence or finite-sample bounds, leaving the comparative advantage unverified.

    Authors: The referee correctly notes that the current version relies primarily on the MCMC real-data illustration and does not contain Monte Carlo evidence for the finite-sample robustness claims. We will add a dedicated simulation section in the revision that compares tail postcoloring with standard prewhitening under correctly specified and misspecified coloring models. The experiments will report bias, variance, and coverage probabilities to document the claimed robustness and the absence of recoloring-induced variance inflation. revision: yes

Circularity Check

0 steps flagged

No significant circularity in tail postcoloring derivation chain

full rationale

The paper introduces a scaling factor to project neglected tail autocovariances from the nonparametric long-run variance estimator onto a parametric coloring model, together with a bandwidth that automatically switches between the two arms. These quantities are defined explicitly in the construction and equipped with separate asymptotic conditions (bandwidth growth rates and moment assumptions) that are not tautological with the target parametric rate. The claim that a parametric rate is achieved when the coloring model is well-specified follows from standard bias-variance analysis under those stated conditions rather than by redefinition or fitting of the same parameters. No load-bearing self-citations, uniqueness theorems imported from prior author work, or ansatzes smuggled via citation are required for the central argument. The derivation remains self-contained with respect to external benchmarks in long-run variance estimation and does not reduce any prediction to its own inputs by construction.

Axiom & Free-Parameter Ledger

2 free parameters · 1 axioms · 0 invented entities

The method depends on the existence of a usable parametric coloring model for tail projection and on the scaling factor correctly linking estimators; bandwidth choice is an additional tuning element.

free parameters (2)
  • bandwidth parameter
    Controls automatic switching between nonparametric base and parametric tail correction.
  • scaling factor
    Bridges nonparametric variance estimator with parametric coloring model.
axioms (1)
  • domain assumption Parametric models can accurately project neglected tail autocovariances onto the nonparametric estimator
    Invoked to achieve parametric rate when model is well-specified.

pith-pipeline@v0.9.0 · 5714 in / 1282 out tokens · 87262 ms · 2026-05-20T17:14:26.864557+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Lean theorems connected to this paper

Citations machine-checked in the Pith Canon. Every link opens the source theorem in the public Lean library.

  • IndisputableMonolith/Cost/FunctionalEquation washburn_uniqueness_aczel unclear
    ?
    unclear

    Relation between the paper passage and the cited Recognition theorem.

    It uses parametric models to project, or color back, the neglected tail autocovariances in nonparametric estimators onto the final estimator. This approach bridges the non-parametric variance estimator and the parametric coloring model through a scaling factor.

What do these tags mean?
matches
The paper's claim is directly supported by a theorem in the formal canon.
supports
The theorem supports part of the paper's argument, but the paper may add assumptions or extra steps.
extends
The paper goes beyond the formal theorem; the theorem is a base layer rather than the whole result.
uses
The paper appears to rely on the theorem as machinery.
contradicts
The paper's claim conflicts with a theorem or certificate in the canon.
unclear
Pith found a possible connection, but the passage is too broad, indirect, or ambiguous to say the theorem truly supports the claim.

Reference graph

Works this paper leans on

268 extracted references · 268 canonical work pages

  1. [1]

    Geyer and Leif T

    Charles J. Geyer and Leif T. Johnson , date-added =. mcmc: Markov Chain Monte Carlo , year =

    work page

  2. [2]

    , date-added =

    Casini, A. , date-added =. Comment on. Econometrica , number =

    work page

  3. [3]

    Wu, W. B. , date-added =. Fourier transforms of stationary processes , volume =. Proc. Amer. Math. Soc. , pages =

    work page

  4. [4]

    On a threshold model , year =

    Tong, Howell , date-added =. On a threshold model , year =. Pattern recognition and signal processing , pages =

    work page

  5. [5]

    Central limit theorem for Fourier transforms of stationary processes , volume =

    Peligrad, Magda and Wu, Wei Biao , date-added =. Central limit theorem for Fourier transforms of stationary processes , volume =. Ann. Probab. , pages =

    work page

  6. [6]

    Killick and P

    R. Killick and P. Fearnhead and I. A. Eckley , date-added =. Optimal Detection of Changepoints With a Linear Computational Cost , volume =. Journal of the American Statistical Association , number =. 2012 , bdsk-url-1 =

    work page 2012

  7. [7]

    Automatic identification of near-stationary traffic states based on the PELT changepoint detection , volume =

    Yan, Qinglong and Sun, Zhe and Gan, Qijian and Jin, Wen-Long , date-added =. Automatic identification of near-stationary traffic states based on the PELT changepoint detection , volume =. Transportation Research Part B: Methodological , pages =

    work page

  8. [8]

    The variance of the mean of a stationary process , volume =

    Hannan, EJ , date-added =. The variance of the mean of a stationary process , volume =. J. R. Stat. Soc. Ser. B Stat. Methodol. , number =

    work page

  9. [9]

    The comparison of means of sets of observations from sections of independent stochastic series , volume =

    Jowett, GH , date-added =. The comparison of means of sets of observations from sections of independent stochastic series , volume =. J. R. Stat. Soc. Ser. B Stat. Methodol. , number =

    work page

  10. [10]

    Periodogram analysis and continuous spectra , volume =

    Bartlett, Maurice S , date-added =. Periodogram analysis and continuous spectra , volume =. Biometrika , number =

    work page

  11. [11]

    Statistical evaluation of climate experiments with general circulation models: A parametric time series modeling approach , volume =

    Katz, Richard W , date-added =. Statistical evaluation of climate experiments with general circulation models: A parametric time series modeling approach , volume =. Journal of Atmospheric Sciences , number =

    work page

  12. [12]

    Statistical procedures for making inferences about climate variability , volume =

    Katz, Richard W , date-added =. Statistical procedures for making inferences about climate variability , volume =. Journal of climate , number =

    work page

  13. [13]

    On the irrelevance of impossibility theorems: the case of the long-run variance , volume =

    Perron, Pierre and Ren, Linxia , date-added =. On the irrelevance of impossibility theorems: the case of the long-run variance , volume =. Journal of Time Series Econometrics , number =

    work page

  14. [14]

    Nonparametric methods in change point problems , volume =

    Brodsky, Emily and Darkhovsky, Boris S , date-added =. Nonparametric methods in change point problems , volume =

    work page

  15. [15]

    Bias correction of quadratic spectral estimators , volume =

    Astfalck, Lachlan C and Sykulski, Adam M and Cripps, Edward J , date-added =. Bias correction of quadratic spectral estimators , volume =. Biometrika , number =

    work page

  16. [16]

    Debiasing

    Astfalck, Lachlan C and Sykulski, Adam M and Cripps, Edward J , date-added =. Debiasing. Biometrika , number =

    work page

  17. [17]

    Statistical analysis of stationary time series , volume =

    Grenander, Ulf and Rosenblatt, Murray , date-added =. Statistical analysis of stationary time series , volume =

    work page

  18. [18]

    The measurement of power spectra from the point of view of communications engineering---Part

    Blackman, Ralph Beebe and Tukey, John Wilder , date-added =. The measurement of power spectra from the point of view of communications engineering---Part. Bell System Technical Journal , number =

    work page

  19. [19]

    Power spectral methods of analysis and their application to problems in airplane dynamics , year =

    Press, Harry and Tukey, John W , booktitle =. Power spectral methods of analysis and their application to problems in airplane dynamics , year =

    work page

  20. [20]

    Hansen, B. E. , date-added =. Least Squares Model Averaging , volume =. Econometrica , pages =

    work page

  21. [21]

    and Hjort, N

    Claeskens, G. and Hjort, N. L. , date-added =. Model Selection and Model Averaging , year =

    work page

  22. [22]

    Robins, J. M. and Rotnitzky, A. and Zhao, L. P. , date-added =. Estimation of Regression Coefficients When Some Regressors are not Always Observed , volume =. J. Amer. Statist. Assoc. , pages =

    work page

  23. [23]

    Scharfstein, D. O. and Rotnitzky, A. and Robins, J. M. , date-added =. Adjusting for Nonignorable Drop-Out Using Semiparametric Nonresponse Models , volume =. J. Amer. Statist. Assoc. , pages =

    work page

  24. [24]

    , date-added =

    Han, P. , date-added =. Multiply Robust Estimation in Regression Analysis With Missing Data , volume =. J. Amer. Statist. Assoc. , pages =

    work page

  25. [25]

    and Hall, P

    Delaigle, A. and Hall, P. , date-added =. Parametrically Assisted Nonparametric Estimation of a Density in the Deconvolution Problem , volume =. J. Amer. Statist. Assoc. , pages =

    work page

  26. [26]

    and Mishra, S

    Martins-Filho, C. and Mishra, S. and Ullah, A. , date-added =. A Class of Improved Parametrically Guided Nonparametric Regression Estimators , volume =. Econometric Rev. , pages =

    work page

  27. [27]

    and Su, L

    Mishra, S. and Su, L. and Ullah, A. , date-added =. Semiparametric Estimator of Time Series Conditional Variance , volume =. J. Bus. Econom. Statist. , pages =

    work page

  28. [28]

    Glad, I. K. , date-added =. Parametrically Guided Non-parametric Regression , volume =. Scand. J. Statist. , pages =

    work page

  29. [29]

    Hjort, N. L. and Glad, I. K. , date-added =. Nonparametric Density Estimation with a Parametric Start , volume =. Ann. Statist. , pages =

    work page

  30. [30]

    and Spiegelman, C

    Olkin, I. and Spiegelman, C. H. , date-added =. A Semiparametric Approach to Density Estimation , volume =. J. Amer. Statist. Assoc. , pages =

    work page

  31. [31]

    Tsiatis, A. A. , date-added =. Semiparametric Theory and Missing Data , year =

    work page

  32. [32]

    Finite Sample Properties of Tests based on Prewhitened Nonparametric Covariance Estimators , volume =

    Preinerstorfer, David , date-added =. Finite Sample Properties of Tests based on Prewhitened Nonparametric Covariance Estimators , volume =. Electron. J. Stat. , number =

    work page

  33. [33]

    Theory of evolutionary spectra for heteroskedasticity and autocorrelation robust inference in possibly misspecified and nonstationary models , volume =

    Casini, Alessandro , date-added =. Theory of evolutionary spectra for heteroskedasticity and autocorrelation robust inference in possibly misspecified and nonstationary models , volume =. J. Econometrics , number =

    work page

  34. [34]

    The fixed-b limiting distribution and the

    Casini, Alessandro , date-added =. The fixed-b limiting distribution and the. J. Econometrics , number =

    work page

  35. [35]

    Prewhitened long-run variance estimation robust to nonstationarity , volume =

    Casini, Alessandro and Perron, Pierre , date-added =. Prewhitened long-run variance estimation robust to nonstationarity , volume =. J. Econometrics , number =

    work page

  36. [36]

    The Annals of Statistics , number =

    Xiao, Han and Wu, Wei Biao , date-added =. The Annals of Statistics , number =

    work page

  37. [37]

    M. J. Bus. Econom. Statist. , number =

    work page

  38. [38]

    Covariance matrix estimation in time series , volume =

    Wu, Wei Biao and Xiao, Han , booktitle =. Covariance matrix estimation in time series , volume =

    work page

  39. [39]

    and Lewis, D

    Lazarus, E. and Lewis, D. J. and Stock, J. H. and Watson, M. W. , date-added =. J. Bus. Econom. Statist. , number =

    work page

  40. [40]

    Alhamzawi, Rahim , journal =

    work page

  41. [41]

    Croissant, Yves and Graves, Spencer , note =

    work page

  42. [42]

    Kozumi, Hideo and Kobayashi, Genya , journal =

    work page

  43. [43]

    Mroz, T. A. , date-added =. The Sensitivity of an Empirical Model of Married Women's Hours of Work to Economic and Statistical Assumptions , volume =. Econometrica , number =

    work page

  44. [44]

    K. W. Chan and C. Y. Yau , date-added =. Asymptotically Constant Risk Estimator of Time-average Variance Constant , volume =. Biometrika , pages =

    work page

  45. [45]

    Lugsail lag windows for estimating time-average covariance matrices , volume =

    Vats, Dootika and Flegal, James M , date-added =. Lugsail lag windows for estimating time-average covariance matrices , volume =. Biometrika , number =

    work page

  46. [46]

    Optimal difference-based variance estimators in time series: A general framework , volume =

    Chan, Kin Wai , date-added =. Optimal difference-based variance estimators in time series: A general framework , volume =. Ann. Statist. , number =

    work page

  47. [47]

    Leung, M. F. and Chan, K. W. , date-added =. Principles of Statistical Inference in Online Problems , volume =. Bernoulli , number =

    work page

  48. [48]

    and Chan, K

    Liu, X. and Chan, K. W. , date-modified =. Positive Definite Converging Kernel Estimation of Long-run Variance , volume =. J. Bus. Econom. Statist. , pages =

    work page

  49. [49]

    Sharp inequalities for martingales and stochastic integrals , volume =

    Burkholder, Donald L , date-added =. Sharp inequalities for martingales and stochastic integrals , volume =. Ast

    work page

  50. [50]

    Mean-structure and autocorrelation consistent covariance matrix estimation , volume =

    Chan, Kin Wai , date-added =. Mean-structure and autocorrelation consistent covariance matrix estimation , volume =. J. Bus. Econom. Statist. , number =

    work page

  51. [51]

    Various Versatile Variances: An Object-Oriented Implementation of Clustered Covariances in

    Achim Zeileis and Susanne K\"oll and Nathaniel Graham , date-added =. Various Versatile Variances: An Object-Oriented Implementation of Clustered Covariances in. 2020 , bdsk-url-1 =. doi:10.18637/jss.v095.i01 , journal =

    work page doi:10.18637/jss.v095.i01 2020

  52. [52]

    Flegal and John Hughes and Dootika Vats and Ning Dai and Kushagra Gupta and Uttiya Maji , date-added =

    James M. Flegal and John Hughes and Dootika Vats and Ning Dai and Kushagra Gupta and Uttiya Maji , date-added =. mcmcse: Monte Carlo Standard Errors for MCMC , year =

    work page

  53. [53]

    Analysis of financial time series , year =

    Tsay, Ruey S , date-added =. Analysis of financial time series , year =

    work page

  54. [54]

    All of Nonparametric Statistics , year =

    Wasserman, Larry , date-added =. All of Nonparametric Statistics , year =

    work page

  55. [55]

    S , date-added =

    Bullen, P. S , date-added =. Handbook of Means and Their Inequalities , year =

    work page

  56. [56]

    Multivariate output analysis for

    Vats, Dootika and Flegal, James M and Jones, Galin L , copyright =. Multivariate output analysis for. Biometrika , language =

    work page

  57. [57]

    Efficient tests for general persistent time variation in regression coefficients , volume =

    Elliott, Graham and M. Efficient tests for general persistent time variation in regression coefficients , volume =. The Review of Economic Studies , number =

    work page

  58. [58]

    Reducing the size distortion of the

    Kurozumi, Eiji and Tanaka, Shinya , date-added =. Reducing the size distortion of the. J. Time Ser. Anal. , number =

    work page

  59. [59]

    Forecasting stock returns , volume =

    Rapach, David and Zhou, Guofu , booktitle =. Forecasting stock returns , volume =

    work page

  60. [60]

    Predicting the equity premium with dividend ratios , volume =

    Goyal, Amit and Welch, Ivo , date-added =. Predicting the equity premium with dividend ratios , volume =. Management Science , number =

    work page

  61. [61]

    A comprehensive look at the empirical performance of equity premium prediction , volume =

    Welch, Ivo and Goyal, Amit , date-added =. A comprehensive look at the empirical performance of equity premium prediction , volume =. The Review of Financial Studies , number =

    work page

  62. [62]

    Residual-based tests for the null of stationarity with applications to US macroeconomic time series , volume =

    Choi, In , date-added =. Residual-based tests for the null of stationarity with applications to US macroeconomic time series , volume =. Econometric Theory , number =

    work page

  63. [63]

    The fragility of the KPSS stationarity test , volume =

    Cappuccio, Nunzio and Lubian, Diego , date-added =. The fragility of the KPSS stationarity test , volume =. Statistical Methods and Applications , number =

    work page

  64. [64]

    Generalizations of the KPSS-test for stationarity , volume =

    Hobijn, Bart and Franses, Philip Hans and Ooms, Marius , date-added =. Generalizations of the KPSS-test for stationarity , volume =. Statistica Neerlandica , number =

    work page

  65. [65]

    Detecting and predicting forecast breakdowns , volume =

    Giacomini, Raffaella and Rossi, Barbara , date-added =. Detecting and predicting forecast breakdowns , volume =. The Review of Economic Studies , number =

    work page

  66. [66]

    Testing the null hypothesis of stationarity against the alternative of a unit root: How sure are we that economic time series have a unit root? , volume =

    Kwiatkowski, Denis and Phillips, Peter CB and Schmidt, Peter and Shin, Yongcheol , date-added =. Testing the null hypothesis of stationarity against the alternative of a unit root: How sure are we that economic time series have a unit root? , volume =. J. Econometrics , number =

    work page

  67. [67]

    Genomic data support the hominoid slowdown and an Early Oligocene estimate for the hominoid--cercopithecoid divergence , volume =

    Steiper, Michael E and Young, Nathan M and Sukarna, Tika Y , date-added =. Genomic data support the hominoid slowdown and an Early Oligocene estimate for the hominoid--cercopithecoid divergence , volume =. Proceedings of the National Academy of Sciences , number =

    work page

  68. [68]

    Molecular evolution: a statistical approach , year =

    Yang, Ziheng , date-added =. Molecular evolution: a statistical approach , year =

    work page

  69. [69]

    Searching for efficient Markov chain Monte Carlo proposal kernels , volume =

    Yang, Ziheng and Rodr. Searching for efficient Markov chain Monte Carlo proposal kernels , volume =. Proceedings of the National Academy of Sciences , number =

    work page

  70. [70]

    Bayesian molecular clock dating of species divergences in the genomics era , volume =

    Dos Reis, Mario and Donoghue, Philip CJ and Yang, Ziheng , date-added =. Bayesian molecular clock dating of species divergences in the genomics era , volume =. Nature Reviews Genetics , number =

    work page

  71. [71]

    MRBAYES: Bayesian inference of phylogenetic trees , volume =

    Huelsenbeck, John P and Ronquist, Fredrik , date-added =. MRBAYES: Bayesian inference of phylogenetic trees , volume =. Bioinformatics , number =

    work page

  72. [72]

    BEAST: Bayesian evolutionary analysis by sampling trees , volume =

    Drummond, Alexei J and Rambaut, Andrew , date-added =. BEAST: Bayesian evolutionary analysis by sampling trees , volume =. BMC evolutionary biology , number =

    work page

  73. [73]

    Evolution of protein molecules , volume =

    Jukes, Thomas H and Cantor, Charles R and others , date-added =. Evolution of protein molecules , volume =. Mammalian Protein Metabolism , pages =

    work page

  74. [74]

    Ancient human genome sequence of an extinct Palaeo-Eskimo , volume =

    Rasmussen, Morten and Li, Yingrui and Lindgreen, Stinus and Pedersen, Jakob Skou and Albrechtsen, Anders and Moltke, Ida and Metspalu, Mait and Metspalu, Ene and Kivisild, Toomas and Gupta, Ramneek and others , date-added =. Ancient human genome sequence of an extinct Palaeo-Eskimo , volume =. Nature , number =

    work page

  75. [75]

    A complete Neandertal mitochondrial genome sequence determined by high-throughput sequencing , volume =

    Green, Richard E and Malaspinas, Anna-Sapfo and Krause, Johannes and Briggs, Adrian W and Johnson, Philip LF and Uhler, Caroline and Meyer, Matthias and Good, Jeffrey M and Maricic, Tomislav and Stenzel, Udo and others , date-added =. A complete Neandertal mitochondrial genome sequence determined by high-throughput sequencing , volume =. Cell , number =

    work page

  76. [76]

    Dating the emergence of pandemic influenza viruses , volume =

    Smith, Gavin JD and Bahl, Justin and Vijaykrishna, Dhanasekaran and Zhang, Jinxia and Poon, Leo LM and Chen, Honglin and Webster, Robert G and Peiris, JS Malik and Guan, Yi , date-added =. Dating the emergence of pandemic influenza viruses , volume =. Proceedings of the National Academy of Sciences , number =

    work page

  77. [77]

    Genomic surveillance elucidates Ebola virus origin and transmission during the 2014 outbreak , volume =

    Gire, Stephen K and Goba, Augustine and Andersen, Kristian G and Sealfon, Rachel SG and Park, Daniel J and Kanneh, Lansana and Jalloh, Simbirie and Momoh, Mambu and Fullah, Mohamed and Dudas, Gytis and others , date-added =. Genomic surveillance elucidates Ebola virus origin and transmission during the 2014 outbreak , volume =. Science , number =

    work page 2014

  78. [78]

    A spectral method for confidence interval generation and run length control in simulations , volume =

    Heidelberger, Philip and Welch, Peter D , date-added =. A spectral method for confidence interval generation and run length control in simulations , volume =. Communications of the ACM , number =

    work page

  79. [79]

    Computational molecular evolution , volume =

    Yang, Ziheng and others , date-added =. Computational molecular evolution , volume =

    work page

  80. [80]

    Evolutionary divergence and convergence in proteins , year =

    Zuckerkandl, Emile and Pauling, Linus , booktitle =. Evolutionary divergence and convergence in proteins , year =

    work page

Showing first 80 references.