Pith. sign in

REVIEW 5 major objections 5 minor 56 references

Handling Sparse Non-negative Data in Finance

T0 review · 5 major / 5 minor · reviewed 2026-08-05 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read Poisson regression can be far from optimal for sparse, heteroskedastic count-like data; a one-parameter family of moment estimators with cross-validated kappa cuts out-of-sample RMSE by 50–90% across four finance datasets.

desk verdict The kappa-family bias-variance framework is a real contribution, but the headline 50-90% RMSE improvements are optimistic because kappa is selected on the same test folds used to measure them, and two outcomes are truncated. read the letter →

arxiv 2509.01478 v1 pith:S3UFAYFL submitted 2025-09-01 econ.EM stat.ME

Agostino Capponi , Zhaonan Qu This is my paper · ORCID
classification econ.EMstat.ME MSC 62F1262J1262P20
keywords count-likedataheteroskedasticitysparsitypseudomaximumlikelihoodPoissonregressionbias-variancetrade-offcross-validationnon-negativeoutcomes
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

The paper takes aim at the near-default use of Poisson pseudo-maximum-likelihood (PPML) regression for non-negative outcomes in finance and economics. It argues that when outcomes are both sparse (many zeros) and heteroskedastic, PPML sits at a poor fixed point of a bias-variance trade-off: estimators that weight large-mean observations reduce censoring bias, while estimators that weight small-mean observations gain efficiency. The authors build a one-parameter family of moment estimators that spans Poisson, gamma PML, and non-linear least squares, and show that the best member depends on the relative strength of sparsity and heteroskedasticity. A simple cross-validation selection procedure picks the member; in simulations and four finance datasets it beats Poisson, often by large margins and with different economic conclusions.

What carries the argument

The generalized pseudo-maximum-likelihood (PML) estimating equations: (1/n) sum (y_i - exp(theta^T x_i)) exp(kappa theta^T x_i) x_i = 0, for kappa in [-1,1] (or all real kappa). They nest Poisson (kappa=0), NLS (kappa=1), and gamma PML (kappa=-1), and reduce to negative binomial when a c=1 shift is included. The exponent kappa tunes the weight given to observations with larger conditional means, and the population version of the equations, under an asymmetric censoring model where P(Y=0|X) decreases with exp(theta_0^T X), yields the bias and variance formulas that drive the trade-off; k-fold cross-validation on out-of-sample MSE then selects kappa.

What would settle it

Simulate data with Var(Y|X)=E^alpha(Y|X) but with a zero probability P(Y=0|X) that is constant or increasing in the conditional mean (for example, structural zeros from a separate binary process). If the cross-validated kappa then selects values near 0 or below, and Poisson has lower out-of-sample MSE than kappa=1, the paper's monotone-censoring-driven bias ordering is falsified.

Watch

Extended reading notes

Core claim

The central claim is that the optimal estimator in the family summing (y_i - exp(theta^T x_i)) exp(kappa theta^T x_i) x_i = 0 is often not the Poisson member (kappa=0) when zeros arise from asymmetric censoring of an underlying exponential-mean process. Bias falls as kappa rises because the exp(kappa theta^T x) weight downweights small-mean observations that are disproportionately censored; variance rises with kappa under heteroskedasticity. Theorems 1 and 2 give asymptotic bias and variance approximations for the family, showing bias independent of the heteroskedasticity index alpha and variance depending on alpha, kappa, and censoring. Cross-validation over kappa selects the MSE-minimizing

Load-bearing premise

The ordering of estimators relies on zeros being censored versions of small underlying means, so that the chance of a zero falls as the conditional mean rises; if many zeros are structural and do not decrease with the mean, the recommended kappa>0 can be wrong.

Editorial extensions

If this is right

  • In data with many zeros and mild heteroskedasticity, NLS (kappa=1) or other kappa>0 estimators can beat Poisson, reversing the common advice to avoid NLS for non-negative outcomes.
  • The same cross-validation procedure can be applied to trade, health, or other non-negative datasets; the selected kappa, not a default Poisson, should define the benchmark model.
  • Parameter estimates and standard economic conclusions can change materially with kappa (e.g., ratings coefficient -0.646 to -2.565, patents overconfidence 0.617 to 0.125), so model choice affects inference, not only prediction.
  • The phase-transition map (optimal kappa as a function of alpha and tau) gives applied researchers a diagnostic: stronger sparsity pushes optimal kappa up, stronger heteroskedasticity pushes it down.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • If excess zeros are generated by structural zeros (e.g., firms that never innovate) rather than censoring of small positive values, the monotone-zero-probability assumption fails and the recommended kappa>0 ordering may reverse; one can check whether P(Y=0|X) declines with the fitted mean before adopting the CV choice.
  • Because the uncensored variance-minimizing kappa is 1-alpha under Var(Y|X)=E^alpha(Y|X), a practitioner who estimates alpha can restrict the CV grid or interpret the selected kappa as an implied trade-off point between censoring bias and heteroskedasticity.
  • The same estimating-equation family with a general f(theta^T x) in place of exp(theta^T x) (Appendix D) could extend the bias-variance logic beyond log-linear conditional means to other nonlinear specifications.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

5 major / 5 minor

Summary. The paper argues that Poisson PML, although widely recommended for non-negative/count outcomes, can be suboptimal when both heteroskedasticity and sparsity are present. It proposes a one-parameter family of generalized PML estimators defined by the moment equations (6), which nests Poisson (κ=0), NLS (κ=1), and gamma PML (κ=-1). The paper derives approximate bias and asymptotic variance formulas under an asymmetric censoring model (3)-(4) and heteroskedasticity (2), showing a bias-variance trade-off indexed by κ. It recommends selecting κ by cross-validation and reports, in four finance datasets, that the selected estimator can reduce out-of-sample RMSE by 50-90% relative to Poisson PML.

Significance. If the results hold, the paper provides a useful framework and a practical data-driven alternative to the default use of Poisson PML. Its strengths are the explicit estimating-equation family (6), the asymptotic bias/variance analysis in Theorems 1-2, and extensive simulations (5000 replications) that convincingly demonstrate a trade-off under the assumed censoring model. The cross-validation proposal is simple and actionable. However, the headline empirical claims are currently overstated because the evaluation protocol is circular and the data pre-processing is not fully justified; these issues are fixable but require substantive revisions.

major comments (5)
  1. [§5.2] The reported 50-90% out-of-sample RMSE improvements are computed by evaluating the test-fold RMSE curve over 2000 random 80/20 splits, selecting the κ that minimizes that same curve, and comparing the minimum with the Poisson point (κ=0) on the same curve. The minimum of a noisy estimated curve is an optimistically biased estimator of the selected model's achievable RMSE; this is effectively in-sample selection. No error bars are given. Please re-evaluate with nested cross-validation (select κ on training folds, evaluate on a separate test split) and report standard errors or confidence intervals for the improvement.
  2. [§5.1 / Figure 8] The patents and permits outcomes are apparently 'truncated to samples with y≤25' and 'y≤50'. If this truncation is applied to the estimation/evaluation sample, the target is a censored outcome and the RMSE comparisons (and the optimal κ) are not for the original economic variable. The extreme 90% gain for gamma PML on permits is especially sensitive to this choice. Please clarify whether truncation is only for plotting; if it is used in the analysis, justify it or use the full sample with a loss appropriate for the actual outcome.
  3. [Appendix C, Corollary 1] The stated denominator for the NLS bias, ∫p(X)X²(1+P)exp(2θ0ᵀX), conflicts with Theorem 1, Eq. (9), which for κ=1 gives A = ∫(P−1)xxᵀp exp(2θ0ᵀX) = −∫(1−P)xxᵀp exp(2θ0ᵀX). The sign of the censoring adjustment matters for the conclusion that NLS is more robust than PPML. In addition, Lemma 1 in Appendix F.2 appears to use scalar denominators E[exp(2θ0ᵀX)] and drops the xxᵀ factor; please reconcile with Theorem 2.
  4. [§4, final paragraph] 'As heteroskedasticity decreases or sparsity increases, the optimal κ decreases' contradicts the paper's own argument (Section 3.1, Figure 7) that greater sparsity favors larger κ. If read literally, this gives the opposite practical advice. Please correct and verify the direction in Figure 7.
  5. [§2.2 / §3.2] The bias-reduction property of κ>0 is derived under the asymmetric censoring model (3)-(4), where P(Y=0|X) decreases monotonically in exp(θ0ᵀX). The paper's own Appendix E.1 notes that the bias approximation is 'less accurate for the gamma PML.' The general claim that Poisson 'can be far from optimal' is only established for this class of DGPs. Please add a robustness simulation with structural zeros (e.g., zero-inflation with P(Y=0|X) independent of or increasing in the mean) to show where the recommendation reverses, and soften the general wording accordingly.
minor comments (5)
  1. [Abstract] 'procedureselects' is missing a space.
  2. [§2.1] 'To To contextualize' should be 'To contextualize'.
  3. [§5.1] The default data description says January 2019-December 2023, but Figure 8 caption says January 2020-December 2023. Please reconcile.
  4. [Appendix E.1] The plots label κ=-0.95 as 'gamma PML'; this should be stated clearly in the text, not only in parentheses.
  5. [Appendix E.1 / Figure E.10] The censoring probability 1/(1+exp(βθ0ᵀx)) used in the appendix differs from the main specification 1/(1+(τ exp(θ0ᵀx))^β) in (4). Please clarify the relationship.

Circularity Check

1 steps flagged · score 6.0 of 10

Empirical RMSE improvements are selected and evaluated on the same test curve; the theoretical bias-variance analysis is otherwise self-contained.

  1. fitted input called prediction [Section 5.2, Figure 9]
    "we split the samples 80%–20% randomly into train and test sets. We estimate generalized PML estimators with varyingκ on the train set, and report the test set RMSEs averaged over 2000 random splittings of the data. ... For credit ratings and corporate patents, estimators withκ >0are recommended with out-of-sample RMSEs about 50% smaller than Poisson PML; for residential permits, the gamma PML is recommended with an out-of-sample RMSE that is about 90% smaller than Poisson PML."

    The κ that the paper recommends is chosen as the minimizer of the very same averaged test-set RMSE curve that is then used to measure the improvement over Poisson (κ = 0). Because the curve is estimated with finite-sample noise, the minimum is an optimistic estimate of the selected model's achievable RMSE. The reported 50–90% reductions are therefore the gap between a fitted minimum and a fixed point on the same fitted curve, so the improvement is in-sample for the selection rule. The paper has no nested or separate holdout to validate the selected κ, and two of the four outcomes are truncated, further obscuring the true effect. A proper nested evaluation is required before these out-of-sample improvement figures can be taken at face value.

full rationale

The theoretical core of the paper is not circular. The generalized PML family (6) is defined independently, and Theorems 1 and 2 derive the bias approximation and asymptotic variance from the population moment conditions and the asymmetric censoring model (3)–(4); the qualitative bias ordering in κ is a derived property of the chosen weights, not an input assumption. There is no load-bearing self-citation: the only overlapping-author citation (Chang et al. 2024) appears in a list of Poisson attributes and does not justify the central claim. The empirical claim, however, is partially circular: the same averaged test-set RMSE curve is used both to select κ and to quantify the improvement over Poisson, so the headline 50–90% reductions are statistically forced by the selection rule rather than honest out-of-sample estimates. This does not undermine the theoretical bias-variance analysis, but it means the practical benefit for the four finance datasets remains unquantified. Hence a partial circularity score of 6.

Assumptions & free parameters 5 free parameters · 6 assumptions · 0 invented entities

Everything in the theory is downstream of four modeling choices: the exponential conditional mean, the power variance model Var = E^alpha, the asymmetric censoring model (3)-(4) for sparsity, and the first-order Taylor approximation of the bias. The first two are standard; the last two are specific to this paper and are not validated against real-world zero-generating processes. The simulations treat alpha, beta, tau, the lognormal error scale, and the kappa grid as hand-chosen inputs. For the empirical claims (CV-selected kappa, 50-90% RMSE improvements), the kappa grid, the outcome scale normalization for ratings, and the outcome truncations for patents and permits are additional choices that affect the reported results. No new entities are postulated; the generalized PML family is a method family, not a postulated object.

free parameters (5)
  • kappa (estimator index) = data-selected; 0.10 (defaults), about 1 to 3 (ratings/patents), about -1 (permits)
    The index of the estimator family (6), selected by out-of-sample MSE cross-validation in the applications. In simulations the RMSE-minimizing kappa is reported as a function of (alpha, tau, beta).
  • alpha (heteroskedasticity index) = 0, 1, 2 in simulations; not estimated in applications
    Parameter of Var(Y|X) = E^alpha(Y|X) in model (2); enters the variance formula (10) and the efficiency rule kappa* = 1 - alpha. In the empirical sections the degree of heteroskedasticity is only assessed informally via box plots (Figure E.12).
  • tau (sparsity scale) = 1 or 2 in simulations
    Controls overall sparsity in the censoring probability (4); chosen by hand in the simulation design, not estimated from real data.
  • beta (sparsity contrast) = 2 in simulations
    Controls how sharply censoring decays with the conditional mean in (4); hand-chosen; Section 4.1 states beta regulates the contrast in censoring probabilities.
  • kappa grid range = [-b, b] with b >= 1 in simulations; empirical grids differ per dataset, up to 3.0 for patents
    The paper considers kappa in [-b, b] (simulations) and uses different grids in the applications; the grid choice affects the selected optimum and the reported improvement over Poisson.
assumptions (6)
  • domain assumption The conditional mean is correctly specified as E[Y|X] = exp(theta_0^T X)
    Section 2.1; consistency of every PML estimator in the family rests on this specification.
  • domain assumption Heteroskedasticity follows Var(Y|X) = E^alpha(Y|X) with alpha >= 0
    Model (2), Section 2.1; not tested on any real dataset. The empirical heteroskedasticity levels are eyeballed from box plots (Figure E.12).
  • ad hoc to paper Excess zeros arise from asymmetric censoring with P(Y=0|X) = exp(-(tau exp(theta_0^T X))^beta) or 1/(1 + (tau exp(theta_0^T X))^beta)
    Model (3)-(4), Section 2.2; this structure makes bias decrease mechanically in kappa. The paper presents no test of this structure against real zero-generating processes and does not compare with zero-inflated or hurdle fits.
  • ad hoc to paper First-order Taylor expansion of exp(theta^T X) and exp(kappa theta^T X) around theta_0 gives an accurate bias approximation
    Theorem 1 and Appendix F.1; the paper concedes the approximation is only good when censoring mass sits at small exp(theta_0^T X) and shows it is inaccurate for gamma PML (Appendix E.1, Figure E.11).
  • standard math Standard Z-estimator regularity conditions hold (compact Theta, dominated score, unique interior solution to (7))
    Theorems 1 and 2 invoke Newey and McFadden (1994) for consistency and asymptotic normality of method-of-moment estimators.
  • domain assumption Covariates have mixed signs so kappa > 0 estimating equations do not degenerate at infinity
    Appendix B: for kappa > 0, if all x_i share one sign, the moment equations can be satisfied only at theta = +/- infinity. The empirical claims therefore depend on mixed-sign covariates, which is not verified for each dataset.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Handling Sparse Non-negative Data in Finance." pith.science (2026). https://pith.science/paper/S3UFAYFL

@misc{pith2026250901478,
  author       = {Pith},
  title        = {Pith review of: Handling Sparse Non-negative Data in Finance},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/S3UFAYFL}},
  note         = {Machine review of arXiv:2509.01478}
}
read the original abstract

We show that Poisson regression, though often recommended over log-linear regression for modeling count and other non-negative variables in finance and economics, can be far from optimal when heteroskedasticity and sparsity -- two common features of such data -- are both present. We propose a general class of moment estimators, encompassing Poisson regression, that balances the bias-variance trade-off under these conditions. A simple cross-validation procedure selects the optimal estimator. Numerical simulations and applications to corporate finance data reveal that the best choice varies substantially across settings and often departs from Poisson regression, underscoring the need for a more flexible estimation framework.

Figures

Figures reproduced from arXiv: 2509.01478 by the authors.

Figure 1
Figure 1. Behavior of the optimal estimator under different levels of heteroskedasticity and sparsity. [PITH_FULL_IMAGE:figures/full_fig_p004_1.png] view at source ↗
Figure 2
Figure 2. Histograms of monthly defaults in Moody’s 35 Industrial Categories (left) and corporate patents [PITH_FULL_IMAGE:figures/full_fig_p005_2.png] view at source ↗
Figure 3
Figure 3. Bias, standard deviation, and RMSE of generalized PML estimators as a function of the [PITH_FULL_IMAGE:figures/full_fig_p017_3.png] view at source ↗
Figures from the paper (6 more)
Figure 4
Figure 4. Figure 4: Bias, standard deviation, and RMSE of generalized PML estimators as a function of the [PITH_FULL_IMAGE:figures/full_fig_p021_4.png]
Figure 5
Figure 5. Figure 5: Bias, standard deviation, and RMSE of PML estimators under censoring probability [PITH_FULL_IMAGE:figures/full_fig_p022_5.png]
Figure 6
Figure 6. Figure 6: Bias, standard deviation, and RMSE of PML estimators under censoring probability [PITH_FULL_IMAGE:figures/full_fig_p023_6.png]
Figure 7
Figure 7. Figure 7: Phase transition of the optimal κ for different levels of heteroskedasticity (α) and sparsity (τ ). Results based on simulated data with true parameter θ0 = (1, 1). the absence of censoring. However, in the presence of asymmetric censoring, the standard deviation is mi…
Figure 8
Figure 8. Figure 8: Histograms of outcome variables in finance datasets, overlayed with predictions of Poisson [PITH_FULL_IMAGE:figures/full_fig_p026_8.png]
Figure 9
Figure 9. Figure 9: Average out of sample RMSEs of generalized PML estimators on corporate defaults (top left), [PITH_FULL_IMAGE:figures/full_fig_p028_9.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

56 extracted references · 54 canonical work pages

  1. [1]

    write newline

    " write newline "" before.all 'output.state := FUNCTION n.dashify 't := "" t empty not t #1 #1 substring "-" = t #1 #2 substring "--" = not "--" * t #2 global.max substring 't := t #1 #1 substring "-" = "-" * t #2 global.max substring 't := while if t #1 #1 substring * t #2 global.max substring 't := if while FUNCTION word.in bbl.in ":" * " " * FUNCTION f...

  2. [2]

    , author Ng, D.T

    author Addoum, J.M. , author Ng, D.T. , author Ortiz-Bobea, A. , year 2023 . title Temperature shocks and industry earnings news . journal Journal of Financial Economics volume 150 , pages 1--45

    work page 2023

  3. [3]

    , author Norton, E.C

    author Ai, C. , author Norton, E.C. , year 2000 . title Standard errors for the retransformation problem with heteroscedasticity . journal Journal of Health Economics volume 19 , pages 697--718

    work page 2000

  4. [4]

    , author Appel, I

    author Akey, P. , author Appel, I. , year 2021 . title The limits of limited liability: Evidence from industrial pollution . journal The Journal of Finance volume 76 , pages 5--55

    work page 2021

  5. [5]

    , year 1985

    author Amemiya, T. , year 1985 . title Advanced econometrics . journal Harvard University Press volume 2 , pages 153--161

    work page 1985

  6. [6]

    , year 1979

    author Anderson, J.E. , year 1979 . title A theoretical foundation for the gravity equation . journal American Economic Review volume 69 , pages 106--116

    work page 1979

  7. [7]

    , author Van Wincoop, E

    author Anderson, J.E. , author Van Wincoop, E. , year 2003 . title Gravity with gravitas: A solution to the border puzzle . journal American Economic Review volume 93 , pages 170--192

    work page 2003

  8. [8]

    , author Cohen, M.C

    author Bekkerman, R. , author Cohen, M.C. , author Kung, E. , author Maiden, J. , author Proserpio, D. , year 2023 . title The effect of short-term rentals on residential investment . journal Marketing Science volume 42 , pages 819--834

    work page 2023

Show all 56 references
  1. [9]

    , author Dietz, E

    author B \"o hning, D. , author Dietz, E. , author Schlattmann, P. , author Mendonca, L. , author Kirchner, U. , year 1999 . title The zero-inflated poisson model and the decayed, missing and filled teeth index in dental epidemiology . journal Journal of the Royal Statistical ...

    1999

  2. [10]

    , author Williford, A

    author Boulton, A.J. , author Williford, A. , year 2018 . title Analyzing skewed continuous outcomes with many zeros: A tutorial for social work and youth prevention science researchers . journal Journal of the Society for Social Work and Research volume 9 , pages 721--740

    2018

  3. [11]

    , author Trivedi, P.K

    author Cameron, A.C. , author Trivedi, P.K. , year 1990 . title Regression-based tests for overdispersion in the poisson model . journal Journal of Econometrics volume 46 , pages 347--364

    1990

  4. [12]

    , author Trivedi, P.K

    author Cameron, A.C. , author Trivedi, P.K. , year 2013 . title Regression Analysis of Count Data . volume volume 53 . publisher Cambridge University Press

    2013

  5. [13]

    , year 2001

    author Card, D. , year 2001 . title Estimating the return to schooling: Progress on some persistent econometric problems . journal Econometrica volume 69 , pages 1127--1160

    2001

  6. [14]

    , author Koehler, F

    author Chang, S. , author Koehler, F. , author Qu, Z. , author Leskovec, J. , author Ugander, J. , year 2024 . title Inferring dynamic networks from marginals with iterative proportional fitting . journal arXiv preprint arXiv:2402.18697

    2024 arXiv

  7. [15]

    , author Roth, J

    author Chen, J. , author Roth, J. , year 2024 . title Logs with zeros? some problems and solutions . journal The Quarterly Journal of Economics volume 139 , pages 891--936

    2024

  8. [16]

    , author Liu, Z

    author Cohn, J.B. , author Liu, Z. , author Wardlaw, M.I. , year 2022 . title Count (and count-like) data in finance . journal Journal of Financial Economics volume 146 , pages 529--551

    2022

  9. [17]

    , author Daniel, N.D

    author Coles, J.L. , author Daniel, N.D. , author Naveen, L. , year 2006 . title Managerial incentives and risk-taking . journal Journal of Financial Economics volume 79 , pages 431--468

    2006

  10. [18]

    , author Guimar \ a es, P

    author Correia, S. , author Guimar \ a es, P. , author Zylkin, T. , year 2020 . title Fast poisson estimation with high-dimensional fixed effects . journal The Stata Journal volume 20 , pages 95--115

    2020

  11. [19]

    , author Von Peter, G

    author Craig, B. , author Von Peter, G. , year 2014 . title Interbank tiering and money center banks . journal Journal of Financial Intermediation volume 23 , pages 322--347

    2014

  12. [20]

    , author Norton, E.C

    author Dow, W.H. , author Norton, E.C. , year 2003 . title Choosing between and interpreting the heckit and two-part models for corner solutions . journal Health Services and Outcomes Research Methodology volume 4 , pages 5--18

    2003

  13. [21]

    , author Saita, L

    author Duffie, D. , author Saita, L. , author Wang, K. , year 2007 . title Multi-period corporate default prediction with stochastic covariates . journal Journal of Financial Economics volume 83 , pages 635--665

    2007

  14. [22]

    , author Tian, X

    author Fang, V.W. , author Tian, X. , author Tice, S. , year 2014 . title Does stock liquidity enhance or impede firm innovation? journal The Journal of Finance volume 69 , pages 2085--2125

    2014

  15. [23]

    , year 1997

    author Frankel, J.A. , year 1997 . title Regional trading blocs in the world economic system . journal Institute for International Economics

    1997

  16. [24]

    , author Wei, S.J

    author Frankel, J.A. , author Wei, S.J. , year 1993 . title Trade blocs and currency blocs . number 11 , publisher National Bureau of Economic Research

    1993

  17. [25]

    , author Monfort, A

    author Gourieroux, C. , author Monfort, A. , author Trognon, A. , year 1984 a. title Pseudo maximum likelihood methods: Applications to poisson models . journal Econometrica: Journal of the Econometric Society , pages 701--720

    1984

  18. [26]

    , author Monfort, A

    author Gourieroux, C. , author Monfort, A. , author Trognon, A. , year 1984 b. title Pseudo maximum likelihood methods: Theory . journal Econometrica: Journal of the Econometric Society , pages 681--700

    1984

  19. [27]

    , author Lawless, A.S

    author Gratton, S. , author Lawless, A.S. , author Nichols, N.K. , year 2007 . title Approximate gauss--newton methods for nonlinear least squares problems . journal SIAM Journal on Optimization volume 18 , pages 106--132

    2007

  20. [28]

    , year 1994

    author Greene, W.H. , year 1994 . title Accounting for excess zeros and sample selection in poisson and negative binomial regression models . journal NYU working paper no. EC-94-10

    1994

  21. [29]

    , author Hall, B.H

    author Hausman, J. , author Hall, B.H. , author Griliches, Z. , year 1984 . title Econometric models for count data with an application to the patents-r & d relationship . journal Econometrica: Journal of the Econometric Society , pages 909--938

    1984

  22. [30]

    , author Tian, X

    author He, J.J. , author Tian, X. , year 2013 . title The dark side of analyst coverage: The case of innovation . journal Journal of Financial Economics volume 109 , pages 856--878

    2013

  23. [31]

    , author Vana, L

    author Hirk, R. , author Vana, L. , author Hornik, K. , year 2022 . title A corporate credit rating model with autoregressive errors . journal Journal of Empirical Finance volume 69 , pages 224--240

    2022

  24. [32]

    , author Low, A

    author Hirshleifer, D. , author Low, A. , author Teoh, S.H. , year 2012 . title Are overconfident ceos better innovators? journal The Journal of Finance volume 67 , pages 1457--1498

    2012

  25. [33]

    , year 1980

    author Holford, T.R. , year 1980 . title The analysis of rates and of survivorship using log-linear models . journal Biometrics volume 36 , pages 299--305

    1980

  26. [34]

    , author Karbownik, K

    author Hollingsworth, A. , author Karbownik, K. , author Thomasson, M.A. , author Wray, A. , year 2024 . title The gift of a lifetime: The hospital, modern medicine, and mortality . journal American Economic Review volume 114 , pages 2201--2238

    2024

  27. [35]

    , year 1988

    author King, G. , year 1988 . title Statistical models for political science event counts: Bias in conventional procedures and evidence for the exponential poisson regression model . journal American Journal of Political Science , pages 838--863

    1988

  28. [36]

    , year 1989

    author King, G. , year 1989 . title Event count models for international relations: Generalizations and applications . journal International Studies Quarterly volume 33 , pages 123--147

    1989

  29. [37]

    , year 1992

    author Lambert, D. , year 1992 . title Zero-inflated poisson regression, with an application to defects in manufacturing . journal Technometrics volume 34 , pages 1--14

    1992

  30. [38]

    , year 1998

    author Manning, W.G. , year 1998 . title The logged dependent variable, heteroscedasticity, and the retransformation problem . journal Journal of Health Economics volume 17 , pages 283--295

    1998

  31. [39]

    , author Mullahy, J

    author Manning, W.G. , author Mullahy, J. , year 2001 . title Estimating log models: to transform or not to transform? journal Journal of Health Economics volume 20 , pages 461--494

    2001

  32. [40]

    , year 2019

    author McCullagh, P. , year 2019 . title Generalized linear models . publisher Routledge

    2019

  33. [41]

    , year 1986

    author Mullahy, J. , year 1986 . title Specification and testing of some modified count data models . journal Journal of Econometrics volume 33 , pages 341--365

    1986

  34. [42]

    , year 1997

    author Mullahy, J. , year 1997 . title Instrumental-variable estimation of count data models: Applications to models of cigarette smoking behavior . journal Review of Economics and Statistics volume 79 , pages 586--593

    1997

  35. [43]

    , year 1998

    author Mullahy, J. , year 1998 . title Much ado about two: reconsidering retransformation and the two-part model in health econometrics . journal Journal of Health Economics volume 17 , pages 247--281

    1998

  36. [44]

    , author Norton, E.C

    author Mullahy, J. , author Norton, E.C. , year 2022 . title Why transform Y? A critical assessment of dependent-variable transformations in regression models for skewed and sometimes-zero outcomes . type Technical Report . National Bureau of Economic Research

    2022

  37. [45]

    , author McFadden, D

    author Newey, W.K. , author McFadden, D. , year 1994 . title Large sample estimation and hypothesis testing . journal Handbook of econometrics volume 4 , pages 2111--2245

    1994

  38. [46]

    , year 1966

    author Park, R.E. , year 1966 . title Estimation with heteroscedastic error terms. journal Econometrica volume 34

    1966

  39. [47]

    , author Tenreyro, S

    author Santos Silva, J. , author Tenreyro, S. , year 2006 . title The log of gravity . journal Review of Economics and Statistics volume 88 , pages 641--658

    2006

  40. [48]

    , author Tenreyro, S

    author Santos Silva, J. , author Tenreyro, S. , year 2010 . title On the existence of the maximum likelihood estimates in poisson regression . journal Economics Letters volume 107 , pages 310--312

    2010

  41. [49]

    , author Tenreyro, S

    author Santos Silva, J. , author Tenreyro, S. , year 2011 . title Further simulation evidence on the performance of the poisson pseudo-maximum likelihood estimator . journal Economics Letters volume 112 , pages 220--222

    2011

  42. [50]

    , author Van Lent, L

    author Sautner, Z. , author Van Lent, L. , author Vilkov, G. , author Zhang, R. , year 2023 . title Firm-level climate change exposure . journal The Journal of Finance volume 78 , pages 1449--1498

    2023

  43. [51]

    , author Xing, Y

    author Vassalou, M. , author Xing, Y. , year 2004 . title Default risk in equity returns . journal The Journal of Finance volume 59 , pages 831--868

    2004

  44. [52]

    , year 1982

    author White, H. , year 1982 . title Maximum likelihood estimation of misspecified models . journal Econometrica: Journal of the Econometric Society , pages 1--25

    1982

  45. [53]

    , author Santos Silva, J.M

    author Windmeijer, F.A. , author Santos Silva, J.M. , year 1997 . title Endogeneity in count data models: an application to demand for health care . journal Journal of Applied Econometrics volume 12 , pages 281--294

    1997

  46. [54]

    , year 1999

    author Wooldridge, J.M. , year 1999 . title Quasi-likelihood methods for count data . journal Handbook of applied econometrics volume 2: Microeconomics , pages 321--368

    1999

  47. [55]

    , author Kim, T

    author Xu, Q. , author Kim, T. , year 2022 . title Financial constraints and corporate environmental policies . journal The Review of Financial Studies volume 35 , pages 576--635

    2022

  48. [56]

    , year 1998

    author Zorn, C.J. , year 1998 . title An analytic and empirical examination of zero-inflated and hurdle poisson specifications . journal Sociological Methods & Research volume 26 , pages 368--400

    1998

Pith tools

Reviewed August 5, 2026 · model on record in the stance chip above.