Pith. sign in

REVIEW 4 major objections 6 minor 45 references

Propensity score with factor loadings: the effect of the Paris Agreement

T0 review · 4 major / 6 minor · reviewed 2026-08-06 · deepseek-v4-flash

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

Pith's one-line read The Paris Agreement lowered green-bond issuers' Q1 2016 returns by an estimated 0.071.

desk verdict A useful new IPW estimator for factor-loading panels, but a sign error in the influence function invalidates the reported standard errors as written. read the letter →

arxiv 2507.08764 v1 pith:23Z5DZCV submitted 2025-07-11 econ.EM stat.AP

Angelo Forino , Andrea Mercatanti , Giacomo Morelli This is my paper
classification econ.EMstat.AP MSC 62P2062F1262H25
keywords ParisAgreementgreenbondspropensityscorefactormodelpaneldataM-estimationaveragetreatmenteffectonthetreatedsuper-populationinference
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

This paper proposes a causal inference method for panel data in which treatment assignment is assumed ignorable once a low-dimensional set of latent factor loadings is conditioned on. The loadings are estimated from pre-treatment outcomes by principal component analysis, converted into a logistic propensity score, and used in an inverse-probability-weighted estimator of the average treatment effect on the treated. The authors derive the estimator's large-sample distribution through M-estimation, producing closed-form standard errors that incorporate uncertainty from the loadings, the propensity score, and the weighting step. Applied to S&P 350 Europe constituents, the method estimates that the Paris Agreement reduced first-quarter 2016 returns of firms that issued green bonds in 2016-2019 by 0.0710, with a standard error of 0.0282. This negative short-run effect supports the interpretation that investors repriced green assets as lower-risk after the Agreement.

What carries the argument

The central object is the propensity score $e_i(\lambda_i) = 1/(1+\exp(-\lambda_i' \beta))$, where $\lambda_i$ is the unit-specific vector of loadings on $r$ common factors in the pre-treatment outcome model $Y_{i,t}(0) = \lambda_i' F_t + \xi_{i,t}$. The loadings are the carriers of unobserved confounding: Assumption 3.4 says treatment is independent of potential outcomes given $\lambda_i$. PCA on pre-treatment outcomes identifies the loadings; logistic regression turns them into a propensity score; the Hajek estimator weights control outcomes by $e_i/(1-e_i)$; and M-estimation stacks the PCA, logistic, and weighting estimating equations so that the asymptotic variance accounts for all three sources of uncertainty at once. The identifying normalization $F'F/T_0 = I_r$ with diagonal $\Lambda'\Lambda$ removes rotational indeterminacy and makes PCA consistent.

What would settle it

Re-estimate the propensity score adding observable pre-treatment covariates such as firm size, sector, and past volatility; if any of them still predicts green-bond issuance given the estimated loadings, latent ignorability fails and the ATT is suspect. A complementary check compares the Q1 2016 estimate for issuers from 2016-2017 with issuers from 2018-2019: under the paper's timing assumption both groups should show similar effects, and a large divergence would indicate the treatment date is mismeasured.

Watch

Extended reading notes

Core claim

On the paper's own terms, the central claim is Theorem 3.1: under Assumptions 3.1-3.4 and standard approximate-factor conditions, as $N, T_0 \to \infty$ with $\sqrt{N}/T_0 \to 0$, the Hajek IPW estimator satisfies $\sqrt{N}(\hat{\tau}_{ATT} - \tau_{ATT}) \to N(0, V_{ATT})$, where $V_{ATT}$ has a closed-form M-estimation influence-function representation that includes the sampling randomness, the PCA estimation of loadings, the logistic estimation of the propensity score, and the counterfactual missingness. This turns a three-step procedure into a single inference problem with a sandwich variance. The empirical discovery carried by the method is a statistically significant negative estimate of the Paris Agreement's short-run effect: $-0.0710$ (SE $0.0282$) on the Q1 2016 returns of treated green-bond issuers, with placebo-date falsification tests showing null results.

Load-bearing premise

The load-bearing premise is that a firm's decision to issue a green bond after the Paris Agreement is unrelated to how its stock would have moved in Q1 2016, once its pre-2016 pattern of ups and downs is captured by the latent factors; if the Agreement moved issuance decisions and returns together beyond what those factors capture, the estimated effect is biased.

Editorial extensions

If this is right

  • Standard errors for the ATT can be computed in closed form, so practitioners do not need bootstrap or jackknife procedures to cover loadings estimation, propensity score estimation, and outcome weighting.
  • The identifying assumption is latent ignorability rather than parallel trends or homogeneous effects, so the method applies where treated and control units follow different pre-treatment trajectories.
  • The estimated ATT of $-0.0710$ (SE $0.0282$) implies that green-bond-issuing European firms experienced lower stock returns in Q1 2016 than they would have without the Paris Agreement.
  • Falsification tests using fictitious policy dates in 2011, 2013, and 2015 produce estimates close to zero with high p-values, supporting the attribution of the main effect to the Paris Agreement.
  • Simulations show the weighted distribution of loadings is balanced and confidence-interval coverage is near nominal, except in the most imbalanced, high-variability design where coverage falls to roughly 0.80.

Reading between the lines

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

  • The same M-estimation template could be applied to other balancing weights, such as overlap weights or matching, or to a propensity score that also includes observable covariates; these variants are not explored in the paper.
  • Because treated units are defined by green bond issuance over 2016-2019 but are treated as of Q1 2016, the estimate applies to firms whose environmental commitment became visible after the Agreement; a dynamic treatment version with issuance timing as the treatment would require assumptions beyond Theorem 3.1.
  • The negative return is consistent with a green-premium story, but the design cannot separate risk-based repricing from preference-based demand; adding trading volume or bid-ask spread data to the same estimator could test the mechanism.
  • The choice of $r=3$ factors follows from IC criteria, and the paper does not report sensitivity of the ATT to $r$; rerunning the procedure with $r=2$ or $r=4$ would be a direct robustness check.
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

4 major / 6 minor

Summary. The paper proposes an inverse-propensity-score weighting estimator for the average treatment effect on the treated (ATT) in panel data, where the propensity score is a logistic function of unit-specific factor loadings estimated by principal components analysis. The authors outline a three-step estimation procedure (PCA loadings, logistic propensity score, Hajek ATT estimator), derive a claimed asymptotic normality result in Theorem 3.1 with a closed-form M-estimation variance, evaluate the method in a simulation study with a known true ATT against the Generalized Synthetic Control method, and apply it to estimate the short-run effect of the Paris Agreement on the stock returns of European firms that issued green bonds, reporting a statistically significant negative ATT. The paper also includes falsification tests using fictitious policy dates to support the empirical claim.

Significance. If the proposed method and its inference are correct, the paper would be a useful contribution to causal panel-data methods: it offers closed-form standard errors without bootstrap or jackknife, relaxes parallel-trends assumptions via latent-factor unconfoundedness, and provides a super-population interpretation that strengthens external validity. The simulation design with a known true ATT and comparison to GSC is a genuine strength, as is the explicit treatment of loadings estimation uncertainty in the influence function. However, a sign error in the derivation of the influence function invalidates the reported variance formula and the standard error used in the application, and the empirical treatment definition (green bond issuance during 2016-2019 for a Q1 2016 outcome) raises a post-treatment selection concern. These issues are load-bearing for the paper's central claims, although the theoretical framework is plausibly fixable.

major comments (4)
  1. [Appendix A.1, Eq. (A.11) and Theorem 3.1] There is a sign error in the derivative of U_i(τ0, β) with respect to τ0. Since U_i(τ0, β) = (1-Z_i,T)(e_i/(1-e_i))(Y_i,T - τ0), the correct derivative is -(1-Z_i,T)e_i/(1-e_i). The paper instead writes -e_i/(1-e_i)(Z_i,T-1); because Z_i,T-1 = -(1-Z_i,T), the printed expression equals +e_i/(1-e_i)(1-Z_i,T), which has the opposite sign. This error propagates into Eq. (A.12), Eq. (A.13), and the influence function I_i in Theorem 3.1: the terms η2^{-1}U_i(τ0,β) and η2^{-1}H_β^T[...] enter with the opposite sign from the correct expansion. Since η2 = E[e/(1-e)(Z-1)] = -E[(1-Z)e/(1-e)], the incorrect sign changes the covariance structure of I_i, so V_ATT in Eq. (3.4) is not the variance of the implemented Hajek estimator, and the standard error 0.0282 reported in Section 5 is not justified.
  2. [Section 5, first paragraph] The treatment indicator is defined by green bond issuance during 2016-2019, while the outcome is Q1 2016 stock returns, making Z_i a post-treatment variable. Assumption 3.4 (latent ignorability) requires {Y_iT(1), Y_iT(0)} ⊥ Z_iT | λ_i, but if the decision to issue a green bond after 2016 reflects shocks after the Paris Agreement that also affect Q1 2016 returns beyond what the pre-treatment factor loadings capture, then the unconfoundedness assumption fails. The paper states an identifying assumption that the issuance decision reflects strategic adjustments initiated shortly after the Agreement, but this is not tested, and the falsification tests in Section 5.1 cannot validate it because they use the same post-treatment issuance definition. The empirical ATT estimate is therefore not a credible causal effect under the stated assumptions.
  3. [Theorem 3.1, Eq. (3.4)] The variance expression V_ATT = N^{-2} Σ I_i^2 + op(1) is dimensionally wrong as a companion to the stated CLT. If √N(τ̂_ATT - τ_ATT) = N^{-1/2} Σ I_i + op(1), then the natural estimator of the asymptotic variance is N^{-1} Σ I_i^2, not N^{-2} Σ I_i^2. As printed, the formula converges to zero, which would imply a degenerate standard error. The authors should clarify whether this is a typo and state the correct scaling, because the reported standard error depends on this choice.
  4. [Section 4, Tables 1] The simulation results report near-nominal coverage in most scenarios despite the sign error in the theoretical variance. This suggests the coverage numbers may have been computed with a different formula or a corrected implementation; otherwise nominal coverage with the erroneous formula would be unexpected except in special cases where the covariances that distinguish the sign-flipped influence from the correct one happen to vanish. The authors should reconcile the simulation coverage with the theorem as stated and re-run the simulations after correcting the sign error to verify that the reported coverage and RMSE remain valid.
minor comments (6)
  1. [Introduction] There are several typographical errors, including "assigmnent" for "assignment" and "H´ajek" for "Hájek" in Section 3.2; these should be corrected.
  2. [Section 4.1] The phrase "Inverse on the Propernsity score Weighting" contains a typo ("Propernsity" should be "Propensity"), and the same error appears in the table captions.
  3. [Equation (3.3)] The formula for τ̂0 is missing a division symbol and a closing parenthesis: the denominator should be clearly written as Σ(1-Z_i,T)(ê_i/(1-ê_i)) to avoid confusion.
  4. [Appendix A.1] The heading "Proof of Theoreom 3.1" contains a typo, and the phrase "the information matrices" should be singular, "the information matrix," since E_{ββ} is a single Hessian matrix.
  5. [Section 5] The sentence containing "S&P350 Europe 5 constituents" has a stray "5" that should be removed.
  6. [Lemma A.1] The rate condition stated as "for N, T0 → ∞ with √N/T0 → 0" is unusually strong; the standard Bai and Ng condition for loadings asymptotics is √T0/N → 0. The authors should clarify the intended rate condition and its role in the proof.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: the ATT is estimated from data by a standard Hajek weighting scheme; no fitted parameter is renamed as a prediction.

full rationale

The paper's central derivation (Theorem 3.1) uses M-estimation to obtain the asymptotic variance of the Hajek estimator in (3.3). The estimating equations for tau1 and tau0 are exactly the moment conditions that define the estimator, so deriving the influence function from them is standard asymptotic theory, not a circular reduction. The propensity score and factor loadings are fitted, but the ATT is not a parameter of the propensity-score or factor model; it is a weighted average of observed outcomes. The simulation exercises use a known true ATT and compare against the Generalized Synthetic Control method, providing external benchmarks that do not rely on the paper's own fitted values. No assumptions are defined in terms of the target effect, and no load-bearing claim is justified solely by a self-citation (the paper does not cite its own authors). The appendix contains a sign issue in the derivative printed as (A.11), but that is a mathematical correctness concern, not a circularity. The treatment definition and latent ignorability assumption are substantive identification choices, and while the post-treatment definition of treatment is debatable, it does not make the derivation circular.

Assumptions & free parameters 3 free parameters · 7 assumptions · 0 invented entities

The central claim rests on the factor model, latent ignorability, logistic propensity specification, and the Bai-Ng regularity conditions. No new entities (particles, forces) are introduced.

free parameters (3)
  • Number of factors r = 3 (selected by IC1/IC2 in application; set to 3 in simulation)
    The ATT estimate depends on the number of factors extracted from pre-treatment returns via PCA; the choice r=3 is data-driven.
  • Propensity score coefficients β = (-3.02, -8.06, -8.27, -17.03)
    Estimated by MLE logistic regression and used to compute weights; these fitted values directly determine the ATT estimate.
  • Factor loadings Λ = Estimated via PCA, values not reported
    Estimated loadings are the covariates in the propensity score and the basis for latent ignorability.
assumptions (7)
  • domain assumption Assumption 3.1: SUTVA (no interference, no hidden treatment versions)
    Standard causal inference assumption stated in Section 3.1.
  • domain assumption Assumption 3.2: Y_i,t(0) = λ_i' F_t + ξ_it (linear interactive fixed-effects outcome model)
    Postulated factor structure for pre-treatment outcomes, Section 3.1.
  • domain assumption Assumption 3.3: PC1 normalization F'F/T0 = I_r and Λ'Λ diagonal with distinct entries
    Ensures uniqueness of PCA solution, Section 3.1.
  • domain assumption Assumption 3.4: Latent ignorability {Y_T(0), Y_T(1)} ⊥ Z_T | λ_i with 0 < P(Z_T|λ_i) < 1
    Core identifying assumption; Section 3.1.
  • standard math Standard approximate factor model assumptions from Bai and Ng (2013)
    Invoked for consistency and asymptotic normality of PCA loadings, Sections 3.2 and A.1.
  • domain assumption Logistic form of the propensity score e(λ_i) = (1+exp(-λ_i' β))^{-1}
    Specified in Section 3.2, equation (3.2).
  • standard math Asymptotic condition N, T0 → ∞ with √N/T0 → 0
    Required for the limiting result in Theorem 3.1; not clearly satisfied in the empirical application.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Propensity score with factor loadings: the effect of the Paris Agreement." pith.science (2026). https://pith.science/paper/23Z5DZCV

@misc{pith2026250708764,
  author       = {Pith},
  title        = {Pith review of: Propensity score with factor loadings: the effect of the Paris Agreement},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/23Z5DZCV}},
  note         = {Machine review of arXiv:2507.08764}
}
read the original abstract

Factor models for longitudinal data, where policy adoption is unconfounded with respect to a low-dimensional set of latent factor loadings, have become increasingly popular for causal inference. Most existing approaches, however, rely on a causal finite-sample approach or computationally intensive methods, limiting their applicability and external validity. In this paper, we propose a novel causal inference method for panel data based on inverse propensity score weighting where the propensity score is a function of latent factor loadings within a framework of causal inference from super-population. The approach relaxes the traditional restrictive assumptions of causal panel methods, while offering advantages in terms of causal interpretability, policy relevance, and computational efficiency. Under standard assumptions, we outline a three-step estimation procedure for the ATT and derive its large-sample properties using Mestimation theory. We apply the method to assess the causal effect of the Paris Agreement, a policy aimed at fostering the transition to a low-carbon economy, on European stock returns. Our empirical results suggest a statistically significant and negative short-run effect on the stock returns of firms that issued green bonds.

Figures

Figures reproduced from arXiv: 2507.08764 by the authors.

Figure 1
Figure 1. Boxplots of the absolute standardized difference, between treated and control units, [PITH_FULL_IMAGE:figures/full_fig_p011_1.png] view at source ↗
Figure 2
Figure 2. Boxplots of the absolute standardized difference, between treated and control units, [PITH_FULL_IMAGE:figures/full_fig_p012_2.png] view at source ↗
Figure 3
Figure 3. Histograms of the estimated propensity score for the treated (blue) and the untreated [PITH_FULL_IMAGE:figures/full_fig_p015_3.png] view at source ↗
Figures from the paper (3 more)
Figure 4
Figure 4. Figure 4: Histograms of the estimated propensity score for the treated (blue) and the untreated [PITH_FULL_IMAGE:figures/full_fig_p021_4.png]
Figure 5
Figure 5. Figure 5: Histograms of the estimated propensity score for the treated (blue) and the untreated [PITH_FULL_IMAGE:figures/full_fig_p021_5.png]
Figure 6
Figure 6. Figure 6: Histograms of the estimated propensity score for the treated (blue) and the untreated [PITH_FULL_IMAGE:figures/full_fig_p022_6.png]

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

45 extracted references · 44 canonical work pages

  1. [1]

    Amjad, D

    M. Amjad, D. Shah, and D. Shen. Robust synthetic control. The Journal of Machine Learning Research, 19 0 (1): 0 802--852, 2015

    work page 2015

  2. [2]

    Estimating the effects of a California gun control program with Multitask Gaussian Processes

    D. Arbour, E. Ben-Michael, A. Feller, A. Franks, and S. Raphael. Using multitask gaussian processes to estimate the effect of a targeted effort to remove firearms. arXiv:2110.07006[stat-ME], 2021

    work page Pith review arXiv 2021

  3. [3]

    D. Y. Arkhangelsky and G. Imbens. Causal models for longitudinal and panel data: A survey. Available at arXiv:2311.15458v1 [econ.EM], 2023

    work page Pith review arXiv 2023

  4. [4]

    Athey, M

    S. Athey, M. Bayati, N. Doudchenko, G. Imbens, and K. Khosravi. Matrix completion methods for causal panel data models. Journal of the American Statistical Association, 116 0 (536): 0 1716--1730, 2021

    work page 2021

  5. [5]

    Bai and S

    J. Bai and S. Ng. Determining the number of factors in approximate factor models. Econometrica, 70 0 (1): 0 191--221, 2002

    work page 2002

  6. [6]

    Bai and S

    J. Bai and S. Ng. Confidence intervals for diffusion index forecasts and inference for factor-augmented regressions. Econometrica, 74 0 (4): 0 1133--1150, 2006

    work page 2006

  7. [7]

    Bai and S

    J. Bai and S. Ng. Principal components estimation and identification of static factors. Journal of Econometrics, 176 0 (1): 0 18--29, 2013

    work page 2013

  8. [8]

    U. S. Bhutta, A. Tariq, M. Farrukh, A. Raza, and M. K. Iqbal. Green bonds for sustainable development: Review of literature on development and impact of green bonds. Technological Forecasting and Social Change, 175: 0 121378, 2022

    work page 2022

Show all 45 references
  1. [9]

    Bolton and M

    P. Bolton and M. Kacperczyk. Do investors care about carbon risk? Journal of Financial Economics, 142 0 (2): 0 517--549, 2021 a

    2021

  2. [10]

    Bolton and M

    P. Bolton and M. Kacperczyk. Global pricing of carbon-transition risk. The Journal of Finance, 78 0 (6): 0 3677--3754, 2023

    2023

  3. [11]

    Bolton and M

    P. Bolton and M. T. Kacperczyk. Carbon disclosure and the cost of capital. Available at SSRN 3755613, 2021 b

    2021

  4. [12]

    Bolton, Z

    P. Bolton, Z. Halem, and M. Kacperczyk. The financial cost of carbon. Journal of Applied Corporate Finance, 34 0 (2): 0 17--29, 2022

    2022

  5. [13]

    Brown and K

    N. Brown and K. Butts. Dynamic treatment effect estimation with interactive fixed effects and short panels. University of Colorado Boulder, Discussion Papers in Economics, Working Paper No. 23-05, 2023

    2023

  6. [14]

    Callaway and S

    B. Callaway and S. Karami. Treatment effects in interactive fixed effects models with a small number of time periods. Journal of Econometrics, 233: 0 184--208, 2023

    2023

  7. [15]

    Chernozhukov, W

    V. Chernozhukov, W. K., and Y. Zhu. An exact and robust conformal inference method for counterfactual and synthetic controls. Journal of the American Statistical Association, 116: 0 1849--1864, 2021

    2021

  8. [16]

    De Angelis, P

    T. De Angelis, P. Tankov, and O. D. Zerbib. Climate impact investing. Management Science, 69 0 (12): 0 7669--7692, 2023

    2023

  9. [17]

    H. Fahmy. The rise in investors’ awareness of climate risks after the paris agreement and the clean energy-oil-technology prices nexus. Energy Economics, 106: 0 105738, 2022

    2022

  10. [18]

    J. Fan, R. Masini, and M. C. Medeiros. Do we exploit all information for counterfactual analysis? benefits of factor models and idiosyncratic correction. Journal of the American Statistical Association, 117: 0 574--590, 2022

    2022

  11. [19]

    C. Flammer. Corporate green bonds. Journal of Financial Economics, 142 0 (2): 0 499--516, 2021

    2021

  12. [20]

    Gobillon and T

    L. Gobillon and T. Magnac. Regional policy evaluation: Interactive fixed effects and synthetic controls. Review of Economics and Statistics, 98 0 (3): 0 535--551, 2016

    2016

  13. [21]

    Hirano, G

    K. Hirano, G. W. Imbens, and G. Ridder. Efficient estimation of average treatment effects using the estimated propensity score. Econometrica, 71 0 (4): 0 1161--1189, 2003

    2003

  14. [22]

    Hsiao, S

    C. Hsiao, S. Ching, and S. K. Wan. A panel data approach for program evaluation: Measuring the benefits of political and economic integration of hong kong with mainland china. Journal of Applied Econometrics, 27 0 (2): 0 705--740, 2012

    2012

  15. [23]

    P.-H. Hsu, K. Li, and C.-Y. Tsou. The pollution premium. The Journal of Finance, 78 0 (3): 0 1343--1392, 2023

    2023

  16. [24]

    Imbens, N

    G. Imbens, N. Kallus, and X. Mao. Controlling for unmeasured confounding in panel data using minimal bridge functions: From two-way fixed effects to factor models. arXiv:2108.03849v1[stat.ME], 2021

    2021 arXiv

  17. [25]

    G. W. Imbens and D. B. Rubin. Causal Inference for Statistics, Social, and Biomedical Sciences: An Introduction. Cambridge University Press, 2015

    2015

  18. [26]

    K. T. Li and D. R. Bell. Estimation of average treatment effects with panel data: Asymptotic theory and implementation. Journal of Econometrics, 197: 0 65--75, 2017

    2017

  19. [27]

    K. T. Li and G. P. Sonnier. Statistical inference for the factor model approach to estimate causal effects in quasi-experimental settings. Journal of Marketing Research, 60 0 (3): 0 449--472, 2023

    2023

  20. [28]

    K. V. Lins, H. Servaes, and A. Tamayo. Social capital, trust, and firm performance: The value of corporate social responsibility during the financial crisis. Journal of Finance, 72 0 (4): 0 1785--1824, 2017

    2017

  21. [29]

    L. Liu, Y. Wang, and Y. Xu. A practical guide to counterfactual estimators for causal inference with time-series cross-sectional data. American Journal of Political Science, 68 0 (1): 0 160--176, 2024

    2024

  22. [30]

    J. K. Lunceford and M. Davidian. Stratification and weighting via the propensity score in estimation of causal treatment effects: a comparative study. Statistics in Medicine, 23 0 (19): 0 2937--2960, 2004

    2004

  23. [31]

    Monasterolo and L

    I. Monasterolo and L. De Angelis. Blind to carbon risk? an analysis of stock market reaction to the paris agreement. Ecological Economics, 170: 0 106571, 2020

    2020

  24. [32]

    Mobilising bond markets for a low-carbon transition

    OECD. Mobilising bond markets for a low-carbon transition. OECD Publications Centre, 2017

    2017

  25. [33]

    X. Pang, L. Liu, and Y. Xu. A bayesian alternative to synthetic control for comparative case studies. Political Analysis, 30 0 (2): 0 269--288, 2022

    2022

  26. [34]

    P \'a stor, R

    L. P \'a stor, R. F. Stambaugh, and L. A. Taylor. Sustainable investing in equilibrium. Journal of Financial Economics, 142 0 (2): 0 550--571, 2021

    2021

  27. [35]

    L. H. Pedersen, S. Fitzgibbons, and L. Pomorski. Responsible investing: The esg-efficient frontier. Journal of Financial Economics, 142 0 (2): 0 572--597, 2021

    2021

  28. [36]

    Rohleder, M

    M. Rohleder, M. Wilkens, and J. Zink. The effects of mutual fund decarbonization on stock prices and carbon emissions. Journal of Banking & Finance, 134: 0 106352, 2022

    2022

  29. [37]

    P. R. Rosenbaum. Observational Studies. Springer Series in Statistics, 2002

    2002

  30. [38]

    D. B. Rubin. Estimating causal effects of treatments in randomized and nonrandomized studies. Journal of Educational Psychology, 66: 0 688--701, 1974

    1974

  31. [39]

    D. B. Rubin. Bayesian inference for causal effects: The role of randomization. The Annals of Statistics, 6 0 (1): 0 34--58, 1978

    1978

  32. [40]

    Samartsidis, S

    P. Samartsidis, S. R. Seaman, Montagna, A. Charlett, M. Hickman, and D. De Angelis. Bayesian multivariate factor analysis model for evaluating an intervention by using observational time series data on multiple outcomes. Journal of the Royal Statistical Society Series A: Stati...

    2020

  33. [41]

    S. UNFCCC. Report of the conference of the parties on its twenty-first session, held in paris from 30 november to 13 december 2015. In Addendum. Part two: Action taken by the Conference of the Parties at its twenty-first session, United Nations Framework Convention on Climate ...

    2015

  34. [42]

    Y. Xu. Generalized synthetic control method: Causal inference with interactive fixed effects models. Political Analysis, 25 0 (1): 0 57--76, 2017

    2017

  35. [43]

    Y. Xu. Causal inference with time-series cross-sectional data: A reflection. Available at SSRN: https://ssrn.com/abstract=3979613, 2023

    2023

  36. [44]

    O. D. Zerbib. The effect of pro-environmental preferences on bond prices: Evidence from green bonds. Journal of Banking & Finance, 98: 0 39--60, 2019

    2019

  37. [45]

    O. D. Zerbib. A sustainable capital asset pricing model (s-capm): Evidence from environmental integration and sin stock exclusion. Review of Finance, 26 0 (6): 0 1345--1388, 2022

    2022

Pith tools

Reviewed August 6, 2026 · model on record in the stance chip above.