IDEAS home Printed from https://ideas.repec.org/p/ris/smuesw/2019_016.html
   My bibliography  Save this paper

Improved Marginal Likelihood Estimation via Power Posteriors and Importance Sampling

Author

Listed:
  • Li, Yong

    (Renmin University of China)

  • Wang, Nianling

    (Renmin University of China)

  • Yu, Jun

    (School of Economics, Singapore Management University)

Abstract

The power-posterior method of Friel and Pettitt (2008) has been used to estimate the marginal likelihoods of competing Bayesian models. In this paper it is shown that the Bernstein-von Mises (BvM) theorem holds for the power posteriors under regularity conditions. Due to the BvM theorem, the power posteriors, when adjusted by the square root of the corresponding grid points, converge to the same normal distribution as the original posterior distribution, facilitating the implementation of importance sampling for the purpose of estimating the marginal likelihood. Unlike the power-posterior method that requires repeated posterior sampling from the power posteriors, the new method only requires the posterior output from the original posterior. Hence, it is computationally more efficient to implement. Moreover, it completely avoids the coding efforts associated with drawing samples from the power posteriors. Numerical efficiency of the proposed method is illustrated using two models in economics and finance.

Suggested Citation

  • Li, Yong & Wang, Nianling & Yu, Jun, 2019. "Improved Marginal Likelihood Estimation via Power Posteriors and Importance Sampling," Economics and Statistics Working Papers 16-2019, Singapore Management University, School of Economics.
    • Handle: RePEc:ris:smuesw:2019_016
    as

    Download full text from publisher

    File URL: https://ink.library.smu.edu.sg/soe_research/2287/
    File Function: Full text
    Download Restriction: no
    ---><---

    Other versions of this item:

    References listed on IDEAS

    as
    1. Abel Brodeur & Nikolai Cook & Anthony Heyes, 2020. "Methods Matter: p-Hacking and Publication Bias in Causal Analysis in Economics," American Economic Review, American Economic Association, vol. 110(11), pages 3634-3660, November.
    2. Campbell R. Harvey, 2017. "Presidential Address: The Scientific Outlook in Financial Economics," Journal of Finance, American Finance Association, vol. 72(4), pages 1399-1440, August.
    3. Vaart,A. W. van der, 2000. "Asymptotic Statistics," Cambridge Books, Cambridge University Press, number 9780521784504, October.
    4. N. Friel & A. N. Pettitt, 2008. "Marginal likelihood estimation via power posteriors," Journal of the Royal Statistical Society Series B, Royal Statistical Society, vol. 70(3), pages 589-607, July.
    5. Han C. & Carlin B. P., 2001. "Markov Chain Monte Carlo Methods for Computing Bayes Factors: A Comparative Review," Journal of the American Statistical Association, American Statistical Association, vol. 96, pages 1122-1132, September.
    6. Waggoner, Daniel F. & Wu, Hongwei & Zha, Tao, 2016. "Striated Metropolis–Hastings sampler for high-dimensional models," Journal of Econometrics, Elsevier, vol. 192(2), pages 406-420.
    7. Li, Yong & Yu, Jun & Zeng, Tao, 2020. "Deviance information criterion for latent variable models and misspecified models," Journal of Econometrics, Elsevier, vol. 216(2), pages 450-493.
    8. Young, Karen D. S. & Pettit, Lawrence I., 1996. "On priors and Bayes factors," Journal of Econometrics, Elsevier, vol. 75(1), pages 113-119, November.
    Full references (including those not matched with items on IDEAS)

    Citations

    Citations are extracted by the CitEc Project, subscribe to its RSS feed for this item.
    as


    Cited by:

    1. Liu, Xiaobin & Li, Yong & Yu, Jun & Zeng, Tao, 2022. "Posterior-based Wald-type statistics for hypothesis testing," Journal of Econometrics, Elsevier, vol. 230(1), pages 83-113.

    Most related items

    These are the items that most often cite the same works as this one and are cited by the same works as this one.
    1. Liu, Xiaobin & Li, Yong & Yu, Jun & Zeng, Tao, 2022. "Posterior-based Wald-type statistics for hypothesis testing," Journal of Econometrics, Elsevier, vol. 230(1), pages 83-113.
    2. Guillaume Coqueret, 2023. "Forking paths in financial economics," Papers 2401.08606, arXiv.org.
    3. Joshua C. C. Chan & Eric Eisenstat, 2015. "Marginal Likelihood Estimation with the Cross-Entropy Method," Econometric Reviews, Taylor & Francis Journals, vol. 34(3), pages 256-285, March.
    4. Engsted, Tom & Schneider, Jesper W., 2023. "Non-Experimental Data, Hypothesis Testing, and the Likelihood Principle: A Social Science Perspective," SocArXiv nztk8, Center for Open Science.
    5. Lefebvre, Geneviève & Steele, Russell & Vandal, Alain C., 2010. "A path sampling identity for computing the Kullback-Leibler and J divergences," Computational Statistics & Data Analysis, Elsevier, vol. 54(7), pages 1719-1731, July.
    6. Aubry, Amandine & Héricourt, Jérôme & Marchal, Léa & Nedoncelle, Clément, 2022. "Does Immigration AffectWages? A Meta-Analysis," CEPREMAP Working Papers (Docweb) 2202, CEPREMAP.
    7. Shi, Huai-Long & Zhou, Wei-Xing, 2022. "Factor volatility spillover and its implications on factor premia," Journal of International Financial Markets, Institutions and Money, Elsevier, vol. 80(C).
    8. Holla,Alaka & Bendini,Maria Magdalena & Dinarte Diaz,Lelys Ileana & Trako,Iva, 2021. "Is Investment in Preprimary Education Too Low ? Lessons from (Quasi) ExperimentalEvidence across Countries," Policy Research Working Paper Series 9723, The World Bank.
    9. Kasy, Maximilian, 2011. "A nonparametric test for path dependence in discrete panel data," Economics Letters, Elsevier, vol. 113(2), pages 172-175.
    10. Fuyu Yang, 2007. "Bayesian Analysis of Deterministic Time Trend and Changes in Persistence Using a Generalised Stochastic Unit Root Model," Discussion Papers in Economics 07/11, Division of Economics, School of Business, University of Leicester.
    11. Atı̇la Abdulkadı̇roğlu & Joshua D. Angrist & Yusuke Narita & Parag Pathak, 2022. "Breaking Ties: Regression Discontinuity Design Meets Market Design," Econometrica, Econometric Society, vol. 90(1), pages 117-151, January.
    12. Stefano DellaVigna & Elizabeth Linos, 2022. "RCTs to Scale: Comprehensive Evidence From Two Nudge Units," Econometrica, Econometric Society, vol. 90(1), pages 81-116, January.
    13. Xing Ju Lee & Christopher C. Drovandi & Anthony N. Pettitt, 2015. "Model choice problems using approximate Bayesian computation with applications to pathogen transmission data sets," Biometrics, The International Biometric Society, vol. 71(1), pages 198-207, March.
    14. Anna Sokolova, 2023. "Marginal Propensity to Consume and Unemployment: a Meta-analysis," Review of Economic Dynamics, Elsevier for the Society for Economic Dynamics, vol. 51, pages 813-846, December.
    15. Filatotchev, Igor & Poulsen, Annette & Bell, R. Greg, 2019. "Corporate governance of a multinational enterprise: Firm, industry and institutional perspectives," Journal of Corporate Finance, Elsevier, vol. 57(C), pages 1-8.
    16. Jeong Eun Lee & Christian Robert, 2013. "Imortance Sampling Schemes for Evidence Approximation in Mixture Models," Working Papers 2013-42, Center for Research in Economics and Statistics.
    17. H. Latan & C.J. Chiappetta Jabbour & Ana Beatriz Lopes de Sousa Jabbour & M. Ali, 2023. "Crossing the Red Line? Empirical Evidence and Useful Recommendations on Questionable Research Practices among Business Scholars," Post-Print hal-04276024, HAL.
    18. Guo, Li & Sang, Bo & Tu, Jun & Wang, Yu, 2024. "Cross-cryptocurrency return predictability," Journal of Economic Dynamics and Control, Elsevier, vol. 163(C).
    19. William Forbes & Egor Kiselev & Len Skerratt, 2023. "The stability and downside risk to contrarian profits: Evidence from the S&P 500," International Journal of Finance & Economics, John Wiley & Sons, Ltd., vol. 28(1), pages 733-750, January.
    20. Will Penny & Biswa Sengupta, 2016. "Annealed Importance Sampling for Neural Mass Models," PLOS Computational Biology, Public Library of Science, vol. 12(3), pages 1-25, March.

    More about this item

    Keywords

    Bayes factor; Marginal likelihood; Markov Chain Monte Carlo; Model choice; Power posteriors; Importance sampling;
    All these keywords.

    JEL classification:

    • C11 - Mathematical and Quantitative Methods - - Econometric and Statistical Methods and Methodology: General - - - Bayesian Analysis: General
    • C12 - Mathematical and Quantitative Methods - - Econometric and Statistical Methods and Methodology: General - - - Hypothesis Testing: General

    NEP fields

    This paper has been announced in the following NEP Reports:

    Statistics

    Access and download statistics

    Corrections

    All material on this site has been provided by the respective publishers and authors. You can help correct errors and omissions. When requesting a correction, please mention this item's handle: RePEc:ris:smuesw:2019_016. See general information about how to correct material in RePEc.

    If you have authored this item and are not yet registered with RePEc, we encourage you to do it here. This allows to link your profile to this item. It also allows you to accept potential citations to this item that we are uncertain about.

    If CitEc recognized a bibliographic reference but did not link an item in RePEc to it, you can help with this form .

    If you know of missing items citing this one, you can help us creating those links by adding the relevant references in the same way as above, for each refering item. If you are a registered author of this item, you may also want to check the "citations" tab in your RePEc Author Service profile, as there may be some citations waiting for confirmation.

    For technical questions regarding this item, or to correct its authors, title, abstract, bibliographic or download information, contact: Cheong Pei Qi (email available below). General contact details of provider: https://edirc.repec.org/data/sesmusg.html .

    Please note that corrections may take a couple of weeks to filter through the various RePEc services.

    IDEAS is a RePEc service. RePEc uses bibliographic data supplied by the respective publishers.