Computer Science in Economics and Management

, Volume 5, Issue 3, pp 183–220 | Cite as

SISAM and MIXIN: Two algorithms for the computation of posterior moments and densities using Monte Carlo integration

  • J. Peter Hop
  • Herman K. Van Dijk
Article

Abstract

Two algorithms, and corresponding Fortran computer programs, for the computation of posterior moments and densities using the principle of importance sampling are described in detail. The first algorithm makes use of a multivariate Student t importance function as approximation of the posterior. It can be applied when the integrand is moderately skew. The second algorithm makes use of a decomposition: a multivariate normal importance function is used to generate directions (lines) and one-dimensional classical quadrature is used to evaluate the integrals defined on the generated lines. The second algorithm can be used in cases where the integrand is possibly very skew in any direction.

Key words

Bayesian analysis Monte Carlo integration importance sampling 

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

References

  1. Abramowitz, M. and Stegun, I.A. (1964) Handbook of Mathematical Functions, National Bureau of Standards, Washington, DC.Google Scholar
  2. Bauwens, L. (1984) Bayesian Full Information Analysis of Simultaneous Equation Models Using Integration by Monte Carlo, Springer-Verlag, New York.Google Scholar
  3. Bauwens, L. and Richard, J.F. (1985) A 1-1 poly-t random variable generator with applications to Monte Carlo integration, J. Econometrics 29, 19–46.Google Scholar
  4. Billingsley, P. (1979) Probability and Measure, Wiley, New York.Google Scholar
  5. Boender, C.G.E. and Van Dijk, H.K. (1991) Bayes estimates of multi-criteria decision alternatives using Monte Carlo integration, Report: 91441A, Econometric Institute, Erasmus University, Rotterdam (forthcoming in Statistica Neerlandica).Google Scholar
  6. Cramér, H. (1946) Mathematical Methods of Statistics, Princeton University Press, Princeton, NJ.Google Scholar
  7. DeJong, D.N. (1991) Bayesian inference in limited dependent variable models: An application to measuring strike duration, Manuscript, Department of Economics, University of Pittsburgh.Google Scholar
  8. DeJong, D.N. and Whiteman, C.H. (1989) Trends and cycles as unobserved components in U.S. real GNP: A Bayesian perspective, Proc. Amer. Statist. Assoc., Business and Economics Section, 63–70.Google Scholar
  9. DeJong, D.N. and Whiteman, C.H. (1990) Are output fluctuations transitory?: A Bayesian perspective, Working paper No. 90-26, Department of Economics, University of Iowa.Google Scholar
  10. DeJong, D.N. and Whiteman, C.H. (1991a) Reconsidering trends and random walks in macroeconomic time series, J. Monetary Econom. (forthcoming).Google Scholar
  11. DeJong, D.N. and Whiteman, C.H. (1991b) The temporal stability of dividends and stock prices: Evidence from the likelihood function, Amer. Econom. Rev. 81, 600–617.Google Scholar
  12. Dhrymes, P.J. (1971) Econometrics, Statistical Foundations and Applications, Harper and Row, New York.Google Scholar
  13. Geweke, J. (1986) Exact inferences in the inequality constrained normal linear regression model, J. Appl. Econometrics 1, 127–141.Google Scholar
  14. Geweke, J. (1988a) Exact inference in models with autoregressive conditional heteroscedasticity, in E. Berndt, H. White, and W. Barnett (eds), Dynamic Econometric Modeling, University Press, Cambridge.Google Scholar
  15. Geweke, J. (1988b) The secular and cyclical behavior of real GDP in nineteen OECD countries, J. Business Econom. Stat. 6, 479–486.Google Scholar
  16. Geweke, J. (1988c) Antithetic acceleration of Monte Carlo integration in Bayesian inference, J. Econometrics 38, 73–89.Google Scholar
  17. Geweke, J. (1989a) Exact predictive densities in linear models with ARCH disturbances, J. Econometrics 40, 63–86.Google Scholar
  18. Geweke, J. (1989b) Bayesian inference in econometric models using Monte Carlo integration, Econometrica 57, 1317–1339.Google Scholar
  19. Geweke, J., Marchal, R.C., and Zarkin, G.A. (1986) Exact inference for continuous time Markov chain models, Rev. Econom. Studies LIII, 653–669.Google Scholar
  20. Hammersley, J.M., and Handscomb, D.C. (1964) Monte Carlo Methods, 1st edn, Methuen, London.Google Scholar
  21. Hammersley, J.M., and Handscomb, D.C. (1979) Monte Carlo Methods, 2nd edn, Chapman and Hall, London.Google Scholar
  22. Hop, J.P. and Van Dijk, H.K. (1990) SISAM and MIXIN: Two algorithms for the computation of posterior moments and densities using Monte Carlo integration, Report 9031/A, Econometric Institute, Erasmus University, Rotterdam.Google Scholar
  23. Judge, G.C., Griffiths, W.E., Lutkepohl, H., and Lee, T.C. (1988) Introduction to the Theory and Practice of Econometrics, 2nd edn. Wiley, New York.Google Scholar
  24. Kim, In-Moo and Maddala, G.S. (1991) Structural change and unit roots, Manuscript, Department of Economics, The University of Florida, Gainesville.Google Scholar
  25. Kinderman, A.J. and Monahan, J.F. (1980) New methods for generating students t and gamma variables, Computing 25, 369–377.Google Scholar
  26. Kinderman, A.J. and Ramage, J.G. (1976) Computer generation of normal random variables, J. Amer. Statis. Assoc. 71, 893–896.Google Scholar
  27. Marsaglia, G. (1984) The exact-approximation method for generating random variables in a computer, J. Amer. Statist. Assoc. 79, 218–221.Google Scholar
  28. Mood, A.M., Graybill, F.A., and Boes, D.C. (1974) Introduction to the Theory of Statistics, 3rd edn. McGraw-Hill, New York.Google Scholar
  29. Oh, Man-Suk and Berger, J.O. (1989) Adaptive importance sampling in Monte Carlo integration, Manuscript, Purdue University and Duke University.Google Scholar
  30. Richard, J.F. and Steel, M.F.J. (1988) Bayesian analysis of systems of seemingly unrelated regression equations under a recursive extended natural conjugate prior density, J. Econometrics 38, 7–37.Google Scholar
  31. Rubinstein, R.Y. (1981) Simulation and the Monte Carlo Method, Wiley, New York.Google Scholar
  32. Steel, M.F.J. (1987) A Bayesian analysis of multivariate exogeneity: A Monte Carlo approach, Doctoral dissertation, University Catholique de Louvain.Google Scholar
  33. Van Dijk, H.K. (1984) Posterior analysis of econometric models using Monte Carlo integration, Doctoral dissertation, Erasmus University, Rotterdam.Google Scholar
  34. Van Dijk, H.K. (1987) Some advances in Bayesian estimation methods using Monte Carlo integration in T.B. Fomby and G.F. Rhodes (eds.), Advances in Econometrics 6, JAI Press, Greenwich Connecticut, pp. 215–261.Google Scholar
  35. Van Dijk, H.K., Hop, J.P., and Louter, A.S. (1987) An algorithm for the computation of posterior moments and densities using simple importance sampling, The Statistician 36, 83–90.Google Scholar
  36. Van Dijk, H.K. and Kloek, T. (1980) Further experience in Bayesian analysis using Monte Carlo integration, J. Econometrics 14, 307–328.Google Scholar
  37. Van Dijk, H.K. and Kloek T. (1983) Experiments with some alternatives for simple importance sampling in Monte Carlo integration, in J.M. Bernardo, M.H. De Groot, D.V. Lindley, and A.F.M. Smith (eds), Bayesian Statistics 2, North-Holland, Amsterdam, pp. 511–530.Google Scholar
  38. Van Dijk, H.K., Kloek, T., and Boender, C.G.E. (1985) Posterior moments computed by mixed integration, J. Econometrics 29, 3–18.Google Scholar
  39. West, M. (1990) Bayesian kernel density estimation, ISDS Discussion paper #90-A12, Duke University.Google Scholar
  40. Zellner, A., Bauwens, L., and Van Dijk, H.K. (1988) Bayesian specification analysis and estimation of simultaneous equation models using Monte Carlo methods, J. Econometrics 38, 39–72.Google Scholar
  41. Zellner, A. and Rossi, P. (1984) Bayesian analysis of dichotomous quantal response models, J. Econometrics 25, 365–394.Google Scholar

Copyright information

© Kluwer Academic Publishers 1992

Authors and Affiliations

  • J. Peter Hop
    • 1
  • Herman K. Van Dijk
    • 1
  1. 1.Econometric InstituteErasmus UniversityDR RotterdamThe Netherlands

Personalised recommendations