Stability Analysis Using Kalman Filtering, Scoring, EM, and an Adaptive EM method

  • Wolfgang Schneider
Conference paper

Summary

This chapter gives a detailed description of the implementation of ML estimation using scoring and EM for the hyperparameters of a particular econometric state-space. Kaiman filtering enters these methods in an essential way. The EM method can be turned into an on-line (adaptive) estimation method, which can be conveniently used for speeding up the ML estimation procedure. We apply these techniques to a random walk parameter model of a standard (Goldfeld type) West German money-demand function testing its stability via testing the variances of the random walk for zero. We compare these results to a descriptive stability analysis that uses so-called flexible least squares—a nonstochastic variant of Kaiman filtering.

Keywords

Covariance Income Assure Eter Summing 

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

References

  1. Anderson, B.D.O. and Moore, J.B. (1979), Optimal Filtering. Englewood Cliffs: Prentice-Hall.Google Scholar
  2. Ansley, C.F. and Kohn, R. (1982), A geometrical derivation of the fixed interval smoothing algorithm. Biometrika, 69, 486–487.CrossRefGoogle Scholar
  3. Aoki, M. (1987), State Space Modeling of Time Series. Berlin: Springer-Verlag.CrossRefGoogle Scholar
  4. Bierman, G.J. (1974), Sequential square root filtering and smoothing of discrete linear systems. Automatica, 10, 147–158.CrossRefGoogle Scholar
  5. Bierman, G.J. (1977), Factorization Methods for Discrete Sequential Estimation. New York: Academic Press.Google Scholar
  6. Brown, R.L., Durbin, J., and Evans, J.M. (1975), Techniques for testing the constancy of regession relationships over time. Journal of the Royal Statistical Society, B-37, 149–192.Google Scholar
  7. Buscher, H.S. (1984), The stability of the West German demand for money. Weltwirt schaftliches Archiv, 120, 256–277.CrossRefGoogle Scholar
  8. Carraro, C. and Sartore, D. (1986), Square root iterative filter: Theory and applications to econometric models. Working Paper 8611, University of Venice.Google Scholar
  9. Cooley, T.F. and Prescott, E.C. (1973), Varying parameter regression: A theory and some applications. Annals of Economic and Social Measurement, 2, 463–473.Google Scholar
  10. Cooley, T.F., Rosenberg, B., and Wall, K.D. (1977), A note on the optimal smoothing for time varying coefficient problems. Annals of Economic and Social Measurement, 6, 453–462.Google Scholar
  11. Cooper, J.P. (1973), Time-varying regression coefficients: A mixed estimation approach and operational limitations of the general Markov structure. Annals of Economic and Social Measurement, 4, 525–530.Google Scholar
  12. de Jong, P. and MacKinnon, M.J. (1988), Covariances for smoothed estimates in state- space models. Biometrika, 75, 601–602.CrossRefGoogle Scholar
  13. Dempster, A.P., Laird, N.M., and Rubin, D.B. (1977), Maximum-likelihood from incom plete data via the EM-algorithm. Journal of the Royal Statistical Society, B-39, 1–38.Google Scholar
  14. Duncan, D.B. and Horn, S.D. (1972), Linear recursive estimation from the viewpoint of regression analysis. Journal of the American Statistical Association, 67, 815–821.Google Scholar
  15. Eraser, D.C. and Potter, J.E. (1969), The optimum linear smoother as a combination of two optimum linear filters. IEEE Transactions on Automation and Control, AC-14, 387–390.CrossRefGoogle Scholar
  16. Gaab, W. and Seitz, H. (1988), Zur Frage der Stabilität der Geldnachfrage Ml, pp. 93 – 122 in W. Franz, W. Gaab, and J. Wolters (eds.), Theoretische und angewandte Wirtschaftsforschung. Berlin: Springer-Verlag.Google Scholar
  17. Garbade, K. (1977), Two methods for examining the stability of regression coefficients. Journal of the American Statistical Association, 72, 54–63.CrossRefGoogle Scholar
  18. Gersch, W. and Kitigawa, G. (1988), Smoothness priors in time series, pp. 431–476 in C. Spall (ed.), Bayesian Analysis of Time Series and Dynamic Models. New York: Marcel Dekker.Google Scholar
  19. Goldfeld, S.M. (1976), The case of the missing money. Brookings Papers on Economic Activity, 3, 683–730.CrossRefGoogle Scholar
  20. Gupta, N.K. and Mehra, R.K. (1974), Computational aspects of maximum likelihood estimation and reduction of sensitivity function calculations. IEEE Transactions on Automation and Control, AC-19, 774–783.CrossRefGoogle Scholar
  21. Harrison, P.J. and Stevens, C.F. (1976), Bayesian forecasting (with discussion). Journal of the Royal Statistical Society, B-38, 205–247.Google Scholar
  22. Jazwinski, A.H. (1970), Stochastic Processes and Filtering Theory. New York: Academic Press.Google Scholar
  23. Kalaba, R. and Tesfatsion, L. (1986), Time-varying regression via flexible least squares. Modelling Research Group Working Paper 8633, University of Southern California, Los Angeles.Google Scholar
  24. Kalaba, R. and Tesfatsion, L. (1988a), U.S. money demand instability: A flexible least squares approach. Modelling Research Group Working Paper 8809, University of Southern California, Los Angeles.Google Scholar
  25. Kalaba, R. and Tesfatsion, L. (1988b), An organizing principle for dynamic estimation. Modelling Research Group Working Paper 8818, University of Southern California, Los Angeles.Google Scholar
  26. Kaminski, P.G., Bryson, A.E., and Schmidt, S.F. (1971), Discrete square root filtering: A survey of current techniques. IEEE Transactions on Automation and Control, AC-16, 727–736.CrossRefGoogle Scholar
  27. Kohn, R. and Ansley, C.F. (1988), Equivalence between Bayesian smoothness priors and optimal smoothing for function estimation, pp. 393–430 in C. Spall (ed.), Bayesian Analysis of Time Series and Dynamic Models. New York: Marcel Dekker.Google Scholar
  28. Kwakernaak, H. and Sivan, R. (1972), Linear Optimal Control Systems. New York: John Wiley.Google Scholar
  29. Lasdon, L.S. (1970), Optimization Theory for Large Systems. New York: John Wiley.Google Scholar
  30. Louv, W.C. (1984), Adaptive filtering. Technometrics, 26, 399–409.CrossRefGoogle Scholar
  31. Luenberger, D.G. (1969), Optimization by Vector Space Methods. New York: John Wiley.Google Scholar
  32. Magnus, J.R. and Neudecker, H. (1988), Matrix Differential Calculus. New York: John Wiley.Google Scholar
  33. Mayne, D.Q. (1966), A solution for the smoothing problem for linear dynamic systems. Automatica, 4, 73–92.CrossRefGoogle Scholar
  34. Meinhold, R.J. and Singpurwalla, N. (1983), Understanding the Kaiman filter. The American Statistician, 37, 123–127.Google Scholar
  35. Morf, M. and Kailath, T. (1975), Square-root algorithms for least-squares estimation. IEEE Transactions on Automation and Control, AC-20, 487–497.CrossRefGoogle Scholar
  36. Morris, C.R. (1983), Parametric empirical Bayes inference: Theory and applications. Journal of the American Statistical Association, 78, 47–55.CrossRefGoogle Scholar
  37. Nabeya, S. and Tanaka, K. (1988), Asymptotic theory of a test for the constancy of regression coefficients against the random walk alternative. The Annals of Statistics, 16, 218–235.CrossRefGoogle Scholar
  38. Oberhofer, W. and Kmenta, J. (1974), A general procedure for obtaining maximum likelihood estimates via generalized regression models. Econometrica, 42, 579–590.CrossRefGoogle Scholar
  39. Pagan, A. (1980), Some identification and estimation results for regression models with stochastically varying coefficients. Journal of Econometrics, 13, 341–363.CrossRefGoogle Scholar
  40. Rao, C.R. (1971), Estimation of variance and covariance components: MINQUE theory. Journal of Multivariate Analysis, 1, 257–277.CrossRefGoogle Scholar
  41. Rao, C.R. (1973), Linear Statistical Inference and Its Applications. New York: John Wiley.CrossRefGoogle Scholar
  42. Rosenberg, B. (1973), A survey of stochastic parameter regression. Annals of Economic and Social Measurement, 2, 381–398.Google Scholar
  43. Rosenberg, B. (1977), Estimation of error covariance in regession with sequentially varying parameters. Annals of Economic and Social Measurement, 6, 457–462.Google Scholar
  44. Ruskeepää, H. (1985), Conditionally Gaussian distributions and an application to Kaiman filtering with stochastic regressors. Communications in Statistics, Theory and Methods, A-14, 2919–2942.CrossRefGoogle Scholar
  45. Sallas, W.M. and Harville, D.A. (1981), Best linear recursive estimation for mixed linear models. Journal of the American Statistical Association, 76, 860–869.CrossRefGoogle Scholar
  46. Sallas, W.M. and Harville, D.A. (1988), Noninformative priors and restricted maximum likelihood estimation in the Kaiman filter, pp. 477–508 in C. Spall, (ed.), Bayesian Analysis of Time Series and Dynamic Models. New York: Marcel Dekker.Google Scholar
  47. Sarris, A.H. (1973), Kaiman filter models: A Bayesian approach to estimation of time varying regression coefficients. Annals of Economic and Social Measurement, 2, 501–523.Google Scholar
  48. Schneider, W. (1986), Der Kaimanfilter als Instrument zur Diagnose und Schätzung variabler Parameter in ökonometrischen Modellen, volume 27 in K.-A. Schäffer, P. Schönfeld, and W. Wetzel (eds.), Arbeiten zur Angewandten Statistik. Heidelberg: Physica.Google Scholar
  49. Schneider, W. (1988), Analytical uses of Kalman filtering in econometrics: A survey. Statistical Papers, 29, 3–33.CrossRefGoogle Scholar
  50. Schneider, W. (1989), Maximum likelihood estimation of seemingly unrelated regression equations with time varying coefficients by Kalman filtering: A comparison of scoring, EM- and an adaptive EM-method. Working paper, University of Kiel.Google Scholar
  51. Shumway, R.H. and Stoffer, D.S. (1982), An approach to time series smoothing and forecasting using the EM-algorithm. Journal of Time Series Analysis, 3, 253–264.CrossRefGoogle Scholar
  52. Watson, M.W. and Engle, R.F. (1983), Alternative algorithms for the estimation of dynamic factor, mimic and varying coefficient regression models. Journal of Econometrics, 23, 385–400.CrossRefGoogle Scholar
  53. Wu, C.F.J. (1983), On the convergence properties of the EM-algorithm. The Annals of Statistics, 11, 95–103.CrossRefGoogle Scholar
  54. Zellner, A. (1962), An efficient method of estimating seemingly unrelated regressions and tests for aggregation bias. Journal of the American Statistical Association, 57, 348–368.CrossRefGoogle Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 1991

Authors and Affiliations

  • Wolfgang Schneider

There are no affiliations available

Personalised recommendations