Hostname: page-component-cd9895bd7-mkpzs Total loading time: 0 Render date: 2024-12-26T20:15:01.423Z Has data issue: false hasContentIssue false
  • English
  • Français

A statistical application of the quantile mechanics approach: MTM estimators for the parameters of t and gamma distributions

Published online by Cambridge University Press:  18 May 2012

A. KLEEFELD  and
V. BRAZAUSKAS
A. KLEEFELD
Affiliation:
Department I (Mathematics, Natural Sciences, Computer Science), Brandenburgische Technische Universität, P.O. Box 101344, 03013 Cottbus, Germany email: [email protected]
V. BRAZAUSKAS
Affiliation:
Department of Mathematical Sciences, University of Wisconsin – Milwaukee, P.O. Box 413, Milwaukee, WI 53201, USA email: [email protected]
Get access Rights & Permissions [Opens in a new window]

Abstract

In this paper, we revisit the quantile mechanics approach, which was introduced by Steinbrecher and Shaw (Steinbrecher, G. & Shaw, W. T. (2008) Quantile mechanics. European. J. Appl. Math.19, 87–112). Our objectives are (i) to derive the method of trimmed moments (MTM) estimators for the parameters of gamma and Student's t distributions, and (ii) to examine their large- and small-sample statistical properties. Since trimmed moments are defined through the quantile function of the distribution, quantile mechanics seems like a natural approach for achieving objective (i). To accomplish the second goal, we rely on the general large sample results for MTMs, which were established by Brazauskas et al. (Brazauskas, V., Jones, B. L. & Zitikis, R. (2009) Robust fitting of claim severity distributions and the method of trimmed moments. J. Stat. Plan. Inference139, 2028–2043), and then use Monte Carlo simulations to investigate small-sample behaviour of the newly derived estimators. We find that, unlike the maximum likelihood method, which usually yields fully efficient but non-robust estimators, the MTM estimators are robust and offer competitive trade-offs between robustness and efficiency. These properties are essential when one employs gamma or Student's t distributions in such outlier-prone areas as insurance and finance.

Keywords

Point estimation (62F10)Asymptotic properties of estimators (62F12)Robustness and adaptive procedures (62F35)Method of trimmed moments (62F99)Computational problems in statistics (65C60)
Type
Papers
Information
European Journal of Applied Mathematics , Volume 23 , Issue 5 , October 2012 , pp. 593 - 610
Copyright
Copyright © Cambridge University Press 2012

Access options

Get access to the full version of this content by using one of the access options below. (Log in options will check for institutional or personal access. Content may require purchase if you do not have access.)

References

[1]Abramowitz, M. & Stegun, I. A. (1972) Handbook of Mathematical Functions, Applied Mathematics Series, No. 55, National Bureau of Standards, Gaithersburg, MD.Google Scholar
[2]Brazauskas, V. & Kleefeld, A. (2011) Folded and log-folded-t distributions as models for insurance loss data. Scand. Actuar. J. 1, 59–74.CrossRefGoogle Scholar
[3]Brazauskas, V., Jones, B. L. & Zitikis, R. (2009) Robust fitting of claim severity distributions and the method of trimmed moments. J. Stat. Plan. Inference 139, 2028–2043.CrossRefGoogle Scholar
[4]Galton, F. (1875) Statistics by intercomparison with remarks on the law of frequency of error. Phil. Mag. 49, 33–46.CrossRefGoogle Scholar
[5]Gilchrist, W. (2000) Statistical Modelling with Quantile Functions, Chapman and Hall, Boca Raton, FL.CrossRefGoogle Scholar
[6]Hald, A. (1998) A History of Mathematical Statistics from 1750 to 1930, Wiley, New York.Google Scholar
[7]Hampel, F. R. (1968) Contributions to the Theory of Robust Estimation. PhD. Dissertation, University of California, Berkeley, CA.Google Scholar
[8]Huber, P. J. (1964) Robust estimation of a location parameter. Ann. Math. Stat. 35, 73–101.Google Scholar
[9]Johnson, N. L., Kotz, S. & Balakrishnan, N. (1994) Continuous Univariate Distributions, 2nd ed., Wiley, New York.Google Scholar
[10]Jones, B. L. & Zitikis, R. (2003), (2004) Empirical estimation of risk measures and related quantities (with discussion). N. Am. Actuar. J. 7, 8, 44–54, 114–117, 117–118.Google Scholar
[11]Klugman, S. A., Panjer, H. H. & Willmot, G. E. (2008) Loss Models: From Data to Decisions, 3rd ed., Wiley, New York.Google Scholar
[12]Maronna, R. A., Martin, D. R. & Yohai, V. J. (2006) Robust Statistics: Theory and Methods, Wiley, New York.CrossRefGoogle Scholar
[13]McNeil, A. J. & Embrechts, P. (2005) Quantitative Risk Management: Concepts, Techniques, and Tools, Princeton University Press, Princeton, NJ.Google Scholar
[14]Necir, A. & Meraghni, D. (2010) Estimating L-functionals for heavy-tailed distributions and application. J. Probab. Stat. Article ID 707146, 34 pp.Google Scholar
[15]Parzen, E. (2004) Quantile probability and statistical data modeling. Stat. Sci. 19, 652–662.Google Scholar
[16]Serfling, R. J. (1980) Approximation Theorems of Mathematical Statistics, Wiley, New York.CrossRefGoogle Scholar
[17]Shaw, W. T. (2006) Sampling Student's T distribution – use of the inverse cumulative distribution function. J. Comput. Finance 9, 37–73.CrossRefGoogle Scholar
[18]Steinbrecher, G. & Shaw, W. T. (2008) Quantile mechanics. European J. Appl. Math. 19, 87–112.CrossRefGoogle Scholar
[19]Wolfram, S. (2004) The Mathematica Book, 5th ed.Wolfram Media, Champaign, IL.Google Scholar