Welcome to

Modelling with parameter mixture copulas

Interactive Supplement CD-ROM

hank you for your interest in my thesis. This CD-ROM is a collection of supplementary material relating to my thesis, titled Modelling with parameter mixture copulas. This thesis was written in 2006 at the University of Sydney as a component of the Bachlor of Commerce (Honours) program.

The supplementary material is divided into the following parts:

Click on each link above to see the detailed list, along with descriptions of each item, and instructions on opening and reading each part. I hope you enjoy this material as much as I have enjoyed writing this thesis. If you would like to contact me, please click here.

Technical note: this and other HTML files in this CD-ROM require Micorosft Internet Explorer 5.0 or above under default settings for full functionality. Some features are unavailable if the display of active contents have been disabled. However, the files have been designed for basic functionality under customised security settings under Microsoft Internet Explorer 4.0 or above, and under Mozilla Firefox 1.0.


This thesis was written using Scientific Workplace. The thesis is available as a single PDF file, and also in individual PDFs by chapter for convenient reading and/or transferring. Note that Adobe Acrobat Reader or another compatible program is required to open PDF files. The Adobe Acrobat Reader is available to download for free at: Get Adobe Reader.

Single file:

Modelling with parameter mixture copulas
PDF 2204 KB

By chapter [browse]:



  1. Introduction
    PDF 70 KB

  2. Copulas and mixture copulas [show/hide details]
    PDF 372 KB

    1. Copulas

    2. Sklar's theorem

    3. Families of Copulas

    4. Dependence Coverage

    5. Modelling using copulas

    6. Parameter-mixing

    7. Parameter-mixing applied to copulas

  3. Properties of mixture copulas [show/hide details]
    PDF 237 KB

    1. Overview

    2. Example of mixture copulas

    3. Equivalent functional form

    4. Identification

    5. Dependence measure

    6. Dependence structures covered

    7. Informative mixing

  4. Simulation [show/hide details]
    PDF 120 KB

    1. Overview

    2. Simulation algorithms

    3. Estimation

    4. Comparing results from mixture copulas with parent copulas

    5. Comparing results at different dependence levels

  5. Application [show/hide details]
    PDF 1653 KB

    1. Overview

    2. NBA Data

    3. Estimation

    4. Estimated marginal distributions

    5. Comparing mixture copulas with parent copulas

    6. Effect of partitioning

  6. Conclusion [show/hide details]
    PDF  62 KB

    1. Further research

    2. Conclusion

    3. Note on appendices



Modelling with parameter mixture copulas - Thesis presentation   Microsoft Powerpoint Presentation 2840 KB
Thesis presentation as presented to the Discipline of Econometrics and Business Statistics, 6 Oct 2006

[back to top]


The following data sets are used in the thesis. Note that Microsoft Excel is required to open Excel files (extension .xls). A text editor or spreadsheet program, such as Microsoft Excel, is required to open CSV files (Comma Separated Values; extension .csv).

Raw data in excel format

Data in csv format

[back to top]

Mathematica codes and results

Computation for this thesis was, for the most part, conducted using Mathematica. Note that Mathematica is required to open Mathematica Notebook files (extension .nb). Note also that some of these notebooks require mathStatica, the add-on to Mathematica. For more information on Mathematica and mathStatica see http://www.wolfram.com/products/mathematica/index.html and http://www.mathstatica.com/ respectively.




[back to top]


The following graphics were used in writing this thesis. The file types of each graphic is as noted. Jpeg and Gif files can be viewed in standard web browsers. However, graphics are best viewed at original size. Animated gifs may not display correctly when viewed inside web browser. Choose "Browse" to view these files in a graphics program.

Browse Graphics

[back to top]