Completing Hedge Fund Missing Net Asset Values Using Kohonen Maps and Constrained Randomization

  • Paul Merlin
  • Bertrand Maillet
Conference paper
Part of the Lecture Notes in Computer Science book series (LNCS, volume 3697)


Analysis of financial databases is sensitive to missing values (no reported information, provider errors, outlier filters...). Risk analysis and portfolio asset allocation require cylindrical and complete samples. Moreover, return distributions are characterised by non-normalities due to heteroskedasticity, leverage effects, volatility feedbacks and asymmetric local correlations. This makes completion algorithms very useful for portfolio management applications, specifically if they can deal properly with the empirical stylised facts of asset returns. Kohonen maps constitute powerful nonlinear financial classification tools (see [3], [4] or [6] for instance), following the approach of Cottrell et al. (2003), we use a Kohonen algorithm (see [2]), altogether with the Constrained Randomization Method (see [8]) to deal with mutual fund missing Net Asset Values. The accuracy of rebuilt NAV estimated series is then evaluated according to a comparison between the first moments of the series.


Hedge Fund Asset Return Asset Allocation Code Vector Performance Persistence 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. 1.
    Agarwal, V., Naïk, N.: Multi-period Performance Persistence Analysis of Hedge Funds. Journal of Financial and Quantitative Analysis 35, 327–342 (2000)CrossRefGoogle Scholar
  2. 2.
    Cottrell, M., Ibbou, S., Letrémy, P.: Traitement des données manquantes au moyen de l’algorithme de Kohonen. In: Proceedings of the tenth ACSEG Conference, in french, 12 pages (2003)Google Scholar
  3. 3.
    Cottrell, M., de Bodt, E., Grégoire, P.: Financial Application of the Self-Organizing Map. In: Proceedings of EUFIT 1998 (1), Verlag Mainz, pp. 205–209 (1998)Google Scholar
  4. 4.
    De Bodt, E., Cottrell, M.: Bootstrapping Self-Organizing Maps to Assess the Statistical Significance of Local Proximity. In: European Symposium on Artificial Neural Networks, pp. 245–254 (2000)Google Scholar
  5. 5.
    Kohonen, T.: Self-Organizing Maps, 362 pages. Springer, Berlin (1995)Google Scholar
  6. 6.
    Maillet, B., Rousset, P.: Classifying Hedge Funds using Kohonen Map. In: Cottrell-Lesage (ed.) Connectionist Approaches in Economics and Management Sciences. Series in Advances in Computational Management Science, vol. 6, pp. 233–259. Kluwer Academic Publisher, Dordrecht (2003)Google Scholar
  7. 7.
    Samad, T., Harp, S.: Self Organization with Partial Data. Network 3, 205–212 (1992)CrossRefGoogle Scholar
  8. 8.
    Schreiber, T.: Constrained Randomization of Times Series Data. Physical Review Letter 80(10), 2105–2108 (1998)CrossRefGoogle Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2005

Authors and Affiliations

  • Paul Merlin
    • 1
  • Bertrand Maillet
    • 2
  1. 1.A.A.Advisors-QCG (ABN Amro Group), Variances and Paris-1, (TEAM/CNRS and SAMOS/MATISSE)Paris
  2. 2.A.A.Advisors-QCG (ABN Amro Group), Variances and Paris-1, (TEAM/CNRS)Paris cedex 13

Personalised recommendations