Clustering (financial) time series according to their predictability by automatically chosen predictors


Type Research Project

Funding Bodies
  • Siemens Austria

Duration June 1, 2000 - Dec. 31, 2002

  • Mathematical Methods in Statistics AE (Former organization)

Tags

Press 'enter' for creating the tag
  • Hauser, Michael (Details) Project Head
 

Abstract (German)

Clustern von (Finanz-)Zeitreihen bezüglich ihrer Vorhersagbarkeit durch automatisch gewählten Prädiktoren


Abstract (English)

The project tries to find a small number of observed predictors (leading indicators) for a large number of time series. Therefore a measure for the degree of predictability for each series by the potential predictors is constructed. Both a time domain (based on the cross correlation function) and a frequency domain (based on the cross spectrum) version is given. <BR>
The result is a possibly rectangular (depending on the choice of the potential predictors) and essentially asymmetric predictability matrix. A heuristic clustering method is developed to cope with this type of problem. It is a generalization of the PAM algorithm of Kaufman and Rousseeuw (1990), a k-medoids method, for symmetric distance matrices. Alternatively, integer programming solutions could be used to find the clusters.<BR>
The approach is applied to a set of 298 daily financial return series for the period January 1998 to November 2000. It is possible to predict 236 (of 298) series reasonably well by 5 automatically chosen series.

Publications

Unpublished lecture

2000 Hauser, M.. 2000. Datenreduktion mittels Clustermethoden für die Prognose von Finanzreihen. Vortrag bei der SIEMENS Österreich AG, Wien, Österreich, 29.08.2000 (Details)

Classification

  • 5707 Time series analysis (Details)
  • 1104 Applied mathematics (Details)

Expertise

  • clustering