Clustering Financial Time Series: How Long is Enough?

Gautier Marti (), Sébastien Andler, Frank Nielsen () and Philippe Donnat ()
Additional contact information
Gautier Marti: LIX - Laboratoire d'informatique de l'École polytechnique [Palaiseau] - X - École polytechnique - IP Paris - Institut Polytechnique de Paris - CNRS - Centre National de la Recherche Scientifique, Hellebore Capital Limited
Sébastien Andler: ENS de Lyon - École normale supérieure de Lyon - Université de Lyon, Hellebore Capital Limited
Frank Nielsen: LIX - Laboratoire d'informatique de l'École polytechnique [Palaiseau] - X - École polytechnique - IP Paris - Institut Polytechnique de Paris - CNRS - Centre National de la Recherche Scientifique
Philippe Donnat: Hellebore Capital Limited

Post-Print from HAL

Abstract: Researchers have used from 30 days to several years of daily returns as source data for clustering financial time series based on their correlations. This paper sets up a statistical framework to study the validity of such practices. We first show that clustering correlated random variables from their observed values is statistically consistent. Then, we also give a first empirical answer to the much debated question: How long should the time series be? If too short, the clusters found can be spurious; if too long, dynamics can be smoothed out.

Keywords: Financial time series; Clustering; Convergence rates; Correlation (search for similar items in EconPapers)
Date: 2016-07-09
New Economics Papers: this item is included in nep-ets
Note: View the original document on HAL open archive server: https://hal.science/hal-01400395v1
References: View references in EconPapers View complete reference list from CitEc
Citations: View citations in EconPapers (6)

Published in Twenty-Fifth International Joint Conference on Artificial Intelligence, IJCAI 2016, Jul 2016, New York, United States

Downloads: (external link)
https://hal.science/hal-01400395v1/document (application/pdf)

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:hal:journl:hal-01400395

Access Statistics for this paper

More papers in Post-Print from HAL
Bibliographic data for series maintained by CCSD ().