Un approccio metrico per lo studio dei dati finanziari
In this paper I present a time series analysis based on a metrical approach. I use a definition of distance which depends on the sample correlation coefficient among bonds. The dataset consists on daily returns of US treasury bonds. By mean of a Linkage-Algorithm bonds are classified according to the distance which show the cluster structure. It is evident how the cluster structure depends strongly on maturity date, bonds are classified in three different clusters, one of them consists on long term bonds. The analysis is focused on long term bonds, introducing a modified time series, I show how is possible to evidentiate a complex cluster structure even in this class of bonds.
|Date of creation:||Feb 2004|
|Publication status:||Published in Metodi Matematici per l'Economia e la Finanza (a cura di Lucia Maddalena), Collana Interdipartimentale di Studi Economici, Volume 4, 2005 ( ISBN: 88-495-1012-8), Edizioni Scientifiche Italiane.|
|Note:||pdf file is available on request|
|Contact details of provider:|| Postal: Largo Papa Giovanni Paolo II, 1 -71100- Foggia (I)|
Web page: http://www.dsems.unifg.it
More information through EDIRC
When requesting a correction, please mention this item's handle: RePEc:ufg:qdsems:lg_igr_2004. See general information about how to correct material in RePEc.
For technical questions regarding this item, or to correct its authors, title, abstract, bibliographic or download information, contact: (Luca Grilli)
If references are entirely missing, you can add them using this form.