We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:

Download:

Current browse context:

cond-mat

References & Citations

Bookmark

(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Condensed Matter > Statistical Mechanics

Title: Detecting long and short memory via spectral methods

Authors: Simone Bianco
Abstract: We study the properties of memory of a financial time series adopting two different methods of analysis, the detrended fluctuation analysis (DFA) and the analysis of the power spectrum (PSA). The methods are applied on three time series: one of high-frequency returns, one of shuffled returns and one of absolute values of returns. We prove that both DFA and PSA give results in line with those obtained with standard econometrics measures of correlation.
Comments: 8 pages, 2 fiures. Submitted to Physica A
Subjects: Statistical Mechanics (cond-mat.stat-mech); Statistical Finance (q-fin.ST)
Cite as: arXiv:cond-mat/0610022 [cond-mat.stat-mech]
  (or arXiv:cond-mat/0610022v1 [cond-mat.stat-mech] for this version)

Submission history

From: Simone Bianco [view email]
[v1] Sun, 1 Oct 2006 15:42:36 GMT (15kb)

Link back to: arXiv, form interface, contact.