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

Download:

Current browse context:

math.ST

Change to browse by:

References & Citations

Bookmark

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

Mathematics > Statistics Theory

Title: How the instability of ranks under long memory affects large-sample inference

Abstract: Under long memory, the limit theorems for normalized sums of random variables typically involve a positive integer called "Hermite rank". There is a different limit for each Hermite rank. From a statistical point of view, however, we argue that a rank other than one is unstable, whereas, a rank equal to one is stable. We provide empirical evidence supporting this argument. This has important consequences. Assuming a higher-order rank when it is not really there usually results in underestimating the order of the fluctuations of the statistic of interest. We illustrate this through various examples involving the sample variance, the empirical processes and the Whittle estimator.
Comments: 31 pages
Subjects: Statistics Theory (math.ST)
MSC classes: 62M10, 60F05
Cite as: arXiv:1610.00690 [math.ST]
  (or arXiv:1610.00690v2 [math.ST] for this version)

Submission history

From: Shuyang (Ray) Bai [view email]
[v1] Mon, 3 Oct 2016 19:35:54 GMT (62kb)
[v2] Thu, 2 Nov 2017 03:01:08 GMT (95kb)

Link back to: arXiv, form interface, contact.