Local asymptotic powers of nonparametric and semiparametric tests for fractional integration

Xiaofeng Shao, Wei Biao Wu

Research output: Contribution to journalArticlepeer-review

Abstract

The paper concerns testing long memory for fractionally integrated nonlinear processes. We show that the exact local asymptotic power is of order O [(log n)- 1] for four popular nonparametric tests and is O (m- 1 / 2), where m is the bandwidth which is allowed to grow as fast as nκ, κ ∈ (0, 2 / 3), for the semiparametric Lagrange multiplier (LM) test proposed by Lobato and Robinson [I. Lobato, P.M. Robinson, A nonparametric test for I (0), Rev. Econom. Stud. 68 (1998) 475-495]. Our theory provides a theoretical justification for the empirical findings in finite sample simulations by Lobato and Robinson [I. Lobato, P.M. Robinson, A nonparametric test for I (0), Rev. Econom. Stud. 68 (1998) 475-495] and Giraitis et al. [L. Giraitis, P. Kokoszka, R. Leipus, G. Teyssiére, Rescaled variance and related tests for long memory in volatility and levels, J. Econometrics 112 (2003) 265-294] that nonparametric tests have lower power than LM tests in detecting long memory.

Original languageEnglish (US)
Pages (from-to)251-261
Number of pages11
JournalStochastic Processes and their Applications
Volume117
Issue number2
DOIs
StatePublished - Feb 2007

Keywords

  • Fractional integration
  • KPSS test
  • Lagrange multiplier test
  • Local Whittle estimation
  • Long memory
  • R/S test

ASJC Scopus subject areas

  • Statistics and Probability
  • Modeling and Simulation
  • Applied Mathematics

Fingerprint

Dive into the research topics of 'Local asymptotic powers of nonparametric and semiparametric tests for fractional integration'. Together they form a unique fingerprint.

Cite this