Common intraday periodicity

A.W. Hecq, S.F.J.A. Laurent*, F.C. Palm

*Corresponding author for this work

Research output: Contribution to journalArticleAcademicpeer-review

Abstract

Using a reduced rank regression framework as well as information criteria, we investigate the presence of commonalities in the intraday periodicity, a dominant feature in the return volatility of most intraday financial time series. We find that the test has little size distortion and reasonable power even in the presence of jumps. We also find that only three factors are needed to describe the intraday periodicity of 30 U.S. asset returns sampled at the 5-minute frequency. Interestingly, we find that for most series, the models imposing these commonalities deliver better forecasts of the conditional intraday variance than those where the intraday periodicity is estimated for each asset separately.

Original languageEnglish
Pages (from-to)325-353
Number of pages29
JournalJournal of Financial Econometrics
Volume10
Issue number2
DOIs
Publication statusPublished - 1 Jan 2012

Keywords

  • Common features
  • Intraday periodicity
  • Realized volatility
  • C10
  • C32
  • PREDICTIVE ABILITY
  • VOLATILITY MODELS
  • FEATURES
  • MARKETS

Cite this