Activity Signature Functions for High-Frequency Data外文翻译.pdfVIP

Activity Signature Functions for High-Frequency Data Analysis∗ Viktor Todorov † and George Tauchen ‡ This Draft: July 22, 2008 First Draft: November 2007 Abstract We define a new concept termed the activity signature function, which is constructed from discrete observations of a process evolving continuously in time. Under quite general regularity conditions, we derive the asymptotic properties of the function as the sampling frequency in- creases and show that it is a useful device for making inferences about the activity level of an Itˆo semimartingale. Monte Carlo work confirms the theoretical results. One empirical applica- tion is from finance. It indicates that the classical model comprised of a continuous component plus jumps is more plausible than a pure-jump model for the spot $/DM exchange rate over 1986-1999. A second application pertains to internet traffic data at NASA servers. We find that a pure-jump model with no continuous component and paths of infinite variation is appropriate for modeling this data set. In both cases the evidence obtained from the signature functions is quite convincing, and these two very disparate empirical outcomes illustrate the discriminatory power of the methodology. Keywords: activity index, Blumenthal-Getoor index, jumps, L´evy process, realized power vari- ation. ∗We thank Cecilia Mancini, Per Mykland, Neil Shephard, Jeannette Woerner as well as seminar participants at Duke University, University of Chicago workshop on finance and statistics, Imperial College 2008 Financial Econo- metrics Conference and Pur