Seminar in applied mathematics and statistics
SPEAKER: Jean Jacod (UPMC-Paris 6)
TITLE: Jump activity estimation, a new method based on the empirical characteristic function approach.
Abstract: By now there are several diff erent methods to estimate the degree of activity of the jumps of an Itô semimartingale. Here we propose a new one, based on the empirical "global" characteristic function of the returns, which allows one to use a single tuning parameter. For simplicity, we restrict our attention to the case where the process is observed without noise and at regularly spaced sampling times, but the method works as well when sampling is irregular (possibly random) and noise is present, although the implementation is more complicated in that case. We will see that we can "almost" achieve the efficient rate when the degree of activity is bigger than 1, and we also can estimate the possible asymmetry between positive and negative jumps.
Based on joint with Viktor Todorov (Northwestern University).
Tea and chocolate will be served in room 04.4.19 after the seminar.
----
Upcoming seminars:
October 27, 13:15: Martin Jönsson (Oxford)
November 3, 11:15: Steven Gabriel (Maryland)
November 3, 13:15: Lars Nørvang Andersen (Aarhus)
November 10, 11:15: Konstantin Pavlikov (Odense)
January 11, 15:15: Ryan Tibshirani (Carnegie Mellon)
February 2, 13:15: Axel Munk (Göttingen)