[Statlist] Research seminar in statistics October 16th 2015, GSEM University of Geneva

[Eva Cantoni]

Organisers :
E. Cantoni - D. La Vecchia - E. Ronchetti -
S. Sperlich - F. Trojani - M-P. Victoria-Feser

Friday October 16th, 2015
at 11h15 - Room M 5220, Uni Mail (40, bd du Pont-d'Arve)

Higher-order asymptotics for frequency domain M-estimators
Davide La Vecchia - GSEM, Université de Genève


ABSTRACT:
Saddlepoint methods provide very accurate approximations to the 
distribution of estimators and test statistics, but except for a few 
simple special cases, these techniques are not available in the time 
series framework. We present two alternative approaches to develop 
saddlepoint techniques for frequency domain M-estimators. In the first 
one, we obtain saddlepoint density and tail probability approximations 
by tilting  the Edgeworth expansion and we illustrate the methodology in 
the case of the widely-applied Whittle estimator. Although the resulting 
saddlepoint approximations are attractive from a theoretical point of 
view, their practical implementation seems to be difficult. In the 
second one, we develop saddlepoint techniques (e.g. for tests in the 
presence of nuisance parameters) using a first-order Gram-Charlier 
differential series, where the developing function is the cumulant 
generating function of an exponential random variable. The methodology 
opens up a wide field of applications, since these methods are linked to 
some existing inferential procedures, they are easy-to-implement and, 
typically, they improve on the routinely applied first-order 
asymptotics. A real data analysis of the Pacific Decadal Oscillation 
index and Monte Carlo exercises illustrate the good performance of the 
proposed methods.


Visit the website: http://www.stat-center.unige.ch/ressem.html


Karen Longden Roure
Program Coordinator
MSc. in Management, MSc. in Economics, MSc. in Statistics
Université de Genève, Uni-Mail
Faculté d'Economie & Management, GSEM
40, bd. du Pont d'Arve, 1211 Genève 4
Tél:  +41.22.379.8109 (10h-14h)


-- 
Prof. Eva Cantoni
Research Center for Statistics and
  Geneva School of Economics and Management
University of Geneva, Bd du Pont d'Arve 40, CH-1211 Genève 4
gsem.unige.ch/rcs/cantoni