Article ID | Journal | Published Year | Pages | File Type |
---|---|---|---|---|
5096020 | Journal of Econometrics | 2014 | 15 Pages |
Abstract
This paper studies taper-based estimates of the spectral density utilizing a fixed bandwidth ratio asymptotic framework, and makes several theoretical contributions: (i) we treat multiple frequencies jointly, (ii) we allow for long-range dependence or anti-persistence at differing frequencies, (iii) we allow for tapers that are only piecewise smooth or discontinuous, including flat-top and truncation tapers, (iv) we study higher-order accuracy through the limit distribution's Laplace Transform, (v) we develop a taper-based estimation theory for the spectral distribution, and show how confidence bands can be constructed. Simulation results produce quantiles and document the finite-sample size properties of the estimators, and a few empirical applications demonstrate the utility of the new methods.
Keywords
Related Topics
Physical Sciences and Engineering
Mathematics
Statistics and Probability
Authors
Tucker S. McElroy, Dimitris N. Politis,