Abstract:
We consider the limiting distribution of sample covariances computed from N x T panel data, comprising N independent copies of a random-coefficient AR(1) time series of length T. We prove that sample covariances may have a variety of stable and non-stable limits depending on how the density function of a random autoregressive coefficient behaves near the unit root and how N, T increase mutually. This is a continuation of earlier work on limits of the sample mean of the same random-coefficient AR(1) panel model. Joint work with Remigijus Leipus, Anne Philippe and Donatas Surgailis.
Biography:
Vytaute Pilipauskaite is a Postdoc at Aarhus University (Denmark). She obtained her Ph.D. in 2017 from Vilnius University (Lithuania) and University of Nantes (France) under the supervision of Donatas Surgailis and Anne Philippe. Vytaute Pilipauskaite's research interests lie in limit theorems, statistical inference for stochastic processes with long-range dependence.
Seminar by the NYU-ECNU Institute of Mathematical Sciences at NYU Shanghai