On Monte Carlo Estimation Of Relative Power  

Oleh:

Paruolo, Paolo

Jenis: Article from Journal - ilmiah internasional
Dalam koleksi: The Econometrics Journal vol. 5 no. 1 (2002), page 65-75.
Topik: Monte Carlo; Design of experiments; (local) Power; Cointegration; Likelihood ratio; Unit roots.
Fulltext: 65.pdf (104.97KB)

Isi artikel

This paper derives standard errors for Monte Carlo (MC) estimators of (relative) power of tests when the critical values under the null have also been estimated. This situation is common, for example, in unit root and cointegration (CI) tests. The associated issue of MC design is discussed. The results are illustrated on likelihood-based tests for CI rank determination.

Opini Anda

Klik untuk menuliskan opini Anda tentang koleksi ini!

Kembali