Extending the validity of frequency domain bootstrap methods to general stationary processes (Q2215743)

The paper focuses on the development of a hybrid periodogram bootstrap (HPB) and on a convolved subsampling approach called the convolved bootstrapped periodograms of subsamples (CBP). The HPB is a new frequency domain bootstrap method that combines features from a special multiplicative approach and the CBP procedure. The second section is devoted to the presentation of some results on spectral means and ratio statistics and specifies the assumptions imposed on the stochastic process. The hybrid bootstrap procedure is developed in the third section. HBP is conceived as a hybrid of two procedures. The one is the already known multiplicative periodogram bootstrap (MPB) and the second one is the new algorithm developed in the paper, the CBP. In the fourth section the asymptotic validity is investigated. Simulation results are more extensively discussed and are to be found in the last section.