Of copulas, quantiles, ranks and spectra: an \(L_{1}\)-approach to spectral analysis (Q2348726)

The authors introduce an alternative method for spectral analysis of univariate stationary time series. The spectral analysis is based on two spectral kernel densities: Laplace and copula. According to these spectral kernel densities, estimation is based on Laplace and copula periodograms. Asymptotic properties of Laplace and copula periodogram kernels are introduced and discussed in Section 3. Smoothed versions of the introduced periodograms are considered in Section 4. In the last section, the authors provide some simulation studies to illustrate properties of the earlier obtained estimators.