Frequency domain minimum distance inference for possibly noninvertible and noncausal ARMA models (Q1750279)

The authors introduce a frequency domain minimum distance procedure for the estimation of autoregressive moving average models which are possibly noninvertible and noncausal. They use information from higher-order moments to achieve identification on the location of the roots of AR and MA polynomials for non-Gaussian time series and propose a minimum distance estimator that optimally combines the information contained in the second, third and fourth moments. The proposed estimator is consistent under some conditions.