A0651
Title: Time domain estimation of non-fundamental ARMA models in the presence of heteroskedasticity of unknown
Authors: Carlos Velasco - Universidad Carlos III de Madrid (Spain) [presenting]
Ignacio Lobato - ITAM (Mexico)
Abstract: Time domain estimation of possibly non-fundamental (that is, non-causal and/or non-invertible) non-Gaussian linear ARMA models with martingale difference innovations that may display conditional heteroskedasticity of unknown form is considered. Instead of explicitly parametrizing the underlying volatility process (the higher order dependence) and employing maximum likelihood procedures, the unpredictability of the true model innovations is exploited, and a time domain minimum distance estimator is proposed based on innovations predictability using second and third powers of past innovations. Using the proposed estimator, which is consistent and asymptotically normal, an ARMA(1,1) model is fit to the squares of exchange rate returns, and evidence of non-invertibility in three out of seven cases is found.
