Parameter Estimation for ARMA Models with Infinite Variance Innovations

On Business Cycle Asymmetries in G7 Countries

“…Maximum likelihood estimation of mixed ARMA models with stable errors poses a challenge, although the Whittle estimator (Mikosch et al, 1995) and minimum dispersion estimators (Brockwell and Davis, 1991) have been used in this context.…”

Stable-GARCH Models for Financial Returns: Fast Estimation and Tests for Stability

“…models in essentially the same way as under a Gaussian assumption. Examples include linear and nonlinear regression (see [75][76][77][78]), ARMA time series (see [79][80][81]), and GARCH-type models (see [24,56]; and the references therein).…”

Some computational aspects of Gaussian CARMA modelling

“…Among typical assumptions about the distribution, is that the ε t 's, are independent zero mean normal, or white-noise with finite variance. Relaxing the finite variance assumption is discussed in Mikosch, Gadrich, Kluppelberg & Adler (1995). Frequently the ARMA(p,q) 1 process is stated in terms of polynomials of the backward operator B (or L), BY t = Y t−1 .…”