The stationary seasonal hyperbolic asymmetric power ARCH model

Most financial time series exhibit seasonality, persistence (hyperbolic decay of the autocorrelation function), asymmetric behavior and leptokurticity. The paper introduces the stationary seasonal hyperbolic APARCH model, which can take into account these previous features. Particularly, we examine sufficient and necessary conditions for existence of strict and weak stationary solution. After looking for long memory property of the process, we provide the expression of the likelihoods, in order to estimate the parameters, in three classical cases which appear as particular case of the hyperbolic likelihood. © 2007 Elsevier B.V. All rights reserved.