Semiparametric generalized long-memory modeling of some mena stock market returns: A wavelet approach

This paper proposes a new class of semiparametric generalized long-memory models with FIAPARCH errors that extends the conventional GARMA model to incorporate nonlinear deterministic trend and allows for time-varying volatility. To estimate the parameters, we implement a wavelet theory. We provide an empirical application to some MENA stock markets and find that the proposed model offers an interesting framework to describe seasonal long-range dependence and nonlinear trend in return as well as persistence to shocks in conditional volatility. The predictive results also indicate that this model outperforms the traditional FARMA-FIAPARCH process.