Long Seasonal Cycle Modeling

Jiří Procházka, Milan Bašta, Samuel Flimmel, Matej Čamaj

Long Seasonal Cycle Modeling

Číslo: 1/2018
Periodikum: Statistika

Klíčová slova: Time series with long seasonal periods are very common. Several methods have been proposed for modeling of long seasonal cycles, the most commonly used ones being those based on basis expansion. In this paper, we present and discuss these methods. We also use them to model seasonality in realized volatility of several major stock market indices and find evidence for the existence of yearly as well as weekly seasonality. The pr

Pro získání musíte mít účet v Citace PRO.

Přečíst po přihlášení

Anotace: Time series with long seasonal periods are very common. Several methods have been proposed for modeling

of long seasonal cycles, the most commonly used ones being those based on basis expansion. In this paper, we
present and discuss these methods. We also use them to model seasonality in realized volatility of several major
stock market indices and find evidence for the existence of yearly as well as weekly seasonality. The presented
approaches can potentially be used for modeling of any seasonal time series with a long seasonal period.