Title: A comparative study on autoregressive models: An application to several financial assets
Authors: Bing Xiao - UCA University Clermont Auvergne (France) [presenting]
Marie-Eliette Dury - UCA University Clermont Auvergne (France)
Abstract: It has become more important for financial institutes to capture the volatility of a financial asset. The autoregressive conditional heteroscedasticity models assume that volatility is not constant. Actually, some of the results of research in the models performance seem conflicting and confusing. It seems that the Students $t$ distribution characterizes better the heavy-tailed returns than the Gaussian distribution. Assets with higher kurtosis are better predicted by a GARCH model with Students distribution while assets with lower kurtosis are better forecasted by using an EGARCH model. Moreover, stochastic models such as stable processes appear as good candidates to take heavy-tailed data into account. These observations lead us to determine how well these different models perform in terms of forecasting volatility and will be assessed based on the forecasts they make. We attempt to model and forecast the volatility of different asset Index during the recent period. The questions to ask are: Which models capture the volatility better? If one model captures the volatility better, would it lead to a more efficient forecast accuracy? The assets concerned are the following: the gold market, the nickel market, the equity market and the maize market. This study contributes to the existing finance literature by investigating the U.S. stock market during the recent period.