The Mathematical Institute, University of Oxford, Eprints Archive

Time Series Model for Forecasting Intraday Volatilities

Serknas, Dominykas (2013) Time Series Model for Forecasting Intraday Volatilities. Masters thesis, Oxford University.

[img]
Preview
PDF (D Serknas Thesis) - Submitted Version
869kB

Abstract

The aim of this dissertation is to construct different intraday volatility forecasting techniques for the future contracts and to evaluate their accuracies in terms of the forecasting errors. In order to achieve these goals the intraday volatility levels are assumed to be reflected by the absolute values of the intraday returns. The periodic intraday volatility factor is modelled by different regression methods and the future intraday absolute returns are forecasted. Since the intraday returns depend on the overall volatility of the particular day, Exponential Weighted Moving Average and GARCH(1,1) techniques are implemented to predict these quantities. Finally, the results reveal that the most accurate method for the intraday volatility forecasting is based on modelling each 5 min interval during the day as an independent variable in the periodic intraday component and this holds not only for the next day forecasting but also for the longer time horizon.

Item Type:Thesis (Masters)
Subjects:H - N > Mathematics education
Research Groups:Mathematical and Computational Finance Group
ID Code:1741
Deposited By: Laura Auger
Deposited On:13 Aug 2013 19:12
Last Modified:29 May 2015 19:26

Repository Staff Only: item control page